A Case of Haemochromatosis and Diabetes A Missed Opportunity

Refer to Page 80- Poster 176
Clinical care and other categories posters: audit
P150
Contrasting relationships between sleep
duration and central obesity in South Asians
and White Europeans with Type 2 diabetes
AA Tahrani1,2, K Dubb1, S Begum2 and MJ Stevens1,2
1
Centre of Endocrinology, Diabetes and Metabolism, University of
Birmingham, Birmingham, UK, 2Department of Diabetes and Endocrinology,
Birmingham Heartlands Hospital, Birmingham, UK
Background: Disorders of sleep have been linked to obesity in the

non-diabetic population; however, this relationship has not been
studied in Type 2 diabetes. Our aim was therefore to assess the
relationship between sleep duration and central obesity in patients
with Type 2 diabetes.
Methods: This was a cross-sectional study of consecutively
recruited patients from the diabetes clinic of a secondary care
hospital. Sleep duration was based on total sleep time (TST)
recorded during a multichannel cardiorespiratory study (Alice
PDX, Phillips Respironics) to assess the apnoea hypopnea index
(AHI). TST was defined as sleep duration minus arousals/
DIABETICMedicine
awakenings. Central obesity was assessed using waist

circumference (WC). Data were analysed for White Europeans
(WE) and South Asians (SA) separately.
Results: In all, 120 WE [mean age 59.3 10.4years, diabetes
duration 11.5 7.4years, body mass index (BMI) 36.8 8.5kg/
m2, WC 119.4 15.8cm, men 59.2%] and 99 SA (age
54.9 12.3 years, diabetes duration 13.0 7.9years, BMI
30.7 5.5kg/m2, WC 106 12.7cm, men 55.6%) were
included in the analysis. TST was similar in WE and SA
(398.4 108.2 vs 388.9 98.8min, p = 0.5). TST correlated
with WC positively in SA (r = 0.22, p = 0.03) and negatively in
WE (r = 0.24, p = 0.008). After adjustment for age, gender,
diabetes duration, AHI, smoking, alcohol intake, insulin treatment
and incretin-based therapy, TST remained independently
associated with WC in SA (B = 0.03, p = 0.04) and WE
(B = 0.03, p = 0.02).
Conclusions: Shorter sleep duration in WE and longer sleep
duration in SA are independently associated with increased central
obesity in Type 2 diabetes. The reasons for the ethnic differences
need to be studied. Whether sleep duration manipulation would
impact central obesity remains to be examined.

P151
New onset hyperglycaemia following acute
coronary syndrome: a re-audit
S Guthrie and P De
Department of Diabetes and Endocrinology, City Hospital, Sandwell and West
Birmingham NHS Trust, Birmingham, UK
Aims: To investigate degrees of glucose intolerance amongst

patients without diabetes admitted to our cardiology unit with
acute coronary syndrome(ACS) and compare with the previous
audit we did in 2009 2010 to see if we could identify any
differences in trend.
Methods: All patients admitted to the cardiology unit with ACS
(October 2010 to June 2013) underwent fasting plasma glucose
and oral glucose tolerance test (OGTT) on day 3 of admission and
had a repeat OGTT 2 weeks later following discharge.
Results: Of these patients, 21% were Caucasian, 62% South
Asians and 17% Afro-Caribbean (N = 65); 35% (23/65) and 12%
(8/65) had impaired glucose tolerance (IGT) and diabetes
respectively (52% and 38% respectively in 2009 2010). Repeat
OGTT showed the majority still continued to have IGT and
diabetes (27% and 25% respectively in 2009 2010). One patient
with IGT progressed to frank diabetes and two developed IGT
from normal glucose tolerance. 50% who developed diabetes were
South Asians (all males, mean age 51 years). 65% of those
developing IGT were South Asians, mostly males (11/15), mean
age 57 years. Non Asians who developed IGT were slightly older
(mean age 62 years).
Conclusions: Incidence of IGT and undiagnosed diabetes among
our multi-ethnic population was much less compared with our
earlier audit of 2009 2010 presented in Diabetes UK 2012.
Additionally, those with IGT/diabetes on initial testing did not
show much change in their overall glycaemic status on re-testing.
Lifestyle advice and education (X-PERT course) have been helpful
but need to be continually reinforced, and these high risk patients
need early identification and aggressive treatment.
2014 The Authors.

Diabetic Medicine 2014 Diabetes UK. Diabetic Medicine, 31 (Suppl. 1), 28183
P152
An audit to determine and improve the
attendance of the X-PERT programme among
a Greater Manchester general practice
population of patients with Type 2 diabetes
V Grannon1 and J Hider2
1
Central Manchester Foundation Trust, Manchester, UK, 2Partington Health

Care Centre, Manchester, UK
Objectives: Type 2 diabetes is usually a result of obesity and is

exacerbated by inactivity and overeating. It carries significant
mortality and the prevention of long-term complications is a
priority. The necessary lifestyle changes needed to achieve this
make patient education paramount. All patients with Type 2
diabetes should be offered a referral to an education programme.
X-PERT provides a structured approach to patient education and
meets the key criteria set by the Department of Health. This audit
aims to investigate and improve the referral and attendance rates to
X-PERT among a general practice population.
Method: The popualtion of a general practice was searched to
identify patients over the age of 16 with Type 2 diabetes.
Documented evidence of referral and attendance or failure to
attend the X-PERT programme was sought. Questionnaires were
distributed to investigate possible low attendance rates.
Results: In all, 245 patients were identified; 18% (n = 44) had
discussed X-PERT during a consultation and 78% (n = 34)
went on to be referred; 24% (n = 8) of those referred
attended. Returned questionnaires revealed the main reasons for
non-attendance were a lack of transport and working hours.
Conclusion: The rates of referral and attendance to X-PERT
among this population was low. Possible explanations for this
finding are poor documentation of referrals or a lack of emphasis
on the need to refer. To improve patient awareness notice boards
were placed in the waiting room. To improve practice awareness a
presentation was prepared to discuss these results and READ codes
were introduced to aid documentation. Providing a nearby session
out of working hours was suggested.
73
DIABETICMedicine
Abstracts of the Diabetes UK Professional Conference, 57 March 2014
P153
An audit investigating how clinical
outcomes on pump services can be achieved
through intensive follow-up via face-to-face
and distance consultation
S Salehi, H Addington, K Turner and K Vithian
Diabetes Centre, Colchester Hospital University NHS Foundation Trust,
Colchester, UK
Aims and objectives: NICE supports using insulin pump therapy

as an alternative to multiple daily injections in selected patients
with Type 1 diabetes. We conducted an audit in 2011 2012
which showed that 31% of patients were not achieving
HbA1c improvements despite commencing insulin pump
therapy. Following this we offered more intensive follow-up via
face-to-face and phone/email contacts and incorporated more
advanced pump features. We re-audited this year to see if this has
resulted in improved clinical outcomes.
Methods: A questionnaire was provided to all patients attending
the pump clinic providing qualitative information such as
hypoglycaemia awareness and quality of life. HbA1c data were
obtained from the trust computer system. Information
on telephone support and clinic time was also collected.
Results: Sixty patients (17 males, 43 females) who had completed
12 months of pump treatment were included in the audit. 85.7%
of patients who initiated pump therapy for hyperglycaemia had
improvements in HbA1c (cf. 69% in the 2011 2012 cohort) with
mean HbA1c reduction being 12.9mmol/mol (13.8mmol/mol in
the 2011 2012 cohort). Non-responders were predominantly
males (80%) and tended to be younger (33 years vs 43 years).
83% of patients who initiated pump therapy for hypoglycaemia
regained awareness (cf. 43% in the 2011 2012 cohort). 85% of
patients reported an improvement in quality of life since starting
pump therapy.
Conclusions: Our audit demonstrates that intensive follow-up
through a dedicated pump service is having a profound effect on
clinical outcomes. However, the audit identifies that a more
focused approach is necessary for younger males.
P154
The use of GLP-1 agonist liraglutide:
adhering to guidelines?
J Fatima and S Pearce
Diabetes, Wrightington, Wigan and Leigh NHS Trust, Wigan, UK
Aims: To assess whether the initiation and continuation of

liraglutide complies with NICE guidelines in a district general
hospital.
Methods: A retrospective medical record audit of 43 patients
with Type 2 diabetes commenced on liraglutide between October
2010 and March 2012. Data were collected at baseline and at 6
months on HbA1c, weight, withdrawal rates and reasons for
continuation of liraglutide at 6 months for patients not meeting
NICE criteria.
Results: The mean diabetes duration of patients was 10.7 years,
mean initial HbA1c 8.7% and mean body mass index 38.9kg/m2.
The mean HbA1c change was 1.2% and mean weight change
was 5.6kg. Three patients stopped treatment due to side effects.
At 6 months, 52% achieved an HbA1c 1% decrease. 69% lost
3% of initial weight. 42% achieved both criteria. 26% achieved
neither criterion. Liraglutide was only stopped in 7% of our patient
population. The reasons for continuation were not always clearly
documented but common reasons were patient preference for
74
liraglutide, a positive outcome in terms of weight loss and patients

and clinicians expecting improvement after a longer time frame
than 6 months.
Summary: Liraglutide was being continued despite patients not
meeting both NICE criteria for continuation. Clinicians agreed to
adhere more closely to NICE guidelines and justify and document
any use outside of guidance, as this is an expensive drug. Perhaps
also the guidance needs to change to allow for patients who benefit
from either weight reduction or HbA1c reduction to continue on
liraglutide.
P155
Diabetes in the very elderly: need for an
individualised approach
S Bellary1,2 and A Oluwatoyin1
1
Aston Research Centre for Healthy Ageing, Aston University, Birmingham,
UK, 2Diabetes Department, Heart of England NHS Foundation Trust,
Birmingham, UK
Background: Intensive risk factor management is recommended

for individuals with diabetes. However, it is not known if such an
approach is appropriate in the elderly with multiple comorbidities
and limited life expectancy. The aim of this study was to
characterise a cohort of very elderly individuals with diabetes
and assess the impact of known risk factors on mortality.
Methods: This was a retrospective audit approved by the clinical
audit lead. All patients aged >80 years who attended diabetes
outpatient clinics 2 years prior to the date of the audit (April 2012)
were identified from clinic records. A detailed history including
demographics, comorbidities and treatment were collected. Blood
pressure readings, HbA1c, cholesterol and renal function were
extracted and the mean of these readings was recorded. Survival
status at 2 years was recorded for all patients. Statistical analysis
was performed using SPSS19.
Results: Data were available for 864 (381 male, 483 female)
patients. The majority (75%) lived in their own home. More
than 60% had multiple comorbidities and 25% had a prior history
of cardiovascular disease. Two-thirds of the patients had more
than one hospital admission in 2 years and a third had more than
three admissions. 60% were on either insulin or a sulfonylurea.
Mean HbA1c was 7.6%, cholesterol 4.2mmol/l, systolic blood
pressure 145mmHg and eGFR 53ml/min. Over 2 years, 174 (20%)
had died. Age, creatinine and previous coronary heart disease were
significant predictors of death.
Conclusion: The benefits of intensive diabetes management
appear to be uncertain in very elderly patients. The need for
intensive treatment must therefore be individualised to each
patient.
P156
Influenza vaccine effectiveness against
hospitalisation and death in people with
Type 2 diabetes
EP Vamos1, UJ Pape1, V Curcin2, MJ Harris1, J Valabhji3,
A Majeed1 and C Millett1
1
Department of Primary Care and Public Health, Imperial College London,
London, UK, 2Department of Computing, Imperial College London, London,
UK, 3Department of Diabetes and Endocrinology, St Marys Hospital, Imperial
College Healthcare NHS Trust, London, UK
Objectives: The clinical benefits of influenza vaccination in older

people and those with chronic conditions are subject to ongoing
debate. This study aimed to examine the effectiveness of influenza
2014 The Authors.

vaccination against hospitalisation for acute cardiovascular and

respiratory conditions and all-cause mortality in people with Type
2 diabetes, while assessing and adjusting for residual confounding
by using estimates obtained for the off-season.
Methods: In this retrospective cohort study, we used data from
the Clinical Practice Research Database between 2003 2004 and
2009 2010. Rates of hospitalisation for acute myocardial
infarction (AMI), stroke, heart failure and pneumonia/influenza
and all-cause mortality were compared in the vaccinated and
unvaccinated cohorts.
Results: There were 124,503 patients with Type 2 diabetes
contributing to 623,591 person-years of observation during the
study period. Influenza vaccine recipients were older and had more
comorbid conditions compared with non-recipients. After
adjustment for covariates and residual confounding, vaccination
was associated with 21% lower hospitalisation rates for AMI [IRR
0.79 (0.64 0.98)], 30% reduction for stroke [IRR 0.70
(0.53 0.91)], 20% reduction for heart failure [IRR 0.80
(0.66 0.94)] and 15% reduction for pneumonia or influenza
[IRR 0.85 (0.74 0.99)]. Vaccinated patients had 28% lower
all-cause mortality compared with non-recipients [IRR 0.72
(0.64 0.82)].
Conclusions: Influenza vaccination was associated with
significant reductions in hospitalisation rates for major acute
cardiovascular and respiratory conditions and all-cause mortality
in people with Type 2 diabetes. Whilst the development and
approval of new vaccines and new approaches are in progress,
efforts should be focused on improvements in vaccine uptake in
this important target group.
P157
Achievement of treatment targets: reaching
a plateau? Results from the National
Diabetes Audit in England
B Young1, N Holman2, L Dunn3, A Uddin3, J Hope3 and
R Gadsby4
1
Diabetes and Endocrinology, Salford Royal Hospital, Salford, UK, 2Public
Health England, York, UK, 3Health and Social Care Information Centre, Leeds,
UK, 4Institute of Diabetes in Older People, University of Bedfordshire, Luton, UK
Aim: To assess the change in the proportion of people with

diabetes meeting the NICE recommended HbA1c, blood pressure
and cholesterol targets between 2003 2004 and 2010 2011. The
Quality and Outcomes Framework was introduced in 2004 and
incentivised the achievement of HbA1c, blood pressure and
cholesterol targets in general practice.
Methods: The National Diabetes Audit (NDA) has collated data
on patients with diabetes from primary care in England since
2003 2004. These have been supplemented by data submitted by
specialist services. Data collected include whether key care
processes have been performed and the results of HbA1c, blood
pressure and cholesterol measurements.
Results: The number of people included in the NDA has risen
from 253,000 in 2003 2004 to 2,235,810 in 2010 2011
(representing 87.6% of people with diagnosed diabetes). In the
2003 2004 audit period 56.2% of people with diabetes had an
HbA1c of 7.5% (58mmol/mol) or less. By 2006 2007 this had
risen to 62.6%. However, by 2010 2011 this proportion had only
increased marginally to 63.3%. A similar pattern is shown in the
proportion of people achieving the NICE blood pressure and
cholesterol targets. There is considerable variation between clinical
commissioning groups(CCGs) and between practices.
2014 The Authors.

DIABETICMedicine
Conclusion: Over recent years there have been substantial

improvements in the attainment of NICE treatment targets for
people with diabetes. However, this improvement appears to have
reached a plateau. Some national improvement could undoubtedly
be achieved if the bottom quartile of CCGs and practices improved
to the mean. Substantial further improvements are likely to require
significant innovations in diabetes care delivery.
P158
Referral pattern and outcome in elderly
inpatients with diabetes: impact of the
diabetes inpatient team on outcome and
glycaemic control
AA Syed, H Maurice, C Holmes, N Taylor and A Mahmood
Diabetes, Heartlands Hospital, Birmingham, UK
Background: Inpatients with diabetes is a common problem

(10% 15% of inpatients). They have a longer length of stay (up to
2.6 days). Clinical studies suggest specialist diabetes inpatient
teams can improve patient outcomes, reduce length of stay,
increase day case rates and reduce the number of admissions.
Economic analysis suggests that the savings from the introduction
of such teams can substantially outweigh the cost of the team.
Objectives: To investigate the referral pattern, the outcome
of interventions following inpatient team review, and HbA1c
pre- and post-intervention.
Methods: An analysis of the data collected on the referrals by the
inpatient diabetes team was carried out.
Results: In all, 106 patients were reviewed in the elderly care
wards in a 6 month period with mean age 78.3 years. 58.5% were
female and 20% Asian. A third of the referrals were due to
hypoglycaemia. 49.1% of the intervention included change in the
treatment and 10.1% needed insulin starts. Mean HbA1c in 84
patients prior to admission was 8.5%, and mean HbA1c in 42
patients 6/12 post admission was 7.7% respectively. In 34 patients
who had pre and post HbA1c the difference in HbA1c was 1.08%.
Conclusion: A significant proportion of elderly patients admitted
had poor glycaemic control. A third of patient referrals were due to
low BMs. Nearly half of the patients were on insulin or a
combination of insulin and OHAs. More than 85% of referred
patients needed intervention.HbA1c was not done in 42 patients
post-discharge. Significant improvement in HbA1c(>1%) was
noted following intervention by the inpatient diabetes team.
P159
SCI-Diabetes inpatient tracking system:
could this support a Scottish National
Diabetes Inpatient Audit?
GM Brennan1, D Voigt1 and C Flach2
1
Diabetes Centre, Ninewells Hospital, Dundee, UK, 2Clinical Technology
Centre, Ninewells Hospital, Dundee, UK
Background: Successive National Diabetes Inpatient Audits

(NaDIAs) have provided important insights into inpatient
diabetes care in England and Wales. However, these have been
labour intensive as they involved manual data collection.
Aims: SCI-DC is the shared electronic record for the care of
people with diabetes in NHS Scotland, used in all outpatient
settings. We extended the function of SCI-DC to permit
identification of all patients with diabetes admitted to our hospital.
Methods: SCI-DC receives admission, discharge and transfer
messages from the hospital Patient Administration System which
75
DIABETICMedicine
identifies where an individual resides during inpatient care. Patients

are located within a clinical domain on a secure system.
Admission date, patient demographics, usual diabetes treatment
and clinical data are available and length of stay can be calculated.
A pilot system was implemented in NHS Tayside and used to
support a snapshot audit in 11 medical wards of Ninewells
Hospital (25 29 March 2013).
Results: Forty-two patients were audited: 35 (83%) aged >65
years; six (14.2%) admitted due to diabetes (four hyperglycaemia,
one hypoglycaemia, one foot ulcer); 36 (85.7%) due to other
illnesses; 23 (55%) treated with insulin; 25 (59.5%) had medium
to high risk feet or active foot disease prior to admission.
Conclusions: In keeping with NaDIA, most patients were
admitted for other illnesses and the majority were elderly. The
next phase locally is to use the data to support service redesign.
This initiative is supported by the Scottish Diabetes Group and
Scotland-wide implementation is planned, which could facilitate a
Scottish NaDIA in due course.
P160
Management of diabetic ketoacidosis:
comparison against standards using an
online audit tool
S Parthasarathy1, K Ellis1, A Smith2, A Crown1 and N Vaughan1
1
2
Diabetes and Endocrinology, Royal Sussex County Hospital, Brighton, UK,

Diabetes and Endocrinology, Princess Royal Hospital, Haywards Heath, UK
Aim: The Joint British Diabetes Societies (JBDS) guidelines

(March 2010) were used as standards to evaluate management of
diabetic ketoacidosis (DKA) in a tertiary centre.
Methods: A retrospective audit of a 17 month period was done
using an online tool using php computer programming language in
conjunction with a MySQL database. Live analysis of data could
be tracked throughout data entry with results to the question posed
already available to view. Of the 113 cases centrally coded as
DKA, 75 cases were analysed.
Results: Only 60% (n = 44) of the coded notes actually met
diagnostic criteria for DKA with 11% of patients outside the
catchment area. Nearly half of the patients met criteria for
intensive monitoring in level 2/3 settings of whom 91% were
appropriately referred. Initial management within the first hour
with intravenous fluids and insulin was suboptimal (39% and 25%
respectively) with monitoring of potassium being achieved in 85%
of patients in the first 24h. Amongst patients previously on a basal
bolus regime, 58% received their basal insulin whilst on variable
rate intravenous insulin. Mean length of stay was 3 days with 19%
who had followed sick day rules prior to admission. Only 57%
were seen by a diabetes specialist nurse before discharge with
diabetes physician follow-up being arranged in 87% of cases.
Conclusion: Management of DKA remains suboptimal needing
robust protocols to achieve favourable outcomes. This online tool
can potentially be used for re-auditing in real time. This audit lays
emphasis on education amongst other healthcare physicians,
patient education and specialist nurse led services in improving
standards.
Clinical care and other categories posters: cardiovascular

P161
Cardiovascular disease and biomarkers of
liver injury in Type 2 diabetes: the
Edinburgh Type 2 Diabetes Study (ET2DS)
JR Morling1, JA Fallowfield2, RM Williamson3, CM Robertson1,
LD Nee4, S Glancy4, IN Guha5, MWJ Strachan3 and JF Price1
1
Centre for Population Health Sciences, University of Edinburgh, Edinburgh,
UK, 2Queens Medical Research Institute, University of Edinburgh, Edinburgh,
UK, 3Metabolic Unit, Western General Hospital, Edinburgh, UK, 4Radiology
Unit, Western General Hospital, Edinburgh, UK, 5Digestive Diseases
Biomedical Research Unit, University of Nottingham, Nottingham, UK
1.28, p = 0.007] and CAD (OR 1.36, p = 0.001) after adjustment

for established CV risk factors. After a mean follow-up of 4.4 years
from baseline attendance there were 121/1,035 (11.7%) incident
CVD and 76/1,035 (7.3%) incident CAD events. Again, only GGT
showed an independent association (CAD, OR 1.35, p = 0.042).
Conclusion: In this population, GGT is independently associated
with both prevalent and incident CV events and this association
persists into older age. Our work supports CAD as the main
element of association within the spectrum of CVD. There was no
statistically significant relationship between CV events and
biomarkers of hepatic steatosis, inflammation or fibrosis in our
study.
Objective: Patients with chronic liver disease (CLD) are said to

have higher cardiovascular (CV) mortality rates than the general
population. There has been little investigation of the relationship
between biomarkers of CLD and CV disease (CVD) in
community-based populations with diabetes. We examined the
association of a range of validated biomarkers of CLD with both
prevalent and incident CV events and mortality.
Design: In
all,
1,066
participants
in
the
ET2DS
(a population-based cohort with Type 2 diabetes aged 60 75
years) were followed up for CVD. Validated markers of CLD were
measured: liver enzymes [alanine aminotransferase, aspartate
aminotransferase, gammaglutamyl transferase (GGT)], platelet
count, liver injury (cytokeratin-18), liver fibrosis (Enhanced Liver
Fibrosis panel) and steatosis.
Results: There were 372/1,035 (35.9%) patients with prevalent
CVD and 319/1,035 (30.8%) with prevalent coronary artery
disease (CAD) at baseline. In multivariate analysis only GGT was
independently associated with prevalent CVD [odds ratio (OR)
Aim: To investigate CVD and its risk factors in people with Type
2 diabetes of different ethnicity living in Scotland.
Methods: Data from Scottish Care Information
Diabetes
Collaboration (SCI-DC) were linked to Scottish Morbidity
Records (SMR01) and National Records of Scotland for
mortality. Of 156,991 people with Type 2 diabetes with coded
76
2014 The Authors.

P162
Ethnicity and risk of cardiovascular disease
(CVD): 4.8 year follow-up of patients with
Type 2 diabetes in Scotland
MO Malik and SDRN Epidemiology Group
British Heart Foundation (BHF) Glasgow Cardiovascular Research Centre,
University of Glasgow, Glasgow, UK
Clinical care and other categories posters: case reports
ethnicity, 121,535 (77.4%) had no CVD at baseline (2005 or

diabetes diagnosis) (114,461 White, 2,554 multiple ethnic, 797
Indian, 319 other Asian, 2,250 Pakistani, 387 Chinese, 301
Afro-Caribbean and 466 other ethnic) and were followed up
(mean SD: 4.8 2.3 years) for the development of CVD
[ICD10 codes: I20I25, I6069 (excluding I62, I68)].
Results: 21,089
(17.3%)
patients
developed
CVD.
Coxproportional hazard ratios (HRs) for Pakistanis (HR = 1.46,
CI 1.271.68, p < 0.001) were significantly higher than for White
after adjustment for age, sex, conventional cardiovascular and
other risk factors. No other group had significant differences from
White except for Chinese (HR = 0.64, CI 0.420.97, p = 0.036).
At baseline Pakistanis were younger (53.8 12 years vs White
63.3 13 years, p < 0.01) but had developed diabetes earlier (age
of onset, mean SD, 48.8 12 years vs White 59.3 13 years,
p < 0.01), had higher HbA1c (69 22mmol/mol vs White
62 22mmol/mol, p < 0.05), more advanced antidiabetic
treatment (insulin 19.3% vs White 17.4%, other antidiabetic
drugs 75.2% vs 69%, p < 0.01) and longer duration of diabetes
(10 7 years vs White 8.6 7 years, p < 0.01). However,
Pakistanis had lower blood pressure, body mass index,
creatinine,
proportion
smokers
and
proportion
on
antihypertensive therapy than White.
Conclusion: People with Type 2 diabetes of Pakistani ethnicity in
Scotland are at increased risk of CVD. This is not explained by
conventional risk factors.
DIABETICMedicine
P163
Management of hyperglycaemia in patients
with acute coronary syndrome
P Hegde, N Ormsby, FM Kaivani and D Bowen-Jones
Department of Diabetes and Endocrinology, Wirral University Teaching
Hospital, Wirral, UK
Introduction: Hyperglycaemia is associated with higher mortality

in patients with acute coronary syndrome (ACS). Conflicting
evidence from studies in this area has resulted in a lack of clarity
over optimal treatment regimes.
Aim: To quantify how this lack of evidence and the inconclusive
guidelines impact the management of hyperglycaemia.
Methods: This is a retrospective case notes analysis of 50 randomly
selected patients between October 2011 and March 2012 at Wirral
University Teaching Hospital against the NICE guidelines.
Results: Of the 50 patients (32 male, 18 female), average age 74.5
years, 43 patients (86%) were known to have diabetes and seven (4%)
were not at the time of admission. Of the 43 patients 14 were on diet,
22 were on oral hypoglycaemic agents and seven were insulin treated.
Thirty-one patients had admission blood glucose above 11mmol/l and
17 did not have any treatment within 48h despite having
hyperglycaemia. Five patients received glucose potassium insulin
infusion and six had their usual treatment. Only 34 patients had blood
glucose chart and regular glucose monitoring. For 22 patients blood
glucose was out of the target range in the first 48h. Only 10 patients
were referred to the diabetes team during their admission.
Conclusion: Our audit demonstrates the lack of knowledge and
inconsistency in managing hyperglycaemia in patients with ACS.
This clearly prompts the need for a robust and comprehensive
guideline in managing this area. Current NICE guidelines are
unhelpful.

P164
Treating fat loss with weight loss:
Roux-en-Y gastric bypass in a patient with
familial partial lipodystrophy
P166
Four cases of radiological diabetic foot
osteomyelitis prior to clinical presentation
CL Hames, C Jenkins-Liu, J Harris, C Adams, RK Semple,

DB Savage, S ORahilly and A Stears
1
Department of Diabetes and Endocrinology, Central Middlesex Hospital,
London, UK, 2Department of Radiology, Central Middlesex Hospital, London,
UK
Department of Diabetes and Endocrinology, Cambridge University Hospitals

NHS Foundation Trust, Cambridge, UK
A Salam1, T Vakilgilani1, A Chambers2, A Hine2 and WM Kong1
Refer to Oral A55
Refer to Oral A53
P165
Diagnostic and therapeutic challenge in a
patient with mixed phenotype diabetes
P167
Use of a ketogenic diet in an adolescent
with Type 1 diabetes and epilepsy on insulin
pump therapy
A Juszczak1,2, G Thanabalasingham1,2, M Khonsari3,

A Webster1 and KR Owen1,2
APB Tasker1,2, V Alexander2, HL Mitchell2, A Jollands2,

N El Tantawi2 and H Grossi3
1
Oxford Centre for Diabetes, Endocrinology and Metabolism, University of
Oxford, Churchill Hospital, Oxford, UK, 2Oxford National Institute for Health
Research (NIHR) Biomedical Centre, Churchill Hospital, Oxford, UK,
3
Radiology Department, Oxford University Hospitals NHS Trust, Oxford, UK
1
Child Health, University of Dundee, Dundee, UK, 2Paediatics, NHS Tayside,
Dundee, UK, 3Paediatrics, NHS Grampian, Aberdeen, UK
Refer to Oral A56
Refer to Oral A54
2014 The Authors.

77
DIABETICMedicine
P168
Two cases of hepatoselective insulin being
less efficacious in patients with
non-alcoholic fatty liver disease and
hypertriglyceridaemia
MB Whyte1, A Quaglia2 and A Hopkins1
1
2
Department of Diabetes, Kings College NHS Foundation Trust, London, UK,

Department of Pathology, Kings College NHS Foundation Trust, London, UK
Introduction: We describe two cases of non-alcoholic fatty liver

disease (NAFLD) and hypertriglyceridaemia whereby detemir
appeared less efficacious than Insulatard. Very high insulin doses
were required and weightgain was problematic.
Patient 1: A 56-year-old man. Four year history of Type 2
diabetes and hypertriglyceridaemia (triglycerides 28.7mmol/l
at diagnosis). No alcohol. On examination, weight 109kg,
body mass index 32kg/m2, palpable splenomegaly. Random
triglycerides 8.4mmol/l. Normal LFTs. NAFLD and 17.7cm
spleen on CT. Bone marrow aspirate and trephine biopsy
normal. Liver biopsy: marked steatosis. Fivemonths after starting
detemir he was using 158 units/day plus NovoRapid 24 units
TDS but remained hyperglycaemic (HbA1c 8.5%). Detemir
then changed to Insulatard 70 units/day (NovoRapid dose
unchanged). Within 4 months HbA1c fell to 6.8% and by
12months he had experienced 6kg weightloss.
Patient 2: A 68-year-old man, 126kg (body mass index
44.6kg/m2). Twenty years of Type 2 diabetes and type V
hyperlipidaemia. Eighteen years previously developed necrotizing
pancreatitis; therefore oral hypoglycaemic agents changed to
Insulatard and Actrapid. Eight years ago he switched to detemir
and NovoRapid but then gained 20kg weight (detemir 120 units
BD and NovoRapid 120 units TDS; total 600 units/day). No
alcohol. Triglycerides 31.9mmol/l, total cholesterol 6.3mmol/l,
LFTs normal, HbA1c 7.2%. NAFLD on ultrasound. Detemir was
then changed back to Insulatard. After 6months, total daily dose
was 430 units/day (a 30% reduction in total daily dose) and
HbA1c 6.7%.
Discussion: Hepatoselective insulin may be less efficacious in
NAFLD and hypertriglyceridaemia. Speculatively this could be due
to increased insulin clearance, porto-systemic shunting, direct
steatosis-induced hepatocellular dysfunction or an effect of
triglycerides on albuminbinding of detemir.
P169
First presentation of autoimmune diabetes
at age 89 years
S Ahmad, M Patel, LKD Premawardhana and M Adlan
Diabetes and Endocrine, Ysbyty Ystrad Fawr, Ystrad Mynach, Hengoed, UK
A 89-year-old woman was admitted with a 3 week history of

increased thirst, polyuria and polydipsia. She was known to have
hypertension, hypothyroidism, mild paraproteinemia and CKD3.
Her sister had Type 2 diabetes requiring insulin. On admission, she
was conscious and dehydrated. Her blood pressure was 136/56,
pulse 100/min and temperature 36.5C. Clinically, there was no
goiter, and cardiovascular, chest, abdominal and neurological
examinations were normal. Her blood sugar was 48.7mmol/l, pH
7.30 and bicarbonate 19.9mmol/l. HbA1c was 72mmol/mol. Urine
and blood ketones were not documented. Subsequent autoimmune
screening showed that she had a GAD-65 level of >2,000U/ml
(reference range 0 5); she was also strongly positive for islet cell
antibodies and thyroid peroxidase antibodies. Adrenal antibody
and TSH receptor antibody screening was negative. A CT scan of
78
the pancreas was entirely normal. She was given an insulin infusion
and later twice daily subcutaneous insulin with symptomatic
improvement and resolution of metabolic abnormalities. This
subject was unusual for several reasons: (a) first presentation of
autoimmune diabetes with significant metabolic derangement at
the age of 89 years is rare; (b) her diabetes was part of an endocrine
diathesis as evidenced by the presence of autoimmune
hypothyroidism; (c) the presence of autoimmune antibodies at
very high levels is also unusual for this age group. Although the
first presentation of Type 1 diabetes is rare at the age of 89 years,
clinicians should be aware of this possibility even in the very
elderly.
P170
Diabetic enthesopathy: a distinct clinical
entity in foot pathology
F Medici1 and S Rowe2
1
Department of Metabolic Medicine, Homerton University Hospital NHS
Foundation Trust, London, UK, 2Department of Radiology, Homerton
University Hospital NHS Foundation Trust, London, UK
Aims: Acute inflammation and plantar foot deformities are

common presentations in diabetes practice. In the absence of
infection and ulceration, these findings are often attributed to
Charcots neuroarthropathy. This condition is characterised by
bone and joint surface disease. In our practice we encountered
three cases where the initial presentation was indistinguishable
from Charcots neuroarthropathy but the radiological evidence
indicated a different disease process.
Methods: This is a retrospective analysis of three cases seen in our
multidisciplinary diabetes foot service.
Results: The patients reviewed were seen between 2009 and
2011. They included two females and one male, aged 41,
58, 74 respectively. Two patients had Type 1 diabetes and one
Type 2 diabetes. All patients had obesity and sensory
neuropathy. They presented with signs of acute foot
inflammation and either already had or developed plantar foot
deformities. Serial X-rays showed subluxation of the tarsal bones
without any evidence of bone or joint surface involvement.
Conclusions: The radiological findings are consistent with purely
primary capsular/ligament disease and resulting joint subluxation.
This is in contrast with the disease process characteristic of
Charcot neuroarthropathy. We postulate that some patients with
diabetic neuropathy and obesity are at risk of severe enthesopathy
due to pressure overload and reduced sensation. The distinction
between this condition and Charcots neuroarthropathy is
important due to the prognostic and management implications.
P171
Truncal neuropathy and generalised
oedema: two rare complications of insulin
initiation in the same patient
NK Tufton, SW Coppack and MSB Huda
Endocrine and Diabetes, Royal London Hospital, Barts Health Trust, London,
UK
A 35-year-old lady, who presented with an infected, necrotic

fourth finger and was admitted for debridement and antibiotics,
developed diabetic ketoacidosis postoperatively. Past medical
history included latent autoimmune disease of adults (LADA)
treated with gliclazide 80mg bd. She was non-compliant with
medication and had had significant weight loss. She was glutamic
acid decarboxylase (GAD) antibody positive, with poor glycaemic
2014 The Authors.

control [HbA1c 133mmol/mol (14.3%)] and underweight (weight

37kg; body mass index 15.1kg/m2). She was treated with
intravenous and then subcutaneous insulin (total daily dose 74
units). Between days 3 to 5 after admission, she developed
significant pitting oedema to her umbilicus. Serum albumin
reduced to 25g/l and liver function tests were transiently
abnormal secondary to sepsis and antibiotic therapy. Oedema
persisted after normalisation of albumin, and after other causes of
oedema were excluded she was treated with furosemide 40mg od
and fluid restriction between days 10 and 17. Peak weight was
42.5kg and oedema resolved 2 weeks later. She also developed a
painful sensory neuropathy over dermatomes T12-L2 on day 5
which persisted for several months before resolution. Magnetic
resonance imaging was normal and nerve conduction studies
showed generalised diabetic sensorimotor neuropathy. It is likely
that oedema and truncal neuropathy developed secondary to
insulin initiation and rapid improvement in glycaemic control. She
was particularly susceptible due to poor glycaemic control, low
bodyweight, hepatic dysfunction and low serum albumin. This case
highlights the importance of being aware of the rare complications
of insulin therapy, particularly in susceptible patients.
P172
A description of two patients with Type 1
diabetes living in a high secure psychiatric
hospital: their journey moving from fixed
dose to self-directed variable dose insulin
using an individualised adaptation of the
Bournemouth Type 1 Intensive Education
(BERTIE) Programme and a unique
calculation tool
D Edson and T Humphries
Rampton High Secure Psychiatric Hospital, Nottinghamshire NHS Trust,
Retford, UK
The High Secure Pychiatric Hospital treats mentally unwell people

who are a grave and immediate danger to themselves and/or
others [1].It is one of only three high secure hospitals in the
country and has its own health centre with associated staff. This
project aimed to educate two patients with Type 1 diabetes in
carbohydrate counting and insulin dose adjustment. Historically
patients have been prescribed fixed dose insulin to avoid errors and
potential self-harm. Allowing them to calculate their own fast
acting insulin doses was initially met with resistance from care
teams at this hospital; however, changes in working relationships
has produced a collaboration of care providing improved outcomes
for health. Empowering patients with learning difficulties and low
numeracy skills to calculate and adjust their fast acting insulin has
been shown, in two cases, to improve glycaemic control. The
principle of calculating a correction dose, estimating the
carbohydrate in any food or drink and administering a dose of
fast acting insulin was taken from training done by educators on
the BERTIE Programme. The programme was adapted in
timescale (to allow for extra education) and delivery. Individual
sessions allowed for the complex needs of the patients and a unique
calculation tool or slide rule was developed to aid the process.
Working with healthcare professionals across all specialities these
patients now show improved biochemistry and health outcomes.
Reference: 1. http://www.rethink.org/living-with-mental-illness/
police-courts-prison/forensic-mental-health-services/referral.
2014 The Authors.

DIABETICMedicine
P173
Chemotherapy for diabetes? Improved
glycaemic control following
commencement of sunitinib
A Puttanna and A Wright
Diabetes and Endocrinology, Walsall Healthcare NHS Trust, Walsall, UK
A 64-year-old lady with known Type 2 diabetes on NovoRapid

14units with meals and Levemir 40units at night was seen in the
diabetes clinic. She had chronic kidney disease stage 4 secondary to
diabetic nephropathy, hypertension and a right nephrectomy for
renal cell cancer. She was commenced on sunitinib chemotherapy 4
years previously and it had recently been stopped. She mentioned
that since stopping the sunitinib she had noticed erratic blood
sugar control. This had been noticed during the treatment course
when she had been off suntinib (as per the regime). She initially
noticed increased hypoglycaemic episodes and decided to stop her
insulin whilst on sunitinib and restart when treatment finished with
no ill effects and no hypoglycaemic or hyperglycaemic episodes.
Fasting blood glucose whilst on sunitinib was 7.4mmol/l; however,
following discontinuation, repeat fasting blood glucose 3 months
later was 13mmol/l. On a previous occasion, fasting blood glucose
was 10.6mmol/l off sunitinib and improved to 7.2mmol/l on the
chemotherapy. Renal function, full blood count, diabetic
medication and diet remained unchanged throughout this time
and there was no other cause for this deterioration. Sunitinib is a
tyrosine kinase inhibitor used in renal tumours by inhibiting signal
transduction. There are scattered reports of its use being associated
with improved glycaemic control in patients with diabetes;
however, there are no largescale studies. The mechanism is
unclear. This report identifies the possible therapeutic benefit of
sunitinib in patients with diabetes and notes more significant
glycaemic improvement compared with previous reports.
P174
Thalamic abscess as the presenting feature
of Type 2 diabetes
I Karageorgiou1, C Chandler2 and MB Whyte1
1
Diabetes, Kings College Hospital, London, UK, 2Neurosurgery, Kings

College Hospital, London, UK
Introduction: Brain abscess is an unusual complication that may

occur in uncontrolled diabetes. A solitary thalamic abscess is an
uncommon type of brain abscess. We report a case of Type 2
diabetes that presented with a solitary thalamic abscess and discuss
the diagnostic protocol and management.
Case presentation: A 56-year-old, Caucasian, British male with
no medical or travel history presented with 5day symptoms
of meningeal irritation. Body mass index 30.6kg/m2. CT
demonstrated a solitary midline lesion with neoplasia as a
differential diagnosis. It was biopsied and cultures grew
Streptococcus milleri. He was treated by stereotactic puncture
and then external drainage and targeted intrathecal and systemic
antibiotic therapy. He was HIV negative but HbA1c was
10.7%. Dental examination revealed a small molar abscess. The
time to radiological resolution of the thalamic abscess was 2
months. Diabetes improved with 2months of (tapered-dose)
insulin and maintained on metformin with HbA1c 6.9%. He was
left with no neurological disability.
Discussion: Brain abscess classically presents with fever,
headache and neurological deficit. It is treated by antimicrobial
therapy and surgical drainage. Haematogenous spread is usually to
the middle cerebral artery territory, although infrequently to
deeper tissues, such as the thalamus. Streptococci are the cause of
79
DIABETICMedicine
most brain abscesses and S. milleri (present in the oral cavity

and upper respiratory tract) is the most common
pathogen. Predisposing factors for S. milleri infections include
mucosal infection (periodontal diseases and paranasal sinusitis),
pneumonia, alcohol abuse and uncontrolled diabetes. This case
highlights the unusual way diabetes may present and the
importance of periodontal care in diabetes.
haemochromatosis. He was started on 6 weekly testosterone

injections with good effect. Research shows that diabetes is
very often misclassified. If this patient was screened for
haemochromatosis earlier this would have had a significant
impact on his disease related complications and life expectancy.
P175
Third nerve palsy as a presenting feature of
Type 2 diabetes
P177
Lucky or unlucky: a case report of
development of latent onset Type 1
diabetes following gastric bypass surgery
for morbid obesity and Type 2 diabetes
L Veryard and RB Poole
CH Hariman, L Pelluri, S Sankar and S Kumar
Diabetes Centre, Poole Hospital NHS Foundation Trust, Poole, UK
Warwickshire Institute for Diabetes, Endocrinology and Metabolism

(WISDEM) Centre, University Hospital Coventry and Warwickshire, Coventry,
UK
A 64-year-old man presented with a 4 day history of a cold

sensation over his left eye and then a sudden onset of drooping of
his left eyelid and blurry double vision. On the day of onset of his
eye symptoms he had a mild headache on the left side but this had
resolved by the time of admission. He was previously fit and well
and had not seen a doctor in 53 years. He denied any polyuria or
polydipsia. On examination his left eye was unable to adduct or
look superiorly and there was a complete ptosis. His blood
pressure was 187/112. There was evidence of a peripheral sensory
neuropathy. His plasma glucose was 24.7mmol/l and his HbA1c
was 109mmol/mol. CRP was 1mg/l. An MRI of his brain was
normal. A diagnosis of third nerve palsy secondary to undiagnosed
Type 2 diabetes and hypertension was made. Causes of third nerve
palsy include trauma, brain stem vascular events or tumours,
demyelination, posterior communicating artery aneurisms, arteritis
and diabetes. Although diabetes is the most common cause of a
third nerve palsy in adults, this usually occurs in patients with a
long duration of diabetes. Presentation of previously undiagnosed
diabetes with a third nerve palsy is very rare. In this man, the
combination of a peripheral neuropathy and the third nerve palsy
suggested that his diabetes may have been present for several years.
His undiagnosed hypertension may also have contributed to the
development of his cranial nerve palsy.
P176
A case of haemochromatosis and diabetes: a
missed opportunity
1
FR Babwah , M Butt and N Waheed
1
Department of Diabetes, The County Hospital, Hereford, UK, 2Diabetes and
Endocrinology Department, Peterborough City Hospital, Peterborough, UK
A 52-year-old lady was referred to the obesity clinic. She was

diagnosed to have Type 2 diabetes 6 years prior to clinic visit. She
initially weighed 114kg with a calculated body mass index of 48kg/
m2,on metformin 500mg twice daily and 60 units daily of mixed
insulin. She was started on a GLP-1 analogue alongside a
multidisciplinary approach with regular dietitian and psychology
input. After 18 months of treatment with GLP-1 analogue and
insulin, she had managed to lose 6kg in weight, but unfortunately
her weight loss stabilised and she was subsequently considered for
laparoscopic gastric bypass surgery. Postoperatively, her weight
was 102kg and this had reduced to 90.2kg over a period of 2
months. Subsequently in the outpatient clinic review, her insulin
treatment was discontinued and she was making good progress
with a reduction of HbA1c from 84mmol/mol to 67mmol/mol, and
the patient was pleased with the management of her obesity. Six
months following surgery, she was admitted acutely unwell with
symptoms of vomiting and abdominal pain. She had biochemical
evidence of diabetes ketoacidosis with a pH of 6.9 and elevated
serum ketones at 6.1mmol/l. A serum c-peptide was found to
be<94pmol/l and GAD antibodies were 9.1kIU/l (NR < 10.1). She
subsequently developed latent onset Type 1 diabetes following
a gastric bypass surgery for management of morbid obesity and
Type 2 diabetes.
P178
A case of non-diabetic ketoacidosis in
pregnancy
L Turner and A Gallagher
Department of Diabetes and Endocrinology, University Hospitals of Leicester
NHS Trust, Leicester, UK
The association between haemochromatosis and diabetes was first

recognised in the late 1800s, when doctors coined the term bronze
diabetes.The aetiology is thought to be multifactorial. A
60-year-old patient presented to his general practitioner in 2003
with a brief history of osmotic symptoms, weight loss and a blood
glucose of 21.2mmol/l. Although he was not ketoacidotic, he was
diagnosed with Type 1 diabetes. He started a BD mix insulin
regime and remained well controlled. At that time he was noted to
have elevated transaminases:AST 49IU/l and ALT 69IU/l. In 2009
a comprehensive liver screen confirmed a hugely elevated ferritin
level. Haemochromatosis was confirmed on genetic testing and he
commenced treatment with weekly venesections. In February 2013
he had a hospital admission for labrinthitis and was found to have
recurrent hypoglycaemic episodes. On review of historic notes it
was discovered that his deranged liver function tests in fact
predated his diabetes. Glutamic acid decarboxylase antibodies
(GADA) were negative. On further questioning he was also
experiencing low energy and loss of libido. He was found to
have hypogonadotrophic-hypogonadism probably secondary to his
A pregnant 19-year-old Caucasian woman presented at 36 weeks

gestation with a 3 day history of vomiting and dyspnoea. She was
usually well and had one previous pregnancy. She had normal
glucose tolerance at 28 weeks gestation which had been assessed
due to her body mass index of 33kg/m2. On admission she was
clinically dehydrated and unable to tolerate oral fluids. Admission
ABG showed pH 7.296, pCO2 1.79, pO2 15.5, bicarbonate
10.9mmol/l, glucose 4.2mmol/l. Her urine revealed 4+ ketones.
Lactate, paracetamol and salicylate levels were all normal as was
her renal function. A diagnosis of non-diabetic ketoacidosis was
made and she was treated with intravenous 10% dextrose,
Pabrinex and antiemetics. This resulted in correction of her
acidosis within 8h, symptomatic improvement and subsequent
clearing of urinary ketones. Regular fetal monitoring occurred
throughout with no concerns. She was subsequently discharged
home and was next admitted in spontaneous labour at 39 weeks
gestation and delivered a healthy baby. Non-diabetic ketoacidosis
80
2014 The Authors.

occurs due to an increased tendency for ketogenesis and high levels

of insulin resistance in advanced pregnancy. It is usually
precipitated by vomiting in the third trimester but is not
frequently seen and therefore diagnosis can be missed or delayed.
In reported cases this had often led to premature delivery of the
fetus. In our patient the diagnosis was recognised on the day of
admission allowing for prompt provision of carbohydrate in the
form of intravenous dextrose which resulted in resolution of
ketosis and continuation of pregnancy.
P179
Alcoholic ketoacidosis mimicking diabetes
ketoacidosis
WMHS Chandrasekara1, FJ Burton2 and C Jayawarna1
1
Acute Medicine, Stepping Hill Hospital, Stockport, UK, 2Endocrinology and
Diabetes, Stepping Hill Hospital, Stockport, UK
Introduction: Chronic alcoholism is a frequently unrecognised

cause of ketoacidosis. Most patients with alcoholic ketoacidosis
present with normal or low glucose, but can present with
hyperglycaemia. This can lead to misdiagnosis of diabetes
ketoacidosis and inappropriate treatment with insulin.
Case: We describe a 37-year-old Caucasian lady with chronic
pancreatitis secondary to excess alcohol consumption and
depression, admitted with abdominal pain and vomiting. She had
abstained from alcohol for a few months but consumed 90 units of
alcohol 2 days prior to admission. She was on Creon (pancrelipase)
for pancreatic exocrine dysfunction. Her system examination was
unremarkable except for generalised abdominal tenderness. Her
body mass index was 17kg/m2. Arterial blood gas showed pH
7.16, standard bicarbonate 3.0mmol/l. Her random blood glucose
was 15.1mmol/l and blood ketones were 7.0mmol/l. Other blood
results showed WBC 7.6, sodium 126mmol/l, potassium 3.0mmol/
l, eGFR > 60ml/min and amylase 24U/l. Her chest X-ray was
normal. She was treated with intravenous fluids and intravenous
insulin according to diabetes ketoacidosis protocol. She had a
hypoglycaemic attack within an hour and insulin therapy. On
review of her old notes she was found to have three similar
episodes over the last 12 months. Her HbA1c ranged from 22 to
39mmol/mol over the same period.
Conclusions: Alcoholic
ketoacidosis
can
present
with
hyperglycaemia due to relative deficiency of insulin and relative
surplus in counter-regulatory stress hormones including
glucagon. Awareness of the syndrome with detailed history helps
to differentiate alcohol ketoacidosis from diabetes ketoacidosis and
prevent iatrogenic hypoglycaemia.
P180
Observational outcome data using
glulisine during pre-conception and
throughout pregnancy in women with
Type 1 diabetes
H Chenoweth and H Probert
Diabetes and Endocrine Centre, Royal Cornwall Hospital, Truro, UK
DIABETICMedicine
included glulisine and glargine. The mean HbA1c during

pre-conception was 58mmol/mol (7.5%), range 54mmol/mol
63mmol/mol (7.1% 7.9%); the National Institute for Clinical
Excellence (NICE) recommendation is to aim for a target of
43mmol/mol (6.1%) during pre-conception. The mean HbA1c
during
mid-term
improved
to
52mmol/mol
(6.9%),
range 50mmol/mol 54mmol/mol (6.7% 7.1%). The diabetes
control remained stable with a pre-delivery mean HbA1c result
of 49mmol/mol (6.6%), range 40mmol/mol56mmol/mol (5.8%
7.3%). The outcomes of the pregnancy resulted in two vaginal
deliveries and two Caesarean sections with no congenital
abnormalities identified. The mean birth weight of the babies
was 3682g, range 3200g4030g.
Discussion: With intensified pre-conception care and regular
monitoring throughout the pregnancy, glulisine appears to have
been successful in keeping good diabetes control with no evidence
of congenital abnormalities.
P181
A case of coexisting lipohypertrophy and
lipoatrophy
RPK Lim1,2, P Joseph2, K Leong2, KS Leong1,2 and
M Banerjee1,2
1
Diabetes and Endocrinology Department, Wirral University Teaching Hospital
NHS Foundation Trust, Upton, UK, 2Diabetes Department, Clatterbridge
Hospital, Wirral University Teaching Hospital NHS Foundation, Bebington, UK
Background: Lipodystrophies are well-documented side effects of

insulin therapy. Modified-release exenatide (exenatide LAR) is
known to cause subcutaneous nodules which can take weeks to
resolve. We describe a case where both forms of lipodystrophies
coexisted in a patient treated with insulin and exenatide LAR.
Case history: A 62-year-old lady with Type 2 diabetes for 13
years had a body mass index of 38kg/m2. She had a history of
hypertension and transient ischaemic attack but no microvascular
complications. Her medications included metformin MR and
pioglitazone. Insulin therapy was added 6 years ago by
basal-bolus regime, and modified-release exenatide (Bydureon)
was subsequently added to optimise glycaemic control. A few
months later, she was found to have developed two areas of a
combination of lipohypertrophy as well as lipoatrophy in the
abdomen. These were painless and asymptomatic, measuring
15cm 9 8cm and 14cm 9 8cm respectively. She admitted to
injecting all her insulin and exenatide in these two sites on her
abdomen using 8mm needles despite previous education. She has
been re-educated to avoid the sites.
Conclusion: Modified-release exenatide had been reported to
cause small subcutaneous injection site nodules. These resolve
spontaneously within 3 6 weeks. Lipodystrophy, however, is
novel and had not been described before with GLP-1 therapy.
Although lipohypertrophy remains a frequent complication of
insulin therapy and lipoatrophy at insulin injection sites is
considered to be immune complexmediated, the patient had not
developed either of them previously. Furthermore, the
simultaneous development of lipohypertrophy and lipoatrophy is
uncommon.
Case: Retrospective data were collected from four females with

Type 1 diabetes known to have used glulisine during
pre-conception and throughout pregnancy. They were all given
pre-conception guidance by the diabetes specialist team and they
chose to remain on glulisine during this time. Their regimes were
intensified during pre-conception; three women were on
continuous subcutaneous insulin infusions (CSII) using only
glulisine and one woman was on multiple daily injections which
2014 The Authors.

81
DIABETICMedicine
P182
Timing of access to secondary healthcare
services and lower extremity amputations
in people with diabetes: a case control study
CM Buckley1,2, F Ali3, G Roberts3, PM Kearney2, IJ Perry2 and
CP Bradley1
1
Department of General Practice, University College Cork, Cork, Ireland,

Department of Epidemiology and Public Health, University College Cork,
Cork, Ireland, 3Department of Medicine and Metabolism, Waterford Institute
of Technology, Waterford, Ireland
2
Objective: To investigate the association between timing of

patient access to secondary healthcare for diabetes management
and lower extremity amputation (LEA) in patients with diabetes.
Method: This was a case control study set in the secondary
healthcare system inIreland. The cases were 116 patients with
diabetes who underwent a first major non-traumatic LEA between
2006 and 2012. Controls were 348 patients with diabetes who
were admitted to hospital for any other cause during the same time
period, frequency matched by gender, type of diabetes, year and
hospital of LEA. Data on predictor variables were collected for 7
years prior to the event year. The outcome measures were odds
ratios (ORs) for LEA in patients with diabetes by early vs late
referral from primary to secondary healthcare. ORs were derived
using unconditional logistic regression with adjustment for
potential confounding variables.
Results: Statistically significant risk factors for LEA in patients
with diabetes identified from the analysis included being single,
chronic kidney disease, retinopathy and hypertension. Higher
HbA1c levels in cases than controls were observed 7, 5 and 3 years
prior to the event whereas no significant difference was detected 1
year prior to the event. In unconditional logistic regression analysis
adjusted for potential confounders and with frequency-matching
variables as covariates, there was no evidence of a reduced risk of
LEA among patients referred earlier to secondary healthcare for
diabetes management.
Conclusions: Referral may need to occur earlier than the 7 year
cut-off used in this study to demonstrate an effect on reducing the
risk of an LEA. Alternatively, efforts to develop more integrated
care between primary and secondary services may be beneficial,
rather than focusing on timing of referral to secondary healthcare.
P183
Mycotic infection in diabetic foot ulcers: a
cause of non-healing
Z Htike1, R Jogia2, D Modha3, R Berrington2, K Nisal2,
S Jackson2 and MF Kong2
1
Leicester Diabetes Centre, Leicester General Hospital, Leicester, UK,

Diabetes Department, Leicester General Hospital, Leicester, UK,
3
Microbiology Department, Leicester General Hospital, Leicester, UK
2
Mycotic infection can affect up to 20% of diabetic foot ulcers often

as secondary infection. However, it is under-suspected leading to
delayed wound healing. We have recently seen four cases in our
foot clinic which highlights the need to raise our index of suspicion
in chronic non-healing ulcers. A 72-year-old man was referred
with 6 months duration of plantar ulceration with a growth
extending to the dorsum. He developed gangrene requiring
amputation of the toes. There was a rapid regrowth of
granulation tissue at the amputation site. The excised tissue
culture confirmed mycetoma of Fusarium solani isolate. A
51-year-old lady presented with a protracted ulcer on the medial
aspect of the left foot. It was surgically debrided and she received
antibiotics for mixed bacterial infection. Non-healing osetomyelitis
82
prompted amputation of the big toe. Bone sequestrum grew

Candida parapsilosis. A 73-year-old man had a trans-metartarsal
amputation for a non-healing ulcer. The amputation site failed to
heal with sinus formation. Bone sample grew Candida albicans. A
52-year-old man with Charcots arthropathy had a recalcitrant
wound with abscess formation over 5 years. Recent tissue sample
grew Candida parapsilosis. In all cases, ulcers healed completely
after antifungal therapy. Fungal infection is commonly acquired
during the course of treatment for bacterial infection. Therefore, a
high index of suspicion is required in protracted diabetic foot
ulcers. We propose that identification and treatment of fungal
infection may lead to better outcomes for patients with non-healing
ulcers who have otherwise received ideal care.
P184
Resistant hypoglycaemia in post-Roux-en-Y
gastric bypass surgery
R Khan, A Jamalfar, P Jacob and SW Coppack
Department of Endocrinology and Diabetes Mellitus, St Bartholomews
Hospital, Barts Health NHS Trust, London, UK
A 46-year-old White man was referred for frequent severe

hypoglycaemic episodes 3 months post-Roux-en-Y gastric bypass
surgery. The surgery had successfully resolved his diabetes, as well
as achieving some weight loss. However, he soon developed
intolerance to solid food and complained of symptomatic features
of hypoglycaemia that required frequent use of hypostop and
sugary drinks. Recurrent admissions over the next 18 months
recorded fluctuating capillary blood glucose levels, ranging from
2 to 5.4mmol/l, which were uncoupled with sporadic
neuroglycopaenic symptoms. Results of a glucose tolerance test
were suggestive of postprandial hyperinsulinaemic hypoglycaemia,
with a blood sugar of 2.4mmol/l 1h post sugar challenge. However,
measurements of his counter-regulatory hormones including
growth hormone, cortisol and c-peptide were not supportive of
this. Visual studies of the alimentary tract revealed no contributory
anatomical obstruction or defect. Treatment for suspected early
dumping syndrome proved to be futile, with no change in his
symptoms with diet alteration and combination of calcium channel
blocker, octreotide and acarbose. Cessation of his symptoms was
noted following insertion of percutaneous endoscopic gastrostomy
which led to the eventual decision to reverse his gastric bypass.
Unfolding of the above events demonstrate that although the
beneficial impact of bariatric surgery on obesity-associated
comorbidities are much to be admired, problems can occur. This
case highlights a previously recognised complication of severe
hypoglycaemia occurring post-gastric-bypass surgery, with
difficulties in distinguishing the underlying pathophysiological
basis, as well as being resistant to medical management to the
extent of warranting reversal of the gastric bypass surgery.
P185
Progressive cerebellar ataxia and new onset
diabetes
M Kong1,3, G Glibert2, F Baleanu1 and R Karmali1
1
Department of Diabetes and Endocrinology, Centre Hospitalier Universitaire
Brugmann, Brussels, Belgium, 2Department of Neurology, Centre Hospitalier
Universitaire Brugmann, Brussels, Belgium, 3Department of Diabetes,
University Hospitals of Leicester NHS Trust, Leicester, UK
A 37-year-old Rwandan business man was transferred from Kenya

where he had been investigated for subacute onset of progressive
cerebellar ataxia and quadriparesis of unclear aetiology. There was
no history of preceding infection or vaccination and the only recent
2014 The Authors.

travel had been to Dubai and Europe. MRI scan of the brain and
spine was essentially normal. Detailed laboratory profile was
normal apart froma fasting blood glucose of 8.6mmol/l and a
dyslipidaemia. A paraneoplastic syndrome was considered but
investigations were negative. A repeat fasting glucose of 9.0mmol/l
confirmed diabetes and he was started on metformin. He had no
family history of diabetes or of autoimmune diseases.
Autoantibodies to glutamic acid decarboxylase (GAD-ab) were
checked and were very high (>250IU/ml). Islet cell antibodies were
positive at a titre of 25,600. A diagnosis of autoimmune cerebellar
ataxia was suspected and confirmed by the presence of high titres
of GAD-ab in the cerebrospinal fluid. He has responded poorly to
treatment. A year after diagnosis of his diabetes his blood sugars
remain well controlled on metformin only. Thyroid, adrenal,
parietal cell and gliadin antibodies were negative. Autoimmune
cerebellar ataxia related to GAD-ab is a rare condition that
typically affects women with late-onset Type 1 diabetes or other
autoimmune disorders. Cerebellar ataxia, the second most
common syndrome associated with high GAD-ab levels, shares
with stiff-man syndrome the same demographic, clinical and
immunological features. Our patient most probably has latent
autoimmune diabetes of the adult.
P186
The chronic and acute management of
glycogen storage disease Type 1 causes
difficulties both in diagnosing diabetes and
managing glycaemic variability: a case
report
J Hazlehurst1,2, J Johnson3, T Geberhiwot1, P Newsome4,
J Webber1 and SG Ghosh1
1
Diabetes Centre, University Hospitals Birmingham, Birmingham, UK, 2Centre
for Endocrinology, Diabetes and Metabolism, University of Birmingham,
Birmingham, UK, 3Dietetics, Metabolic Liver Service, University Hospital
Birmingham, Birmingham, UK, 4Liver Biomedical Research Unit, University of
Birmingham, Birmingham, UK
A 32-year-old male of Pakistani origin was diagnosed in childhood

with glycogen storage disease Type 1a (GSD1a) on the basis of
failure to thrive, hypoglycaemia, lactic acidemia and genetic
analysis (G6PC mutation). His brother has the same mutation.
He also has Crohns disease, epilepsy and chronic pancreatitis. On
examination his body mass index is 21.4 and he has severe
hepatomegaly and growth retardation. He regularly presents with
infective symptoms or abdominal pain, hypoglycaemia and very
high lactate. His acute management involves treatment of sepsis
and initiation of either an emergency feeding regimen or parenteral
dextrose/saline solution. There is often a delay between correcting
his hypoglycaemia and the lactic acidosis resolving so he becomes
hyperglycaemic and is managed with a variable rate insulin
infusion. HbA1c is consistently 30 40mmol/mol with normal
fructosamine levels. Islet cell and anti-glutamic acid decarboxylase
antibodies are negative. Continuous glucose monitoring has been
limited by variable compliance. He can become hypoglycaemic
2 4h after a carbohydrate load, and has periods of unpredictable
postprandial transient hyperglycaemia (25mmol/l) with osmotic
symptoms. His GSD1 is now managed with oral cornstarch (as
glycosade). The hyperglycaemia is thought to be due to pancreatic
insufficiency although c-peptide levels are appropriate for glucose
concentrations and remains not fully explained. Most of the oral
hypoglycaemics are contraindicated and insulin administration
may put him at risk of development of prolonged intractable
hypoglycaemia.
2014 The Authors.

DIABETICMedicine
P187
Diabetes ketoacidosis precipitated by
natalizumab infusion
N Choudhary, H Chenoweth and DL Browne
Diabetes and Endocrinology, Royal Cornwall Hospital Trust, Treliske, Truro,
UK
Case: We present the case of a 42-year-old lady with Type 1

diabetes for 18 years who developed two episodes of
hyperglycaemia and diabetic ketoacidosis following natalizumab
(Tysabri) infusion for relapsing remitting multiple sclerosis. She
was commenced on continuous subcutaneous insulin infusion
(CSII) therapy for disabling hypoglycaemia in 2011. Her glycaemic
control was impeccable with an HbA1c of 53mmol/mol (7%).
After several relapses of multiple sclerosis, she was commenced on
natalizumab infusion in July 2012 by the neurologist. Following
the first infusion, she developed uncontrolled hyperglycaemia with
ketosis requiring hospital admission for 24h despite appropriate
self-management strategies of increased insulin infusion via CSII.
Following the second infusion in August 2012, she again developed
diabetes ketoacidosis requiring hospitalisation while on holiday.
Both episodes happened 17 days after natalizumab infusion and no
other precipitating cause was identified. Subsequent natalizumab
infusions were discontinued. She has suffered no further events of
ketoacidosis in the following 12 months. A drug alert was
submitted.
Discussion: Natalizumab,
a
selective
adhesion-molecule
inhibitor, is used in relapsing remitting multiple sclerosis and
Crohns disease. There are sporadic reports of development of
diabetes and poor glycaemic control with natalizumab infusion,
although this does not feature in the side effect profile of the drug.
The mechanism of hyperglycaemia mediation and the 17 day
interval following administration is unclear. This case highlights
the importance of recognising and reporting new and rare side
effects via drug alert reports. Patients with diabetes should be
advised to increase capillary glucose monitoring frequency to
identify and treat hyperglycaemia.
P188
Sudecks atrophy: an uncommon
complication of diabetes foot infection
SU Srinivas1,3, C Elliot1, N Majid1, R Narayanan1, S Keigan1,
P Chiran1 and N Mohsin2
1
Department of Diabetes and Endocrinology, St Helens and Knowsley
Teaching Hospitals NHS Trust, St Helens, UK, 2Department of Radiology,
St Helens and Knowsley Teaching Hospitals NHS Trust, St Helens, UK,
3
University of Manchester, Manchester, UK
Case history: A 49-year-old woman with Type 2 diabetes for 17

years, diabetic retinopathy and sensory neuropathy presented with
a 2week history of left forefoot, plantar ulcer and cellulitis.
Peripheral pulses were palpable; clinically and radiologically there
was no osteomyelitis. Wound swab revealed methicillin resistant
Staphyloccus aureus (MRSA), Group B Streptococcus and
anaerobes. She was initially treated with intravenous (IV)
vancomycin and metronidazole for 10 days and then with IV
teicoplanin for 4 weeks. The antibiotics were then changed to oral
doxycycline and fuscidic acid for 3 weeks. The ulcer healed in 10
weeks, but the patient developed left forefoot oedema, erythema
and pain. A repeat X-ray of the left foot revealed patchy bone
demineralisation, significantly different from the initial X-ray. MRI
of the left foot showed subcutaneous soft tissue thickening with
marrow oedema of the first metatarsal, most pronounced in the
subarticular region. There was enhancement of these areas with IV
83
DIABETICMedicine
contrast. A transverse fracture line extending through the navicular

was noted. The clinical features of pain, swelling, erythema (after
the ulcer healed) and radiological findings were consistent with
complex regional pain syndrome or Sudecks atrophy. The patient
was provided analgesia, air-cast walker and referred for passive
ankle physiotherapy. The oedema, erythema and tenderness settled
in about 6 weeks.
Conclusion: This case highlights the importance of considering
Sudecks atrophy in patients with treated diabetes foot infections in
the presence of unexplained pain, erythema, oedema and the
typical radiological change of patchy demineralisation.
P189
Hemichorea-hemiballismus in an adolescent
with Type 1 diabetes
R Srinivasan1, CI Woodward2, K Cudmore3, L Krishnan4,
V Raman5, M Venu1, J Saada2 and J Randall1
1
Diabetes and Endocrinology, James Paget University Hospitals NHS
Foundation Trust, Great Yarmouth, UK, 2Radiology, Norfolk and Norwich
University Hospitals, Norwich, UK, 3Accident and Emergency, James Paget
University Hospitals NHS Foundation Trust, Great Yarmouth, UK, 4Diabetes
and Endocrinology, Queen Elizabeth Hospital, Kings Lynn,UK, 5Paediatrics,
James Paget Hospitals NHS Foundation Trust, Great Yarmouth, UK
A 17-year-old with Type 1 diabetes of 4 years duration was

admitted with abnormal involuntary movements of left arm and leg
for 3weeks. On examination he had hemichoreic-hemiballismic
movements. Blood glucose was 44mmol/l with no ketonaemia and
serum osmolality was 316mmol/kg. MRI of the brain showed
bilateral, predominantly right sided, T1 signal hyperintensity in the
caudate, lentiform and thalamus with prominent low signal on the
gradient echo T2 sequence consistent with haemosiderin deposition
and enhancement following gadolinium administration. He was
diagnosed to have hemichorea-hemiballismus due to non-ketotic
(hyperosmolar) hyperglycaemia. He was started on clonazepam
and his insulin regime was intensified with improvement of his
symptoms. Hemichorea-hemiballismus is a movement disorder
characterised by involuntary, irregular, flinging movements of
variable amplitude involving commonly upper and lower limbs.
Hemichorea-hemiballismus is associated with lesions affecting the
contralateral subthalmic nucleus. Hemichorea-hemiballismus due
to hyperglycaemia has been reported, typically in older adults with
Type 2 diabetes. We believe our patient is the youngest affected to
date. The MRI appearances in this case are typical of those
described in the literature along with the clinical features and
biochemical abnormalities. The pathogenesis of this disorder due to
hyperglycaemia is not clear although gliosis due to micro infarction
has been noted in an autopsy report. The involvement of
microhaemorrhages has been debated. However, the blooming
artefact within the globus pallidus in this case would support a role
for microhaemorrhage or iron deposition in the pathological
process.
P190
Thick skinned
K Duncan, M Pass, P Ogilvie and R Nayar
Departments of Diabetes and Endocrinology, City Hospitals Sunderland NHS
Foundation Trust, Sunderland, UK
Inspection of his injection sites across the abdomen revealed

lipohypertrophy despite being on analogue insulin. A detailed
history revealed difficulty in blood glucose testing and insulin
administration due to stiffness and tightness in hands and fingers
over the preceding 12 months; episodes where his fingers turned blue
with pain especially when taking a bath occurred. Examination
revealed periungual erythema, blisters on the tips of his fingers with
pitting of the finger pulp. Evidence of sclerodactyly in a number of
digits with scleroderma on the dorsal aspects of both hands, across
his forearms, trunk and some involvement below the knees was
present. Telangectasia across the right side of his face was noted. A
diagnosis of systemic sclerosis was confirmed with positive ANA,
ENA antibodies (speckled pattern) and Raynauds phenomenon
with thermography and capillaroscopy. We further established
symptoms of dyspepsia but no dysphagia. He was commenced on
ACE and proton pump inhibitors, and calcium channel antagonist.
Insulin sites were rotated leading to improvement in glycaemic
control.
Learning points: Clinicians need to remain vigilant to the
associated dermatological problems in patients with Type 1
diabetes. Systemic sclerosis can have a significant impact on the
practicalities of blood glucose monitoring and insulin
administration.
P191
One foot, two diseases: a case of Charcots
foot with superimposed osteomyelitis
D Blundell, U Aziz and V Oguntolu
Wolverhampton Diabetes Centre, Royal Wolverhampton NHS Trust,
Wolverhampton, UK
A 25-year-old female with Type 1 diabetes complicated with

retinopathy and peripheral neuropathy presented to the foot clinic
with swollen and hot left foot. A diagnosis of midfoot Charcots
arthropathy was made. Standard treatment with offloading was
commenced. The foot was offloaded with a removable cast walker,
as total contact casting was not available in our foot clinic. After 6
months of offloading, the cast walker was removed as the Charcot
was quiescent. Two weeks after resuming weight bearing she
presented again with grossly swollen hot, left foot. A diagnosis of
Charcots reactivation was made and offloading with a cast walker
was commenced again. The swelling persisted despite offloading.
X-ray, magnetic resonance imaging and Doppler did not reveal any
significant pathology. Six weeks later, she re-attended with
persistent foot swelling, systemic symptoms and cellulitis of the
left foot with no ulcers present. C-reactive protein was elevated
and a clinical diagnosis of osteomyelitis was made. Treatment
was commenced with broad spectrum antibiotics and abscess
debridement and drainage was performed multiple times. Vacuum
therapy was used for post surgical wound treatment. The foot was
offloaded with a removable cast walker and antibiotics were
continued for 6 weeks. Concomitant Charcots foot and
osteomyelitis is very rare, especially in patients without foot
ulcers. This case demonstrates the rare possibility of coexisting
Charcots and osteomyelitis. It is imperative that clinicians should
have high index of suspicion in patients with Charcots foot, if
swelling and temperature does not resolve within a few weeks of
offloading.
Introduction: An unusual cause of deteriorating glycaemic

control in a patient with Type 1 diabetes. A 48-year-old man
with Type 1 diabetes for 14 years presented to clinic with
suboptimal glycaemic control; his HbA1c was rising despite
adherence to a healthy eating plan and basal-bolus regimen.
84
2014 The Authors.

P192
Transforming diabetes of pregnancy
K Duncan, M Pass and R Nayar
Department of Diabetes and Endocrinology, City Hospitals Sunderland NHS
Introduction: Women who develop gestational diabetes have an

increased risk of Type 2 diabetes. We report the development of
Type 1 diabetes post partum.
Case history: Our patient, a 33-year-old woman, was referred to
the combined antenatal clinic with a new diagnosis of gestational
diabetes (GD) during her first pregnancy (fasting glucose 6.3mmol/
l, 2h glucose 9.8mmol/l); body mass index 24kg/m2. Despite
adherence to a healthy eating plan and lifestyle blood glucose
targets were not met. Metformin was commenced, but our patient
was intolerant of this agent and was converted to twice daily
human biphasic insulin. She maintained excellent control
throughout the pregnancy. Delivery was uncomplicated at 38
weeks. The lack of risk factors for GDM raised the possibility of
Type 1 diabetes; anti GAD antibodies during the antenatal period
were normal and subsequent postnatal checks confirmed normal
glucose tolerance. Our patient returned 6 months later to the
pre-conceptual service with a short history of osmotic symptoms,
3kg weight loss but no ketonuria, a random glucose was
13.5mmol/l and anti GAD antibodies were elevated at 79.7U/ml.
A diagnosis of Type 1 diabetes was established and a basal-bolus
regimen of insulin was commenced.
Learning points: The development of Type 1 diabetes in women
with previous GDM is rare but recognised. Autoantibody
assessment should be considered in primigravidas women with
gestational diabetes who have a normal body mass index, have no
risk factors GDM, or are intolerant of metformin, and require
insulin in pregnancy.
P193
Severe oedema: a rare complication of
insulin therapy
RS Manjunatha, UY Raja, R Mahto and PM Horrocks
Department of Diabetes and Endocrinology, Warwick Hospital, Warwick, UK
DIABETICMedicine
P194
Charcot neuroarthropathy in a patient with
Type 1 diabetes and Charcot Marie tooth
disease
N Ibraheem and K Jacob
Diabetes and Endocrine, Pilgrim Hospital, Boston, UK
A 46-year-old female, with Type 1 diabetes with suboptimal

diabetes control presented with difficulty walking.At age 26,
pulmonary tuberculosis was diagnosed and treated with
antitubercular therapy.Foot examination revealed bilateral pes
cavus and hammer toes. There was reduced pin prick sensation to
the mid foot with normal proprioception. There was no distal
muscle wasting and power was normal. There was global
areflexia. Nerve conduction studies confirmed demyelination.
Following neurology review, with strong family history, a
diagnosis of Charcot Marie tooth disease was made. Genetic
testing was negative. Blood tests revealed low vitamin B12 which
was replenished with some improvement in gait. Referral
to orthopaedicians to correct foot deformity resulted in left
plantar fascia release. Postoperatively, the left foot became hot
and swollen with a temperature difference of 7 C. An MR
scan suggested Charcot neuroarthropathy.
Discussion: Diabetic peripheral neuropathy is a diagnosis of
exclusion. Our case demonstrates the importance of considering
other causes for peripheral neuropathy in a patient with diabetes.
The neuropathy in our case could have been multifactorial
with hereditary motor sensory neuropathy (HSMN), diabetic
neuropathy, antitubercular drugs or vitamin B12 deficiency.
Charcot neuroarthropathy develops in a neuropathic foot and is
thought to be initiated by trauma. In our patient, the surgical
intervention was done to correct the foot deformity secondary to
Charcot Marie tooth disease. This could possibly have
precipitated the Charcot neuroarthropathy. We couldnot find any
published literature of patients with coexisting diabetes and
HSMN presenting with acute Charcot neuroarthropathy.
Caution is advised and a high index of early suspicion is
necessary to make the diagnosis of Charcot neuroarthropathy.
P195
Case: A 46-year-old lady with Type 1 diabetes since age 19 was

referred to the diabetes clinic in view of a 3 4 month history of
generalised oedema. She had been started on insulin pump at another
hospital a year ago. Since then, her diabetes control had improved
dramatically. There were no clinical signs of heart failure. She had
pitting oedema of both her legs, ascending up to the groin and
abdominal wall. U&Es, LFTs and CXR were normal. Urine dipstick
was negative for protein; 24h urine collection for protein was not
raised. Serum albumin was normal at 38g/l. Echocardiogram was
normal. A CT scan of her abdomen and pelvis revealed extensive
subcutaneous oedema, small bilateral pleural effusions, and a small
pericardial effusion with multiple enlarged pelvic varices surrounding
the uterus. No thrombus was seen in the iliac veins or inferior
venacava. There was no mass lesion seen compressing the IVC.
Dopplar ultrasound of her legs was negative for deep vein
thrombosis. Given the negative investigation, the working
diagnosis was that of oedema caused by insulin therapy. She has
been started on bumetanide with good response. This case serves as a
reminder that, rarely, crucial fluid retention can occur with insulin
therapy. The oedematous state usually develops briefly after the
initiation of intensive insulin therapy. The severity of peripheral
oedema is variable, most cases being mild. In the absence of heart,
liver or renal involvement, insulin induced oedema should be thought
of as a cause of peripheral or generalised odema.
A 69-year-old Asian man with Type 2 diabetes for 24 years presented

with a 2 week history of bilateral leg pain and fevers. He described a
sharp pain worse with movement and palpation; there was no
history of trauma. On examination, his temperature was 38C and
pulse 120. He had evidence of a proximal myopathy: wasted
quadriceps muscles and unable to stand from sitting with power 4/5
on hip flexion and extension. He had brisk knee reflexes, absent
ankle jerks and a waddling broad-based gait. Five months prior to
presentation his HbA1c was 6.9% (52mmol/mol). Investigations
showed CRP 272, HbA1c 11.3% (97mmol/mol). CK 41units/l,
LDH 177units/l and an autoimmune screen (ANA negative <1:160)
were all within normal limits. Staphylococcus aureus was isolated in
his urine. MRI of the lower limbs demonstrated signal abnormality
throughout the anterior compartments bilaterally; there was no
evidence of osteomyelitis. These changes were consistent with a
diabetic radiculoplexopathy and electromyography (EMG) studies
confirmed the diagnosis. The pain continued despite antibiotic
2014 The Authors.

85
Severe diabetic amyotrophy: acute changes

on magnetic resonance imaging (MRI)
resolve following insulin treatment
AO Adjene, S Robinson and M Yee
Department of Metabolic Medicine, St Marys Hospital, London, UK
DIABETICMedicine
treatment for his sepsis and neuropathic agents; it improved

following insulin treatment. Six months later his symptoms had
completely resolved with resolution of the initial signal changes seen
within the musculature on MRI. Further EMG studies showed no
features of diabetic amyotrophy. Diabetic amyotrophy is a rare
complication and the pathogenesis is unclear. MRI of the affected
limbs can play a complementary role in the diagnosis. Randomised
trials do not support the use of immunotherapy. Good glycaemic
control remains the mainstay of treatment.
diet and exercise. Having identified the success of CGM for the
diagnosis and management of CFRD, we have incorporated this
into our annual screening programme for all patients.
P197
Dapagliflozin in primary care: three case
studies
N Milne and N Kanumilli
Northenden Group Practice, Manchester, UK
P196
The sweet success of continuous glucose
monitoring for diagnosis and management
of cystic fibrosis related diabetes
P Dyce
Cystic Fibrosis, Liverpool Heart and Chest Hosptial, Liverpool, UK
A 28-year-old female patient with cystic fibrosis (CF) was admitted

to hospital due to pulmonary exacerbation. Previous history
showed a decline in pulmonary function measured by forced
expiratory volume (FEV1) from 76% to 63% and an increased
number of hospital admissions (three in 9 months). Previous
investigations included HbA1c and 2 weeks of serial glucose
monitoring. HbA1c levels were 41 (IFCC) and all postprandial
serial glucose monitoring recordings were <7mmols/l. On
discharge, a continuous glucose monitor (CGM) was attached for
5 days and the patient was asked to complete a food and exercise
diary. After 5 days, the diary was analysed and the monitor was
uploaded onto a computer software program. CGM showed
elevated glucose levels and the patient was subsequently diagnosed
with cystic fibrosis related diabetes (CFRD). Early detection and
treatment of this growing complication of CF is paramount to
avoid clinical decline and long-term complications. Following
diagnosis and insulin initiation, the patient had no further
admissions and Fev1 has risen to 82%. Without CGM, the
patient would never have been diagnosed using conventional
methods and oral glucose tolerance testing is not an effective tool
for the detection of CFRD. Furthermore, the graphical information
from the uploaded monitor was found to have educational benefits,
enabling a better understanding of the association of glucose levels,
Background: Dapagliflozin is a new, selective sodium glucose

co-transporter 2 inhibitor (SGLT-2) which has shown to reduce
HbA1c in a wide range of patients with Type 2 diabetes through
increasing renal glucose excretion. Weight loss and low rates of
hypoglycaemia were observed in clinical trials.
Objective: To evaluate the effect of dapagliflozin therapy in a
primary care setting through the description of three case studies.
Cases: Case 1 is a female, 56 years, metformin 500mg bd,
gliclazide 80mg bd and insulin levemir 100IU/d. HbA1c 80mmol/
mol (9.5%), weight 101kg, declines to change insulin regime. Case
2 ismale, 57 years, metformin 850mg tds, gliclazide MR 120mg/
day, HbA1c 70mmol/mol (8.6%), weight 99kg, refusing injectable
therapy. Case 3 is afemale, 67 years, metformin MR 1000mg bd,
gliclazide MR 60mg/day, insulin glargine 20IU/d, no baseline
HbA1c, weight 73kg and static.
Results after 3 months: Case 1: HbA1c reduced by 19mmol/
mol ( 1.3%) to 66mmol/mol (8.2%) ( 22%), weight unchanged.
Case 2: HbA1c reduced by 14mmol/mol ( 1.3%) to 56 mmol/
mol (7.3%) ( 20%), weight reduced by 2kg to 97kg ( 2%). Case
3: weight reduced by 4.76kg ( 6.5%), insulin treatment
discontinued. All patients were very happy with the progress.
No hypoglycaemia or adverse effects were reported in these three
patients.
Conclusion: Dapagliflozin showed significant improvements in
HbA1c and weight loss in selected patients. In one case, a patient
experienced a significant weight loss and was able to discontinue
insulin glargine treatment. Dapagliflozin as add-on therapy to
concomitant medication was well tolerated in patients with Type 2
diabetes, improving glycaemic control and weight.
Clinical care and other categories posters: cerebral and

cerebrovascular complications
P198
P199
Increased mortality and poorer functional

outcomes in diabetes patients admitted
with stroke
Presence of retinopathy and a high

microaneurysm count increase the risk of
subsequent cardiovascular disease
S Kassim1, L Hunt1, S Muniyappa2, N Ranjant1, O Hudson1,

K Hudson1, C Nelson1, G Pratt3, FM Creagh1 and
D Selvarajah1,4
A Fleming1, S Philip2, G Prescott1, J Olson2 and P Sharp1
1
Diabetes and Endocrinology, Sheffield NHS Teaching Hospitals Foundation
Trust, Sheffield, UK, 2Diabetes and Endocrinology, Dewsbury NHS Trust,
Dewsbury, UK, 3Stroke Medicine, Sheffield NHS Teaching Hospitals
Foundation Trust, Sheffield, UK, 4Department of Human Metabolism,
University of Sheffield, Sheffield, UK
Refer to Oral A66
86
1
College of Life Sciences and Medicine, University of Aberdeen, Aberdeen,
UK, 2Grampian Retinal Screening Service, NHS Grampian, Aberdeen, UK
Aim: Can manual or automated assessment of retinopathy predict

future risk of cardiovascular disease in people with diabetes?
Methods: This was a retrospective cohort study using data from
the regional diabetes register linked to hospital morbidity and
mortality data. Retinal photographs and routine clinical data
related to cardiovascular risk and cardiovascular outcome data
were obtained for 13,124 people with diabetes who attended
2014 The Authors.

Clinical care and other categories posters: cerebral and cerebrovascular complications
retinal screening between 2003 and 2004. The retinal photographs

were also assessed using an automated system for detection of
diabetic retinopathy that has been validated for use in screening.
Patients with a history of cardiovascular disease were excluded
leaving 9,232 subjects [male 52.6%, median age 61years(IQR 21),
Type 1 diabetes 16.7%]; 1,038 cardiovascular events (123 deaths)
were recorded over the following 5 years.
Results: The adjusted risk of developing cardiovascular disease
was higher in patients with any retinopathy by manual assessment
at baseline (HR 1.21, p < 0.05) and the risk was higher in patients
with Type 1 diabetes (HR 2.24, p < 0.01). The increased risk was
maintained with the sub-categories of cardiovascular disease
including coronary heart disease (HR 1.21, p < 0.05) and
peripheral vascular disease(HR 1.54, p < 0.001) but not with
cerebrovascular disease. Automated features associated with an
increased cardiovascular risk were high microaneurysm count (HR
1.23, p < 0.05) and high haemorrhage score (1.57, p < 0.05), and
the risk was higher in people with Type 1 diabetes for high
microaneurysm count (HR2.70, p < 0.001).
Conclusion: In our cohort the presence of retinopathy or a high
automated microaneurysm count at retinal screening increased the
future 5 year risk of cardiovascular disease.
P200
Is diabetes an important risk factor for
posterior circulation stroke? Lessons learnt
from retrospective case series
P Boovalingam, L Brawn, C Miller, N Pipalia, V Moore, R Day,
M Blake, V Srinivasan and D OKane
Department of Medicine, Northampton General Hospital, Northampton, UK
Introduction: Diabetes is increasingly recognised as an

independent risk factor of cardio cerebrovascular disease in
particular posterior circulation stroke. Our aim was to inspect
the association between diabetes and posterior circulation stroke
and its clinical outcome and to assess whether diabetes, in
particular with chronic kidney disease (CKD), affects the
outcome and mortality in a large cohort of patients with
ischaemic strokes.
Methods: We examined the association between diabetic patients
with and without CKD and 3 month outcomes (using mRS)
including mortality in 123 consecutive patients with posterior
circulation stroke admitted to Northampton General Hospital
for 6 months from January 2013 toJune 2013.
Results: Among 123 patients with a posterior circulation stroke
(mean age 74.5 6.3 years; 63% men), 66% had diabetes (98%
Type 2; 1.8% Type 1) and amongst those 43.2% had CKD (37.2%
stage 3a, 24.1% stage 3b and 38.7% stage 4). Diabetic patients
with moderate to severe renal impairment (CKD stages 3b and 4)
had >7-fold adjusted hazard ratio for mortality over 3 months
(6.32, 95%CI 1.92 12.84) compared with patients with no
2014 The Authors.

DIABETICMedicine
impairment. Also the location of the stroke was brainstem and

occipital cortex in 12% with no impairment, 32% with mild
impairment (CKD stage 3a) and 44% with moderate/severe
impairment (CKD stages 3b 4) (p = 0.004) and more brainstem
strokes with moderate/severe impairment. The clinical outcome
(using mRS) and mortality at 3 months were 42.3% and 30.9%
respectively.
Conclusion: Diabetes is a strong independent predictor of
mortality in patients with posterior circulation stroke. The
presence of moderate/severe CKD among patients with
diabetesand ischaemic stroke is linked with brainstem strokes
and with unfavourable clinical outcome.
P201
Glycaemic management in acute stroke
S Muktar, N Hill, J Meegan, C Jairam, S Misra, O Geraghty and
N Oliver
1
Diabetes and Endocrinology, Charing Cross Hospital, Imperial College NHS
Trust, London, UK, 2Hyperacute Stroke Unit, Charing Cross Hospital, Imperial
College NHS Trust, London, UK
Background: Diabetes increases the risk of stroke 2 4 times

compared with people without diabetes. Hyperglycaemia
co-presenting with stroke is common and has been associated
with negative outcomes although information supporting tight
glycaemia in acute stroke is limited.
Aims: To review hyperglycaemia management in patients
presenting to the hyperacute stroke unit.
Methods: We prospectively assessed all patients with stroke and
known diabetes or hyperglycaemia over a 4 week period.
Management was compared with local Glycaemic Management
in Acute Stroke guidelines.
Results: Twenty-three patients were included; 22 had infarcts
(three were thrombolysed). Sixteen (70%) were female, with
median age 77 years (IQR 66 84). All but one had pre-existing
Type 2 diabetes; three were dietcontrolled and 7/19 (37%) insulin
treated. Six (26%) had an initial capillary blood glucose
(CBG) > 12mmol/l. In the first 48h, 12/23 (52%) had CBG
readings maintained in the normal range (4 11mmol/l) and 14/
23 (61%) had CBG assessed every 4 6h. Nine of 11 with a
CBG > 11mmol/l were referred to the diabetes specialist nurse.
Only 2/12 (16%) patients on metformin had this withheld during
the first 48h. All seven patients with an intact swallow continued
their usual insulin. There were no episodes of hypoglycaemia in the
first 48h.
Discussion: Local glycaemic management in acute stroke
guidance supports safe glycaemia without hypoglycaemia in
people with diabetes and acute stroke on a tertiary hyperacute
stroke unit. Further work is needed to minimise hyperglycaemia in
the first 48h and a prospective glycaemic control study with stroke
outcomes is planned.
87
DIABETICMedicine
Clinical care and other categories posters: children,

adolescents and young adults
P202
Improving clinical outcomes in a childrens
diabetes service using tools from national
healthcare improvement and patient safety
programmes
V Alexander, SA Greene, S Bluff, V Ramsay, C Sturrock,
V Dunbar and HL Mitchell
Paediatrics, NHS Tayside, Dundee, UK
Refer to Oral A57
P203
Hospital admissions in young people with
Type 1 diabetes: results from a large
national matched cohort study
D Thayer1, A Sayers2, J Harvey3, M Atkinson1, J Warner4,
S Wong5, CM Dayan5, S Luzio1 and J Gregory5
1
College of Medicine, Swansea University, Swansea, UK, 2Department of
Social and Community Based Medicine, University of Bristol, Bristol, UK,
3
School of Medical Sciences, Bangor University, Bangor, UK, 4Department of
Child Health, University Hospital of Wales, Cardiff, UK, 5Institute of Molecular
and Experimental Medicine, Cardiff University, Cardiff, UK
Refer to Oral A58
P204
A third of patients within paediatric clinics
have endogenous insulin production:
implications for diagnosis and management
M Shepherd1,2, The UNITED team1, C Moudiotis3, B Fraser4,
K Mallam5, J Cox6, R Smith7, E Pearson8, B Shields1 and
AT Hattersley1
P206
Influence of a short-term school-based
intervention on the health behaviours and
prevalence of obesity among schoolchildren
and their parents: results of RATIONAL
HEALTH pilot programme
S Watson, AK Gupta and N Poulter
International Centre for Circulatory Health, Imperial College London, London,
UK
Refer to Oral A61
P207
Insulin adjustment for blood glucose and
carbohydrate content of meals from day 1
of diagnosis in children and adolescents
with Type 1 diabetes improves glycaemic
control by 3 months in a home-based
education programme
HL Mitchell, SA Greene, V Alexander, S Bluff, L McKenzie,
V Ramsay and C Sturrock
Child Health, NHS Tayside, Dundee, UK
Refer to Oral A62
P208
Hearing voices: do it in STYLE (Safe
Transition to Young Adult Life)
AD Thynne, PS Kar, J Head and L Albon
Diabetes Centre, Queen Alexandra Hospital, Portsmouth Hospitals NHS Trust,
Portsmouth, UK
Refer to Oral A60
Aims: Transitioning from paediatric to adult services is difficult

due to change of staffing combined with the typical challenging
nature of engaging young people(YP). This niche group is difficult
to engage in traditional didactic methods, so a shift from
traditional NHS style clinics to more modern strategies in
keeping with todays lifestyles is required. An existing YP service
was thus re-structured to encourage independence, autonomy,
adopting responsibility for their own diabetes and developing
confidence in using services.
Methods: Social media and patient focus groups provided
platforms to develop new ways of thinking based on thoughts and
ideas presented by YP. From this, strategies were developed (Safe
Transition to Young Adult Life, STYLE) to enhance engagement and
deliver alternative educational techniques including a formalised
relaxed transition clinic in partnership with a paediatric team
involving YP/parent, an educational group focusing on approaching
risky behaviours safely delivered to YP/parent separately allowing
freedom of discussion, pre-clinic topic discussion sheets to guide
consultations completed by YP, local university clinics, direct and
adhoc access to familiar staff, IT based routine screening tracker etc.
Results: Satisfaction questionnaires completed by the YP/parent
following educational groups were overwhelmingly positive. DKA
admission rates have reduced by 16%;DNA rates have reduced
from 45% to 15%. Frequent flyers have dropped from 7 to 2.
88
2014 The Authors.

1
National Institute for Health Research (NIHR) Exeter Clinical Research Facility,
University of Exeter Medical School, Exeter, UK, 2Research and Development,
Royal Devon and Exeter NHS Foundation Trust, Exeter, UK, 3Paediatric Unit,
Royal Devon and Exeter NHS Foundation Trust, Exeter, UK, 4Diabetes
Research Department, South Devon Healthcare NHS Foundation Trust,
Torquay, UK, 5Directorate of Child Health, Royal Cornwall Hospitals NHS
Trust, Truro, UK, 6Paediatric Department, Northern Devon Healthcare NHS
Trust, Barnstaple, UK, 7Children and Young Peoples Outpatient Department,
Plymouth Hospitals NHS Trust, Plymouth, UK, 8Division of Cardiovascular and
Diabetes Medicine, Medical Research Institute,University of Dundee, Dundee,
UK
Refer to Oral A59
P205
Improving transition from paediatric to
adult diabetes services: an audit of
changing models of care in a district general
hospital
J Witczak, M McClatchey, B Mumford, V Oldham, O Okosieme
and N Agarwal
Diabetes Centre, Cwm Taf Local Health Board, Merthyr Tydfil, UK
Clinical care and other categories posters: children, adolescents and young adults
Conclusion: Commissioner agreement allowed for a specific YP

focused and dedicated team. Risky behaviours are addressed in
non-judgemental fashion,with novel education techniques ensuring
YP remain safe while enjoying relatively normal lifestyles. There
are no additional cost implications but an ethos of working
differently within the same financial envelope.
P209
Young adolescent diabetes transition clinic
in a city hospital: are we making a
difference?
J Lloyd1, Z Khan1, L Hudson2, H Gandhi1, M Bandhakavi2 and
P De1
1
Department of Adult Diabetes and Endocrinology, City Hospital,
Birmingham, UK, 2Department of Paediatric Diabetes, City Hospital,
Birmingham, UK
Aim: To audit the outcome of our local hospital adolescent

diabetes clinic.
Method: Adolescents aged 16 24 years are seen by a
multidisciplinary team [adult and paediatric diabetes consultant,
diabetes specialist nurse (DSN) and a dietitian] every 4 6 months.
We looked at the clinic records and electronic data of 52 patients
over 2 years (2011 2012).
Results: Forty-four had Type 1 diabetes, eight had Type 2
diabetes (31 males, 21 females), mean age 20 (range 17 23
years) and average duration of diabetes 9 years (Asian 27,
Afro-Caribbean 10, Caucasian 15). Two patients had coeliac
disease. Two used recreational drugs, four were smokers and 12
took alcohol. Retinopathy was seen in four and microalbuminuria
in five patients over these 2 years. Mean HbA1c in 2011 was 81
and in 2012 was 77 (year on year decrease in HbA1c since 2010).
All 52 patients are on multi-dose insulin injection; 22 patients are
DAFNE trained. There was no increase in DKA related hospital
admissions (five in 2011, five in 2012). 64% did not experience
any hypoglycaemia; four one episode per week and another four
one per month. Overall 20% experienced hypoglycaemia in 2012
compared with 34% in 2010. DNA rates remained low at 22%
(2012) compared with 23% (2011).
Conclusions: Our ethnically mixed clinic audit now has 100%
patients on multi-dose insulin injections. DAFNE enrolment needs
to be improved. Clinic attendance is better in this difficult cohort:
DKA hospital admission rates have remained stable. Through more
frequent DSN contact (by emails, text messaging, drop-ins), we
have seen HbA1c improvements for the first time.
DIABETICMedicine
programme and (ii) the key issues for young people with long-term
conditions and professionals involved in the transition process.
Methods: The following databases were searched from inception
to August 2012: MEDLINE, EMBASE, PsychINFO, CINAHL,
ASSIA, Social Services Abstracts, Academic Search Complete,
Social Science Citation Index, Cochrane and Campbell Libraries.
Selected studies included young people aged 11 25 diagnosed with
long-term conditions who were in transition from paediatric to
adult secondary healthcare services.
Results: Sixteen systematic reviews and 13 primary studies were
included from 9,992 records retrieved. No single transition model
was uniquely effective. The most successful transitions centred
around the following factors: young personfocused; age
and developmentally appropriate content and delivery; selfmanagement education; family participation; paediatric and adult
collaboration; designated transition clinics; transition coordinator;
young persons portfolio; specific professionals training;
multidisciplinary approach; structured process embedded in
service delivery. There were no distinctive characteristics of
condition-specific Type 1 diabetes services.
Conclusion: This important and timely review summarises the
key factors that need to be considered for the development of
transition programmes for young people with long-term
conditions, including those with Type 1 diabetes.
P211
A jointly staffed adolescent transitional
diabetic clinic improves glycaemic control
and attendance rates on transfer to the
adult service at a local district general
hospital
Y Khalid1, N Martin2 and S Joseph1
1
Diabetes and Endocrinology, East Kent Hospitals University and NHS
Foundation Trust, Margate, UK, 2Paediatric Endocrinology, East Kent
Hospitals University Foundation NHS Trust, Canterbury, UK
Aims: For many young people with Type 1 diabetes, transition

from paediatric to adult care can result in a marked deterioration
in glycaemic control. A systematic review assessed the effectiveness
of transition models, or components of models, for managing the
transition process in young people with long-term conditions,
including Type 1 diabetes. This involved identifying (i) the main
barriers and facilitators in implementing a successful transition
Objectives: Guidelines recommend that young patients with Type

1 diabetes should be seen in a well supported transitional clinic
(TC) before their care is transferred to the adult clinic (AC).
However, studies that demonstrate improvement of glycaemic
control and other clinical indicators by this approach are scarce in
the literature. We set out to review clinical parameters and
attendance rates of patients as they transition through a jointly
staffed adolescent transitional clinic.
Methods: A retrospective analysis of patients seen in 2011 2012
was carried out. Glycaemic control (HbA1c), blood pressure (BP)
and attendance rates were analysed at three time points: last
paediatric clinic (PC), at TC and 12months into AC. Data are
expressed as the median of the obtained values.
Results: Thirty patient notes were reviewed. Median HbA1c at
PC was 11.1% (range 5.3 12.9%), 10.3% (range 6.2 to >14%) at
TC and 9.5% (range 6.3 13.9%) in AC. HbA1c was performed
on 28/30(PC), 23/24(TC) and 25/26(AC) patients respectively. BP
was recorded for 11/30 in PC, 5/30 TC and 18/30 AC. Median BP
in PC was 119/75 (range 161 113 over 68 89), and 145/76
(range 164 124 over 88 72) and 137/82 (range 155 104 over
106 60) in TC and AC respectively. Non-attendance rate was 1/
30, 6/30, 1/27 for PC, TC and AC respectively.
Conclusion: Our study demonstrates that the review of patients in
TC improves metabolic parameters as they transition to
AC. Attendance rate for AC was maintained despite a dip during
TC. Although BP checks were done less frequently in TC, this rose
2014 The Authors.

89
P210
A systematic review of transition models
for young people with long-term
conditions, including Type 1 diabetes
N Kime1, AM Bagnall2 and R Day3
1
Carnegie Faculty, Leeds Metropolitan University, Leeds, UK, 2Centre for

Health Promotion Research, Leeds Metropolitan University, Leeds, UK,
3
Institute for Health and Wellbeing, Leeds Metropolitan University, Leeds, UK
DIABETICMedicine
insignificantly on transfer to AC. The clinical relevance of this will

need to be explored in further larger studies.
P212
Barriers to good transitional diabetes care
in UK universities
KS Myint1, MJ Sampson1, C Price2, A Howes3, FM Swords1,
V Datta1, NK Thalange1 and J Kellett1,4
1
Diabetes and Endocrinology, Norfolk and Norwich University Hospitals NHS
Foundation Trust, Norwich, UK, 2University East Anglia Medical Practice,
University of East Anglia, Norwich, UK, 3Medical School, University of East
Anglia, Norwich, UK, 4Clinical Research and Trials Unit, Norfolk and Norwich
University Hospitals NHS Foundation Trust, Norwich, UK
Aims: We aim to describe the current model(s) of diabetes care

for UK university students (UUS) through a national survey of
healthcare professionals (HCP) in university medical centres/GP
practices (GPP) and diabetes specialist teams (DST).
Methods: A comprehensive online survey was developed and sent
to 220 sites. NVivo-10 software was used for the qualitative data
analysis.
Interim results: We received 170(88 GPP and 82 DST) completed
questionnaires (77% response rate as of 1 October 2013).The
responding GPP were spread across the UK, with 53% located
within a university campus. Diabetes care was provided by named
doctors (82.6%) and practice nurses (81%). There were no
protocols for emergency care (62.5%), for insulin management
(67%) or for diabetes in pregnancy (70.8%). In the DST survey,
service/facilities provided included young person clinics(89%),
inpatient young person units (6.8%), best practice tariffs
(53.5%), evening clinics (29.2%) andout-of-hours service
(27.8%).In both surveys, access to psychologist/counselling
facilities was felt to be inadequate or not available (26.6% in
GPP and 50.7% in DST). Only 32% of GPP and 32.4% of DST
were aware of the You are Welcome Department of Health
quality criteria for young people. The biggest challenges identified
were patient engagement, attendance, student lifestyle, ensuring
continuity of care, communication between healthcare providers,
patients reluctance to change providers, and inflexibility of
services. Suggested improvements were standardised national
protocols for transitional care, better communication/
coordination between providers/services, increased flexibility and
accessibility of services.
Conclusion: We have quantified areas of concern which will
provide a platform for national policy makers to improve diabetes
services for this at risk group in the UK.
P213
Introduction of the paediatric diabetes best
practice tariff associated with significant
improvements in HbA1c in a single centre
EJ Peters and NW Hopper
The Children and Young Peoples Diabetes Service, City Hospitals Sunderland
NHS Foundation Trust, Sunderland, UK
whether the increased funding and the implementation of changes

required to achieve the criteria set in the tariff were associated with
any change in HbA1c results in the clinic population.
Methods: We conducted a retrospective analysis of HbA1c results
from our clinic cohort over the last 6 years, specifically looking at
results in 2011, before BPT introduction, and 2013, after its full
implementation.
Results: Comparing 2011 to 2013, the median HbA1c fell from
74mmol/mol to 63mmol/mol (p < 0.01), the proportion of patients
with an HbA1c < 58mmol/mol rose from 13% to 26.3%
(p < 0.01) and those with an HbA1c > 80mmol/mol fell from
30.2% to 12.9% (p < 0.01).
Conclusions: There has been a dramatic improvement in HbA1c
in the clinic population from 2011 to 2013. This occurred at the
time of introduction of the BPT and significant investment in the
clinical service.
P214
Developing stakeholder agreed guidelines
for the support of children and young
people with diabetes in the school setting
MG Williams1, K Wilson2, C Pesterfield3 and CL Acerini4
1
Paediatric Diabetes, Queen Elizabeth Hospital NHS Foundation Trust, Kings
Lynn, UK, 2The East of England Paediatric Diabetes Network, Cambridge, UK,
3
Paediatric Diabetes, Cambridge University Hospitals NHS Foundation Trust,
Cambridge, UK, 4Paediatric Endocrinology and Diabetes, Cambridge
University Hospitals NHS Foundation Trust, Cambridge, UK
Currently in the UK children and young people with diabetes

receive variable provision of care and support in the school setting.
There are concerns that this impacts on the young persons
glycaemic control, their quality of life, and their educational
performance and outcome. Whilst various guidelines exist in some
schools, none has been developed based on a consensus,
multi-agency approach and none has been regionally or national
adopted. Our aims and objectives were to develop a
comprehensive, consensus-based, guideline that would inform
education providers about diabetes and provide a framework for
the best practice management and support of young people in
schools. This was achieved by convening a series of multi-agency
stakeholder (including for example clinicians, patients/families,
teachers; local authority and charity representatives) workshops in
an established regional paediatric diabetes network. The final
document produced provides specific guidance for all key parties
regarding their roles and responsibilities in the day-to-day care of
children with diabetes, including their legal (indemnity/liability)
obligations and their approach to safe diabetes care in a number of
school related activities (e.g. sport, school trips). These guidelines
are currently being trialled in the east of England and have been
positively received by education providers and families with
diabetes;they have attracted the attention of the Department of
Education who will link them to their Managing long term
conditions at school protocol.
Acknowledgement: All authors on behalf of the East of England
Paediatric Diabetes Network.
Introduction: In April 2012 the Department of Health launched

the paediatric diabetes best practice tariff (BPT). This required
strict criteria to be fulfilled in the care of each patient, in order to
receive a yearly tariff of 3,189, providing the opportunity for
units to significantly increase income, allowing for increased
expenditure on care. The children and young peoples diabetes
service achieved the tariff targets in 2012, which allowed for
significant expansion of the team in 2012 and 2013. We assessed
90
2014 The Authors.

Clinical care and other categories posters: children, adolescents and young adults
P215
Could diabetic ketoacidosis at presentation
of Type 1 diabetes in children predict higher
insulin requirements?
B Khodaghalian1, AU Nayak1, GI Varughese1 and P Raffeeq2
1
Diabetes and Endocrinology, University Hospital North Staffordshire NHS
Trust, Stoke-on-Trent, UK, 2Paediatrics, University Hospital North
Staffordshire NHS Trust, Stoke-on-Trent, UK
Background and aims: The regional paediatric diabetes register

at Stoke-on-Trent is validated, capturing 100% data on children
with diabetes in the region. We aimed to study the standard process
parameters at diagnosis and 3 months post diagnosis of Type 1
diabetes in the paediatric cohort. We explore the potential
association of diabetic ketoacidosis (DKA) at diagnosis with
insulin requirement and glycaemia at 6 months post diagnosis.
Methods: All children (age < 16 years) diagnosed with diabetes
between January 2010 and May 2012 were identified (n = 43).
Data on demographics, DKA at diagnosis, anti-GAD and anti-islet
cell (ICA) antibodies at diagnosis, glycaemia (HbA1c) at baseline
and 3 months, and insulin doses in the same time frames, were
obtained.
Results: Of n = 43 identified, the age was 9.6 4.4 years
(mean SD); 51% males. At diagnosis ICA was checked in all
and positive in 83.7%; anti-GAD was checked in 41 of 43 and
positive in 82.9%. HbA1c was checked in all at diagnosis and 3
months. The proportion with DKA at presentation was 12 (28%).
On regression analysis, HbA1c at diagnosis was a significant factor
(but not GAD, ICA, age, body mass index) associated with DKA at
presentation (HbA1c 130 31 vs 98 23mmol/mol, p = 0.001).
Those with DKA required higher basal insulin at 3 months, which
remained significant when corrected for HbA1c (F = 5.8, p
= 0.021).
Conclusions: A fair proportion of children have DKA as the
initial presentation of Type 1 diabetes (28% in our study). There
was a clear association of this initial presentation with higher
HbA1c at diagnosis and higher basal insulin requirements at 3
months post diagnosis.
P216
Ten years on: a review of paediatric patients
with diabetic ketoacidosis admitted to a
district general hospital
JS Woodside and ECA Stewart
Paediatric Department, Antrim Hospital, Antrim, UK
Aims: To compare diabetic ketoacidosis (DKA) admissions in

children to a district general hospital in 2000 with 2010.
Methods: This was a retrospective analysis of the case notes of all
children with Type 1 diabetes less than 18 years of age admitted
with DKA from 1 January to 31 December in 2000 and 2010.
Results: Attendance at the paediatric diabetic clinic increased
from 107 patients in 2000 to 165 patients in 2010. In 2000 there
were 44 episodes of DKA in 13 patients (average age 11.2 years)
compared with 23 episodes in 14 patients (average age 12.7 years)
in 2010. In 2000, 25% of newly diagnosed children initially
presented with DKA compared with 15.4% in 2010. The average
pre-admission HbA1c in the previously diagnosed patients was
9.9% in 2000 compared with 12% in 2010. Precipitating factors
were identified in 59% of those admitted in 2000 and 86% of those
admitted in 2010 with infection being the commonest precipitant
for both groups. The average length of hospital admission was 4
2014 The Authors.

DIABETICMedicine
days in 2000 and 2.4 days in 2010. In 2000, 75% admissions

received an intravenous insulin infusion compared with 100%
admissions in 2010.
Conclusions: This study supports that Type 1 diabetes in children
is increasing. Over the 10 year period, there has been a reduction in
the percentage of children with Type 1 diabetes admitted with
DKA. In future, an increase in those attending structured education
programmes and public advertising campaigns highlighting the
symptoms of Type 1 diabetes will hopefully lead to a continued
reduction in DKA.
P217
Giving diabetes street cred: improving
young persons diabetes service
MZ Chen, D Stern, C Grove, V Parfitt and F Chau
Diabetes Unit, Southmead Hospital, Bristol, UK
Aims: Transition into adult diabetes services can be challenging

for young adolescents. A tailored young persons diabetes service
(YPDS) is recommended to maximise patient engagement and
support diabetes management. Yet, there are few tools available to
establish the reasons for non-engagement and how to engage
young people with diabetes. To explore this, we conducted a pilot
survey evaluating patient experience using a trust-approved
questionnaire.
Methods: Surveys in paper and online forms were given to clinic
attendees using YPDS. Twenty patients, 9% of our patient
population, 50% female, age between 17 and 25, diabetes
duration 8.8 2 years, responded. Eight questions were
included to establish patients preferred clinic style, choice of
personnel, education updates and their basic understandings of
diabetes management.
Results: Multidisciplinary clinic with access to all healthcare
professionals away from adult outpatient services was the most
preferred set-up. Diabetes specialist nurse, dietitian and physician
were the personnel of choice. Two-fifths of responders asked for
regular educational updates through email. One-fifth of the
responders preferred meeting peers with diabetes in a support
group. Overall, responders relished our existing clinic set-up, staff
and educational materials. They highlighted areas requiring
reinforcement amongst our patients including insulin dose
adjustments during examination period and illnesses.
Conclusion: We identified areas of strengths and weaknesses
within our YPDS using this survey. This enables us to enhance our
services to motivate patient engagement and provide focused
educational updates for our young patients. We intend to develop
an easy-to-use assessment tool to regularly evaluate and improve
our YPDS based on this.
P218
A stable incidence rate of childhood onset
Type 1 diabetes in Wales: the influence of
birth cohort
JN Harvey1, R Hibbs1, H OConnell2 and JW Gregory2
1
Wrexham Academic Unit, Bangor University, Wrexham, UK, 2Department of

Paediatrics, Cardiff University, Cardiff, UK
Aims: To determine the time trend in incidence of Type 1 diabetes

in Wales and the relative influence of period (around time of
diagnosis) and birth cohort factors.
Methods: Data on new cases of Type 1 diabetes before age 15
were obtained from all paediatric diabetic clinics seeing Welsh
children. Ascertainment was checked by capture recapture analysis
91
DIABETICMedicine
in 2006 and 2013 using data from primary care. Incidence rate was
modelled using Poisson regression analysis. Age-period-cohort
modelling was undertaken using a novel method we derived to
avoid the problem of identifiability with birth cohorts constructed
using exact date of birth.
Results: From 1995 to 2012 2,676 newly diagnosed cases were
identified. Ascertainment was satisfactory. Best fit models indicated
no further increase in incidence rate since 2000 and no further
increase before age 5years. The rate is currently 29 per 100,000
children annually. Changes in incidence rate were primarily
determined by birth cohort with period having minimal
influence. The 1985 1989 birth cohort had a risk of diabetes

1.5 times that of the 1980 1984 cohort (p < 0.001). Subsequent
birth cohorts showed a similar difference to the 1980 1984 cohort
indicating no further increased risk.
Conclusions: The incidence rate of childhood Type 1 diabetes in
Wales is high but not currently increasing, unlike in most other
nations. The major influence of birth cohort suggests that
aetiological factors in the perinatal period are predominant.
Acknowledgement: On behalf of the Brecon Group.
Clinical care and other categories posters: delivery of care

P219
The Lost Tribe: an audit of transition care
in Lothian
CJ Kyle and N Zammitt
Department of Diabetes and Endocrinology, Royal Infirmary Edinburgh, NHS
Lothian, Edinburgh, UK
Aims: Adolescence is a time of transition when chronic disease

management is challenging. This audit examined the follow-up
status, glycaemic control and complication rates in 18 23-yearolds with Type 1 diabetes in Lothian. Sub-group analysis compared
outcomes at the Royal Infirmary Edinburgh (RIE) and Roodlands
Hospital Haddington (RHH). Secondary analysis assessed
pregnancy outcomes.
Methods: Subjects were identified by date of birth using SCI-DC.
Follow-up status, mean HbA1c, albumin creatinine ratio (ACR),
retinopathy status and complications were recorded for the last
year of adolescent follow-up and years 1 5 from transition.
Results: A hundred subjects were included in the audit: 77 from at

RIE (39 female) and 23 from RHH (seven female). RIE and RHH
respectively lost 27.3% and 39.1% to follow-up. Mean HbA1c at
RIE fell from 87mmol/mol to 77mmol/mol by year 5 and at RHH
from 81mmol/mol to 65mmol/mol by year 4. Those lost to
follow-up had higher HbA1c at baseline (RIE, 90mmol/mol; RHH,
87mmol/mol). Twelve patients had established microalbuminuria
after transition to adult care. Retinopathy was recorded in 36%
at year 5, of whomfive were referrable. There were 0.23
diabetes-related hospital admissions per subject per year and no
recorded deaths. Ten pregnancies occurred, two of which ended in
miscarriage. Only one patient took pre-conceptual folic acid. Mean
booking HbA1c was 83mmol/mol.
Conclusions: High losses to follow-up are of concern and those
with higher HbA1c appear most at risk. Complication and
admission rates are high, underlining the importance of
engagement with patients during transition.
Clinical care and other categories posters: diabetes

specialist nurse
P220
Patient satisfaction survey: views and
experiences of patients with diabetes
attending a consultation in a community
clinic with a community specialist diabetes
nurse
S Smith
Diabetes, Royal Wolverhampton NHS Trust, Wolverhampton, UK
Aim: To ascertain care experiences of those individuals with

diabetes who reside in the Trust, undergoing consultation in the
community with a community diabetes specialist nurse as part of
their diabetes care.
Method: The Picker Institute Europe pre-approved questionnaire
was utilised to ascertain 100 patients views and experiences
of their consultation with their community diabetes specialist
nurse.
Results: The majority of participants were White British with
Type 2 diabetes, aged 41 75 with an even distribution of gender.
92
Both knowledge and expertise are provided by the nurse, with

particular emphasis on partnership in care planning. 85% of
participants had agreed a care plan and 75% stated goals had been
discussed. However, only a third had recieved a written care plan
and 33% who did not receive one would have liked one. The level
of understanding of participants disease, tests and treatment
options was very good; however, more information on
the reason for tests is required. 99% stated that their nurse
listened carefully and explained things well, with length of time
spent a recurring theme. Issues with differing advice being given
amongst the wider team were evident.
Conclusion: The role of the community diabetes specialist nurse
within the community is important to patients. Three clear areas of
value emerged: approach to patient care; time; education/
empowerment. There is a need to implement formal written care
plans and education is required with regard to the need for
diagnostic tests. Advice within the wider multidisciplinary team
requires synchronisation.
2014 The Authors.

Clinical care and other categories posters: diet, obesity, exercise and inflammation
P221
Community diabetes specialist nurse: value
for money or a role the National Health
Service can ill-afford?
S Smith
Diabetes, Royal Wolverhampton NHS Trust, Wolverhampton, UK
Aim: To provide financial evidence of the role, value and cost

effectiveness of the community diabetes specialist nurse.
Method: Local financial and activity data were utilised to
determine activity and income for a single full-time diabetes
specialist nurse and a team of four full-time specialist nurses over
a 1 year period.
Results: Trust data identify a 12% non-attendance rate. Trust
data also identified that a single diabetes specialist nurse could
provide an income of 108,839 in 12 months if all appointments
were utilised. A team of four full-time nurses could provide a net
DIABETICMedicine
income of 599,057 per annum. However, data were deemed

inaccurate; amounts of contacts and types of contacts were not
recorded correctly resulting in a loss of recorded chargeable
income. Lost income for new patients alone totalled 202,980. If
income was recorded correctly for all new patients and follow-up
appointments a team of four full-time nurses income could be
802,037 on clinic activity alone. Activity excluded from financial
charging included home visits, telephone review, urgent care,
antenatal, adolescents, education/training.
Conclusion: Community diabetes specialist nurses are value for
money and if real charges as opposed to virtual charges were
being made a substantial six figure net income could be achieved.
Poor recording of data, inconsistences in data and poor
clarification of charging issues are preventing true income
generation from being unearthed. Excluded activity does not
incur a charge and is further lost income. Further income
generation would occur if 12% non-attendance was eliminated.
Clinical care and other categories posters: diet, obesity,

exercise and inflammation
P222
Patterns of weight change in the first 5
years after diagnosis of Type 2 diabetes and
their impact on glycaemic control and
mortality
L Aucott, S Philip and A Avenell
College of Life Sciences and Medicine, University of Aberdeen, Aberdeen, UK
Refer to Oral A67
P223
Patients with Type 2 diabetes achieve
similar weight loss in a community weight
management programme to those without
Type 2 diabetes
AC Wright1, A Brown1, R Golics2, E Maxwell2, R Davies1,
R John1, S Bellary1,3 and A Tahrani1,2
1
Specialist Weight Management Services, Heart of England NHS Foundation
Trust, Birmingham, UK, 2Centre of Endocrinology, Diabetes and Metabolism,
University of Birmingham, Birmingham, UK, 3Department of Life and Health
Sciences, Aston University, Birmingham, UK
Background: It is perceived that patients with Type 2 diabetes

lose less weight compared with patients without diabetes. Our aim
is to assess and compare the efficacy of a community-based weight
management (WM) programme in patients with and without Type
2 diabetes.
Methods: We conducted a longitudinal analysis of patients who
attended a 12 month community-based WM programme, which
includes education and regular contact with a specialist dietitian, a
consultant and a psychologist, between 2011 and 2013.
Results: In all, 144 patients were included in the analysis (mean
age 48 11 years, weight 126.9 24kg, body mass index
46 7.6kg/m2, women 82%); 90% completed 12 months. 46/
144 patients (32%) had Type 2 diabetes, of whom 10 patients were
insulintreated. There was no difference in attendance between
2014 The Authors.

patients with or without Type 2 diabetes. Patients with Type 2

diabetes were older (54 10.3 vs 46 11.3 years, p = 0.018)
and heavier at baseline (weight 130.4 27 vs 125.3 22.4kg,
p = 0.16). Patients with Type 2 diabetes lost a greater percentage
of body weight (BW) compared with patients without Type 2
diabetes (6.45 6.7% vs 4.4 9.1%, p = 0.23). 22.2% of
patients lost >10% BW with a similar proportion of patients
with and without Type 2 diabetes achieving >10% weight loss
(22.4% vs 21.7%). Insulin-treated patients with Type 2 diabetes
lost a similar percentage of BW to those not on insulin
(4.6 4.9% vs 6.9 7.8%, p = 0.27). After adjusting for age,
gender and baseline weight, Type 2 diabetes was not a predictor of
weight change (B = 0.28, p = 0.5).
Conclusions: Attending a community WM programme resulted in
significant weight loss in patients with Type 2 diabetes which is
comparable to those without Type 2 diabetes. Insulin-treated
patients lost similar weight to those not on insulin.
P224
An improved indicator of risk for
obstructive sleep apnoea: estimated
visceral adiposity compared with body mass
index and waist hip ratio in individuals
with diabetes
S Mazumdar, D Hughes and P Narendran
Diabetes Research Unit, Queen Elizabeth Hospital, Birmingham, UK
Introduction: Obstructive sleep apnoea (OSA) is strongly

associated with central obesity and Type 2 diabetes. Central
obesity is caused by an increase in visceral adipose tissue. Visceral
adiposity can be accurately measured with CT, MRI or DEXA
scans. The recent publication of a formula to estimate visceral
adiposity could provide a new inexpensive means to determine risk
of OSA.
Methods: Participants with diabetes were recruited to the the
CDR
Diabetes Study (an ongoing, observational, multicentred
study). As part of the study, body mass index (BMI), waist hip
93
DIABETICMedicine
ratio and estimated visceral adiposity (eVA) were recorded. Also,

participants completed the validated STOP-BANG screening tool
for OSA. A STOP-BANG score of >3 is considered moderate high
risk of OSA. SPSS statistical software was then used to determine
correlation coefficients and statistical probability.
Results: A total of 31 participants were analysed (Type 1
diabetes 21, Type 2 diabetes 10). The mean STOP-BANG score
was 3.88. The correlation coefficient values for STOP-BANG
scores were BMI 0.616 (p < 0.01), waist hip ratio 0.696
(p < 0.01) and eVA 0.852 (p < 0.01). Sub-group analysis of
those individuals with Type 1 diabetes: eVA = 0.809 (p < 0.01).
Conclusions: BMI, waist hip ratio and eVA all significantly
correlate to risk of OSA. eVA correlates stronger to OSA risk
than either BMI or waist hip ratio (even in individuals with Type
1 diabetes). Further research is required to determine whether eVA
has a role as a diagnostic tool for obstructive sleep apnoea.
P225
Development of a care pathway for patients
with Type 2 diabetes undertaking a very low
calorie liquid diet (VLCLD)
J Jackson, D Till, DW Lipscomb, JJ Bending, S Faulkner,
K Colvin, L Mackay and H ODonnell
Eastbourne District Diabetes Centre, Eastbourne District General Hospital,
Eastbourne, UK
Aim: To develop a multidisciplinary team (MDT) care pathway

for patients who choose a VLCLD to lose weight.
Methods: Following research by Lim et al.[1] the MDT had
begun to discuss (with selected patients with Type 2 diabetes) the
use of a VLCLD to enhance weight loss. Anecdotally these
patients subsequently demonstrated both a considerable weight
loss and an improvement in glycaemic control. It became apparent
that a care pathway was necessary to ensure a consistent approach
across the MDT and safety for the patient. The subject was fully
discussed at the weekly MDT meeting and ideas for the pathway
were generated. It was decided that the pathway should contain
the following elements: inclusion and exclusion criteria,
cardiovascular assessment, diabetes and lifestyle assessment,
motivation score, quality of life questionnaire, capillary blood
glucose monitoring, drug adjustment and patient information. A
flow chart summarised the pathway, detailing the medical and
dietetic reviews felt necessary by the MDT. Medical references
were to be included. Individual members of the MDT (consultant,
specialist nurse,dietitian) developed each of the above sections and
subsequent discussions took place resulting in changes being made
before the pathway was completed.
Results: A draft care pathway was finalised in August 2013 by the
MDT. It was agreed that the pathway would be registered with the
Trusts clinical audit department and that it will be evaluated.
Conclusion: Development of care pathways using members of the
MDT ensures the best use of each members expertise. This
pathway has resulted in a document which is both auditable and
owned by the MDT.
Reference: 1. Lim et al. Diabetologia 2011.
94
P226
Effectiveness and cost-effectiveness of
providing a local dietetic service to general
practitioner (GP) practices for people with
Type 2 diabetes and those at significant risk
of developing hyperglycaemia
E Smith
Nutrition and Dietetics Department, East Sussex Healthcare NHS Trust,
Hastings, UK
Aims: In 2010 GPs in Lewes, Newick and The Havens proposed a

new GP based diabetes dietetic service with the aim of identifying
whether having a dietitian working in the 10 GP practices in their
localities with people with Type 2 diabetes and those at high risk of
developing it (i.e. body mass index (BMI) > 30 with a comorbidity)
was cost effective at improving outcomes for their patients.
Method: Overall 564 patients were seen by the project dietitian. In a
two-way related samples design the independent variable was dietetic
input,pre- and post-intervention, with the dependent variables being
diet, physical activity, quality of life, patient satisfaction, confidence
to make changes, and a number of anthropometric and biochemical
outcomes (weight, BMI, waist circumference, HbA1c, total
cholesterol, LDL, HDL, triglycerides and blood pressure).
Results: Significant positive outcomes were demonstrated in this
study with 74% of patients with weight loss as a goal losing weight.
The average weight loss of these patients was 3.51kg, t(62) = 8.21,
p < 0.000 (two-tailed test), with this weight loss being sustained 6
12 months post-discharge. Additionally HbA1c reduced by
5.6mmol/mol 612 months post-discharge, t(9) = 1.48,
p = 0.173. The service was highly rated being scored 9.0 out of
10 on levels of satisfaction,t(59) = 62.3, p< 0.000 (two-tailed test).
Summary: The 2 year pilot project has been successful in
demonstrating the positive impact and cost-effectiveness that
having a GP based dietetic service for those with diabetes and
those at significant risk of developing it has on both health and
wellbeing outcomes.
P227
Effect of intensive preoperative and
postoperative glucose management on
glycaemic outcome after Roux-en-Y gastric
bypass surgery
LL Chuah1, A Miras1, T Olbers2 and CW le Roux1,3
1
Imperial Weight Centre, Imperial College Healthcare NHS Trust, London, UK,
Department of Surgery, Sahlgrenska Academy, University of Gothenburg,
Gothenburg, Sweden, 3Conway Institute, University College Dublin, Dublin,
Ireland
2
Introduction: Roux-en-Y gastric bypass (RYGB) surgery rapidly

improves glycaemic control in patients with Type 2 diabetes
post-surgery; however, there is no consensus on perioperative
glucose management of these patients, and its impact on glycaemic
outcome.
Method: In two randomised, non-blinded, single-centre trials, we
assessed the impact of intensive glucose management before and
after RYGB on glycaemic outcome and perioperative complications
of obese Type 2 diabetes patients. In study 1, 41 patients with
HbA1c > 8.5% undergoing RYGB were randomly assigned to
either glucose optimisation or non-optimisation 3 months before
surgery. In study 2, 42 patients on insulin undergoing RYGB were
randomly assigned to either intensive or standard glucose
management up to 2 weeks post-surgery. All patients were
followed up for 12 months.
2014 The Authors.

Clinical care and other categories posters: early detection and prevention
Results: In study 1, HbA1c improved from 10.3 (9.5 11.0)% to

6.3 (5.9 6.8)% in the non-optimised group (p = 0.0001, n = 18)
and from 9.9 (8.9 10.4)% to 6.9 (5.9 7.8)% in the optimised
group (p = 0.0002, n = 16). There was no significant difference in
HbA1c changes and hospital length of stay between the two
groups. In study 2, HbA1c improved from 8.5 (7.6 10.1)% to 6.2
(5.8 6.8)% in the standard group (p = 0.001, n = 17)and from
8.6 (7.3 10.3)% to 6.2 (5.7 7.1)% in the intensive group
(p < 0.001, n = 18). There was no significant difference in
HbA1c changes and length of stay between the two groups.
Conclusion: Optimisation of glycaemic control 3 months before
surgery or intensive management of fasting glucose 2 weeks after
RYGB has no significant impact on glycaemic outcome at 1 year.
P228
A newly developed obesity pathway
implemented by community dietitians and
dietetic assistants to measure outcomes
more accurately
K Treslove
Clinical Nutrition and Dietetics, Kent Community Health NHS Trust, Herne
Bay, UK
Aims: A specific package of care for obesity was identified as

being needed by our dietetic team in 2011, making costs of a
DIABETICMedicine
weight loss programme easier to calculate and outcomes measured

more accurately. Evidence suggests that as little as 5% 10%
weight loss can start to improve health related risk factors, such as
serum lipids, glucose tolerance and blood pressure. Our aim is for
participants to achieve an average weight loss of 5% or more over
6 months.
Methods: A 6 month programme was developed with an initial
60min appointment with the dietitian, followed by two further
30min appointments at 3 and 6 months. Up to 12 appointments
are offered in between with the dietetic assistant for weight check
and review of goals. All data are entered onto a database once
the participant has completed the 6 months. All staff have been
behaviour change trained including community dietetic assistants
who are relatively new within our team and have been a vital part
in delivering this new pathway. Signposting participants to
appropriate community schemes such as health walks, exercise
on referral and health trainers are also encouraged to
support further lifestyle changes.
Results: In all, 161 obese individuals have completed the pathway
over the last 12 months and have achieved an average weight loss
of 5.2%.
Conclusion: Without action, obesity related diseases will cost an
extra 45.5 billion per year. By offering a more structured,
supportive programme to obese patients, this could help
to encourage positive changes to be made for both the short and
longer term.
Clinical care and other categories posters: early detection

and prevention
P229
Categorising individuals at high risk of Type
2 diabetes is dependent on which validated
risk assessment is used
BJ Gray1,2, RM Bracken1,2, D Turner1,2, K Morgan3,
SD Mellalieu2, M Thomas4, SP Williams5, M Williams3, S Rice3
and JW Stephens1
1
Diabetes Research Group, Swansea University, Swansea, UK, 2Applied
Sports Technology Exercise and Medicine Research Centre, Swansea
University, Swansea, UK, 3Hywel Dda Health Board, Prince Philip Hospital,
Llanelli, UK, 4Public Health Wales, Carmarthen, Carmarthenshire, UK, 5TATA
Steel Packaging Recycling, Llanelli, UK
Refer to Oral A41
P230
HbA1c values are associated with changes in
risk scores in four commonly used validated
diabetes risk assessment tools
BJ Gray1,2, RM Bracken1,2, D Turner1,2, K Morgan3, SD
Mellalieu2, M Thomas4, SP Williams5, M Williams3, S Rice3 and
JW Stephens1
1
Diabetes Research Group, Swansea University, Swansea, UK, 2Applied
Sports Technology Exercise and Medicine Research Centre, Swansea
University, Swansea, UK, 3Hywel Dda Health Board, Prince Philip Hospital,
Llanelli, UK, 4Public Health Wales, Carmarthen, Carmarthenshire, UK, 5TATA
Steel Packaging Recycling, Llanelli, UK
Refer to Oral A43
2014 The Authors.

P231
Higher prevalence of undiagnosed glucose
intolerance in an urban adult Indian
population than in South Asians living in
the UK
V Kumar1, R Selvarajan1, R Gupta2 and S Rajbhandari3,4
1
Diabetacare, Bangalore, India, 2University of Bolton, Bolton, UK, 3Lancashire
Teaching Hospital, Chorley, UK, 4University of Central Lancashire, Preston, UK
Aims: Glucose intolerance is common amongst South Asian (SA)

adults in the UK and in the urban population of India. The aim of
the study was to find differences between the adult SA population
in the UK and Bangalore.
Methods: The general public were invited to come for a check up
including a capillary blood glucose test during events held at public
places. Invitations were sent by advertisement in local media,
pamphlets in public places and word of mouth. Simple history,
anthropometric data, blood pressure and random blood glucose
were measured. Only those subjects above the age of 30 and of SA
ethnicity were analysed.
Results: In all, 230 adults of SA origin in the UK and 412 in
Bangalore were screened at various events. In the UK 14.6% and in
India 42.7% (p < 0.001) were already known to have diabetes.
Amongst those not known to have diabetes, 8.5% in the UK and 21%
in India (p < 0.01) had impaired blood glucose (>7.8mmol) and
2.3% in the UK and 9.1% in India had possible diabetes
(>11.1mmol). Amongst people with diabetes, 22.7% in the UK and
52.8% in India (p < 0.01) had high blood pressure (>160/90mm Hg).
Discussion: In the UK the proportion of subjects with known
diabetes was low and their blood pressure was better controlled,
95
DIABETICMedicine
possibly due to universal NHS. There is no screening programme

for detection of diabetes in thetwo countries, despite which more
cases of undiagnosed impaired glucose tolerance and possible
diabetes were detected in India.
Conclusion: The resulting guidelines represent a notable example

of the value clinical networks can bring when seeking to
collaborate across organisational boundaries to develop regional
change.
P232
The value of clinical networks when seeking
to improve patient outcomes:
implementing a Merseyside diabetes
prevention pathway for those at highest
risk
P233
Missed opportunities for patients with
pre-diabetes: a study of the Salford
integrated record
A Ahmad , D McKinlay , RA du Plessis , J Rooney and

M Timoney1
1
Merseyside Diabetes Network, Cheshire and Merseyside Strategic Clinical
Networks, Warrington, UK, 2Public Health, Public Health England, Liverpool,
UK, 3Cheshire, Warrington and Wirral Area Team, NHS England,
Warrington, UK
Aims: To develop a Merseyside pathway and guidelines for the

identification and management of impaired glucose regulation
(IGR) with the aim of delaying or preventing the onset of Type 2
diabetes.
Methods: The clinically led Merseyside diabetes network worked
with a multi-stakeholder group including five public health teams,
six clinical commissioning groups and the Merseyside diabetes
patient action group to develop evidence-based guidelines for the
identification and management of IGR. The quality, innovation,
productivity and prevention agenda provided the platform on
which a business case was developed and funding for research
agreed. Activities undertaken include initial evidence review,
business case and cost benefit analysis; a systematic evidence
review and recommendations for effective interventions; a
questionnaire sent to all GP practices in Liverpool, Knowsley
and Sefton; a clinical audit of >600,000 patient primary care
records;the establishment of a multi-agency steering group;insight
with 65 people across Merseyside with risk factors for Type 2
diabetes.
Results: The Merseyside IGR pathway will be formally launched
in November 2013 to coincide with World Diabetes Day. The
pathway includes guidelines for the identification of IGR, initial
review in primary care, follow-up appointment with a health
trainer/lifestyle advisor, facilitated access into lifestyle services, and
annual review in primary care. In addition evidence-based
supporting materials have been produced including patient
information leaflets, goal planners and a bespoke package of
professional education.
96
H Elkhenini1,2, P Baker2, ND Stein2 and JP New1,2

1
Obesity Medicine and Endocrinology, Salford Royal NHS Foundation Trust
and University Teaching Hospital, Salford, UK, 2Manchester Academic Health
Science Centre, University of Manchester, Manchester, UK
Introduction: The evidence is suggestive that an increasing

number of patients present with macrovascular complications at
the time of diagnosis of diabetes. Efforts are required to identify
people with undiagnosed diabetes and people at high risk of
diabetes and offer them proper management.
Aims: To study the prevalence and cardiovascular risk of people
with pre-diabetes in our local community during the period
between January 2009 and June 2013.
Methods: Read codes are used to extract the study data from the
electronic medical record (EMR). Ethical approval was obtained
from the EMR board.
Results: In all, 225,346 EMR records are included in the study. A
higher percentage of the study population either have undiagnosed
diabetes or are at high risk of developing diabetes (7.35%)
compared with the people treated for diabetes (4%). 1,914
(0.85%) have read codes for diagnostic test results for diabetes
but are not on oral hypoglycaemic medications or insulin. A
comparable number of people with pre-diabetes have high systolic
blood pressure (SBP) ( 140mmHg, 25.2%) to those with diabetes
(25.8%, p = 0.3187) and a statistically significant number have
serum total cholesterol (TC) > 4.5mmol/l (68.3% vs 60.5%, p =
0.0001).
Conclusion: A high number of people with pre-diabetes, high SBP
and TC are identified using the EMR. Preventive interventions are
required in order to reduce complications. We believe that further
development of the EMR to flag patients at high risk of diabetes
and to create electronic alerts to prompt healthcare providers has
the potential to improve early management of people with
pre-diabetes and reduce vascular complications.
2014 The Authors.

Clinical care and other categories posters: education and self-management
DIABETICMedicine
Clinical care and other categories posters: education and

self-management
P234
Comparison of weight loss outcomes
between obese patients with and without
Type 2 diabetes attending a weight
management focused structured
educational group
P238
DAFNE-HART, a psychoeducational
programme to reverse hypoglycaemia
unawareness in Type 1 diabetes: report on
sustained biomedical benefit at 1 year, and
the user experience
G Abernethy1, V Mann1, R John1, J Shakher1, S Bellary1,2,

AA Tahrani1,3, AC Brown1, A Wright1 and R Davies1
M Stadler1, EM Shuttlewood1, H Rogers2, C Gianfrancesco3,

S Beveridge3, E Britneff1, P Choudhary1, J Elliott4, AD Rankin5,
SA Heller4, J Lawton5, SA Amiel1 and N deZoysa1
1
Specialist Weight Management Services, Heart of England NHS Foundation
Trust, Birmingham, UK, 2Life and Health Sciences, Aston University,
Birmingham, UK, 3Centre of Endocrinology, Diabetes and Metabolism,
University of Birmingham, Birmingham, UK
Refer to Oral A17
P235
Detrimental effects of dietary acculturation
in UK communities of black African and
Caribbean ancestry
LM Goff, L Timbers, H Style and A Knight
Division of Diabetes and Nutritional Sciences, Kings College London, London,
UK
Refer to Oral A19
P236
Pregnancy planning indicators following
the implementation of an educational
pre-conception counselling resource (DVD)
for women with diabetes into routine care
in Northern Ireland: a service evaluation
Diabetes Research Group, Kings College London, London, UK, 2Kings

College London Hospital, London, UK, 3Sheffield Teaching Hospitals, NHS
Trust Northern General Hospital, Sheffield, UK, 4Academic Unit of Diabetes,
Endocrinology and Metabolism, The Medical School, University of Sheffield,
Sheffield, UK, 5Centre for Population Health Sciences, University of
Edinburgh, Edinburgh, UK
Refer to Oral A22
P239
Influence of basal insulin (BI) regimen on
outcome after structured education in
adults with Type 1 diabetes in a Scottish
diabetes service
KJ Gardner1, RM Jacques2 and HE Hopkinson1
1
Diabetes Department, Victoria Infirmary, Glasgow, UK, 2School of Health
and Related Research, University of Sheffield, Sheffield, UK
Department of Diabetes, South Tyneside Foundation Trust, South Shields, UK
Aims: In current clinical practice BI regimens include once or

twice daily neutral protamine Hagedorn (NPH) insulin or
long-acting insulin analogue. This analysis investigates the
relationship between BI regimen and outcomes.
Method: HbA1c data from a large Scottish DAFNE service were
analysed in relation to BI type and frequency of injection at 12
months using a linear random intercepts model allowing for
clustering by hospital site, adjusting for baseline HbA1c, age and
duration of diabetes.
Results: Of 330 cases 311 pairs of complete data were identified
[41.8% male, mean age 39.7 years (SD = 12.4), mean duration of
diabetes 21.9 years (SD = 11.7)]. At 12 months there was a
significant reduction in HbA1c for the whole group from 71.8 to
68.3mmol/mol ( 3.5, 95% CI 4.7 to 2.3, p < 0.001) but the
benefit was only in those taking BI twice daily 72.5 to 67.6mmol/
mol ( 4.9, 95% CI 6.6 to 3.3, p < 0.001) vs once daily 70.9 to
69.2mmol/mol ( 1.7, 95% CI 3.5 to 0.1, p = 0.067). HbA1c in
people with poor control (baseline HbA1c 59mmol/mol, n = 267)
fell further: 75.2 to 70.8mmol/mol ( 4.4, 95% CI 5.7 to 3.0,
p < 0.001); twice daily 75.6 to 69.9mmol/mol ( 5.8, 95% CI
7.5 to 4.0, p < 0.001); once daily 74.6 to 72.1mmol/mol ( 2.5,
95% CI 4.6 to 0.5, p = 0.017). Linear modelling showed no
significant difference between different basal insulins (p = 0.296),
nor significant interaction between insulin type and injection
frequency (p = 0.171).
Conclusion: DAFNE is effective in a Scottish population and a
twice daily rather than once daily BI regimen is associated with
better HbA1c outcome at 12 months.
Refer to Oral A21
Refer to Oral A18
VA Holmes1, LL Hamill1, M Spence2, FA Alderdice3,

CC Patterson1, R Harper4, S Loughridge5, S McKenna5 and
DR McCance5
1
Centre for Public Health, Queens University Belfast, Belfast, UK, 2School of
Biological Sciences, Queens University Belfast, Belfast, UK, 3School of
Nursing and Midwifery, Queens University Belfast, Belfast, UK, 4Ulster
Hospital, South Eastern Health and Social Care Trust, Dundonald, UK,
5
Regional Centre for Endocrinology and Diabetes, Belfast Health and Social
Care Trust, Belfast, UK
Refer to Oral A20
P237
The 12 month impact of a group patient
education programme (Diet and Insulin to
Suit Yourself DAISY2) in patients with
Type 2 diabetes on basal bolus insulin
therapy
EH Gan, A Stewart, R Raine, B Marron, J Weir, K Dukhan,
J Mettayil and ST Wahid
2014 The Authors.

97
DIABETICMedicine
P240
Alcohol health literacy in young adults with
Type 1 diabetes and impact on diabetes
management
KD Barnard1, RIG Holt1, J Lawton2, JMA Sinclair3, AJ Young4,
A VanDenTol1 and P Dyson5
1
Human Development and Health Academic Unit, University of
Southampton, Southampton, UK, 2Centre for Population Health Sciences,
University of Edinburgh, Edinburgh, UK, 3Faculty of Medicine, University of
Southampton, Southampton, UK, 4National Institute for Health Research
(NIHR) Evaluation, Trials and Studies Coordinating Centre (NETSCC),
University of Southampton, Southampton, UK, 5Oxford Centre for Diabetes,
Endocrinology and Metabolism (OCDEM), University of Oxford, Oxford, UK
Background: Alcohol-specific health literacy is poor in the

general population, and this applies equally to young adults with
Type 1 diabetes. Alcohol directly affects glycaemic control and
accurate estimation of alcohol intake can support effective diabetes
management. Little is known about the accuracy of alcohol
knowledge and management strategies of young adults with Type
1 diabetes when consuming alcohol.
Aims: To investigate knowledge of alcohol and carbohydrate
content of commonly consumed beverages among young adults
with Type 1 diabetes; to explore alcohol consumption and to
identify diabetes self-management strategies for minimising
alcohol-associated risk.
Method: This was an open-access, 20 question, multiple-choice
websurvey investigating knowledge of alcohol and carbohydrate
content of typical alcoholic drinks using images. Participants
also recorded current alcohol consumption and diabetes selfmanagement strategies when drinking.
Results: In all, 547 participants aged 18 30 years (341 women,
192 men; mean age 24.5, SD 3.7 years) took part, of whom 365
(66.7%) drank alcohol. 84 (32.9%) female and 31 (22.6%) male
respondents scored above the AUDIT-C cut-off for increased risk
drinking. Knowledge accuracy of alcohol units was poor: only
7.3% (n = 40) correctly identified the alcohol content of 6 of 10
drinks. Knowledge of carbohydrate content was also poor: no
participant correctly identified the carbohydrate content of 6 of
10 drinks. Various and inconsistent strategies to minimise
alcohol-associated risk were reported.
Conclusions: Alcohol consumption was common amongst
participants; however, alcohol and carbohydrate health literacy
are poor. Greater alcohol related health literacy is required to
minimise alcohol-associated risk. Further research is required to
develop effective strategies to improve health literacy and support
safe drinking for young adults with Type 1 diabetes.
P241
A twice daily basal insulin (BI) regimen
achieves better glycaemic control than a
once daily regimen during structured
education in adults with Type 1 diabetes in
routine UK clinical practice
HE Hopkinson1, RM Jacques2, I Lawrence3, SA Amiel4 and
PM Mansell5
1
Diabetes Department, Victoria Infirmary, Glasgow, UK, 2School of Health
and Related Research, University of Sheffield, Sheffield, UK, 3Leicester
University Hospitals, Leicester, UK, 4Kings College Hospital, London, UK,
5
Queens Medical Centre, Nottingham, UK
Aim: Long-acting insulin analogues are often used once daily in

clinical practice, but the basal regimen in early effective structured
education programmes was twice daily neutral protamine
98
Hagedorn (NPH) insulin. This analysis investigates the

relationship between BI regimen and outcomes in people with
poor glycaemic control (HbA1c 7.5%) at presentation for
education in flexible insulin therapy.
Method: HbA1c data for people entering the Dose Adjustment
for Normal Eating (DAFNE) Research Database with HbA1c
7.5% were analysed in relation to BI type and frequency of
injection at 12 months using a multiple linear regression model
allowing for clustering by centre, adjusting for baseline HbA1c, age
and duration of diabetes.
Results: In all, 447 pairs of complete data were identified [47.2%
male, mean age 42.5 years (SD 14.4) and mean duration of
diabetes 18.2 years (SD 13.8)]. After DAFNE there was a
significant reduction in HbA1c from 9.15% to 8.87% (difference
0.28, p < 0.001), but the benefit was only in people using BI
twice daily (n = 277) 9.03% to 8.67% (difference
0.36,
p < 0.001) vs once daily 9.35% to 9.20% (difference 0.14,
p = 0.147). Linear regression modelling confirmed a significantly
lower HbA1c for the twice daily group (p = 0.002) and no
significant difference between different basal insulins (p = 0.156).
Relative risk of severe hypoglycaemia and ketoacidosis were 0.30
(p < 0.001) and 0.36 (p = 0.010) 12 months after DAFNE.
Conclusion: DAFNE is effective in improving glycaemic control
in routine UK practice but the benefit is associated with twice daily
rather than once daily BI regimen at 12 months.
Refer to Oral A18
P242
Understanding the impact of patient
education on the lives of people with
diabetes, inside and outside the health
system
R Snow1, J Sandall2 and C Humphrey1
1
2
School of Nursing and Midwifery, Kings College London, London, UK,

Womens Health and Social Science, Kings College London, London, UK
Aims: Research on patient education has looked mainly at

cost-effectiveness and medicallydefined outcomes. This study
explored several unaddressed areas, including what it means to
patients to take part and consequences for their lives long term.
Focusing on DAFNE (Dose Adjustment for Normal Eating), it is
one of the first studies in the field to be entirely service-user led.
Methods: A qualitative approach was taken, using narrative
analysis. Participants were recruited from established education
centres in three English NHS hospitals. Eleven new students were
interviewed pre-course, observed during education, and
re-interviewed 3 months later. Retrospective narratives were
collected from 10 former students. Data were gathered from 32
interviews and 146h of observations, including 5 days of complete
participation in DAFNE by a user-researcher.
Results: Most participants saw DAFNE as positive, speaking
about life-stories transformed. However, they were taught to aim
for biomedical targets they could not physically reach. This led to
some positioning their self-management work as failure, even
when they had better-than-average health outcomes. Obstacles
to self-management also arose post-course where generalist
healthcare professionals were uncomfortable with patients new
expertise. Personnel with the greatest power over medication
supplies had the least condition-specific knowledge, resulting in
tension when participants required things that their primary care
team deemed unnecessary.
2014 The Authors.

Conclusions: In terms of impact on patients life-stories,

DAFNE-style education can be said to be a success; but attitudes
in the wider health system, plus a lack of context on the
achievability of targets, may be undermining patients
self-efficacy and motivation post-course.
P243
Insulin safety and tomorrows doctors
AK Potts, N Robinson and A Crown
1
Diabetes Department, Brighton and Sussex University Hospitals Trust,
Brighton, UK
Aims: Patient safety is often compromised when insulin therapy is

used within secondary care. Local audits indicate 40.3% of
inpatients experience at least one error related to prescribing or
incorrect administration. A contributory factor may be the lack of
confidence and need for further training, which is self-reported by
qualified doctors. A diabetes education programme relating to
inpatient management was designed for third year medical students
to address these learning needs and to satisfy the standards relating
to therapeutic procedures and insulin therapy for undergraduates.
Methods: Evaluation included anonymous, qualitative comments
from students about which element/s of the course they found
useful. These were then grouped into nine clinical themes. Likert
scores also identified the students perception of the overall
relevance of the course content.
Results: Sixty-seven students attended the programme over 3
years. Enhancing knowledge about diabetes treatment, insulin
types and prescribing issues were ranked as the most useful aspects
of the course, scoring 16 20 responses each. Other themes scored
no higher than three responses each. Likert scores showed that
70% thought the overall relevance to practice was excellent and
29% felt it was good.
Conclusions: Third year medical students attending this course
valued the opportunity to enhance knowledge around diabetes
inpatient care. Increasing understanding about insulin therapy and
safe prescribing was perceived as the most relevant to their future
practice. Integrating these key areas early during undergraduate
education may improve practitioners confidence and patient safety
post registration.
P244
Integrating an evidence-based behaviour
change intervention targeting physical
activity behaviour into routine primary
care: Movement as Medicine for Type 2
Diabetes: an open pilot study
L Avery1, SJ Denton1, FF Sniehotta2, N Steen2, E McColl2,
R Taylor1 and MI Trenell1
1
Institute of Cellular Medicine, Newcastle University, Newcastle-upon-Tyne,
UK, 2Institute of Health and Society, Newcastle University,
Newcastle-upon-Tyne, UK
Objectives: To assess and optimise acceptability, feasibility and

fidelity of a multifaceted behavioural intervention Movement as
Medicine for Type 2 Diabetes in routine primary care.
Methods: An open pilot study was conducted in two primary care
practices in northeast England with six healthcare professionals
(two GPs, three practice nurses and one healthcare assistant) and
28 adults with non-insulin dependent Type 2 diabetes (14 males
and 12 females; mean age 70 10years). Healthcare professionals
completed an online accredited evidence-based training
programme prior to delivering a behavioural intervention
2014 The Authors.

DIABETICMedicine
targeting physical activity to participating patients. A qualitative

process evaluation identified (1) opportunities for optimisation of
the intervention and (2) barriers/enabling factors to successful
implementation. Video recordings of consultations assessed
treatment fidelity.
Results: The median time to complete the training programme
was 3h and 35min. Qualitative data identified opportunities for
intervention optimisation following initial delivery. Intervention
components (training programme and patient intervention) were
revised, reintroduced and delivered at 1month follow-up.
Adherence to intervention components was evident for the
majority of intervention components (e.g. goal setting, selfmonitoring). Barriers to implementation included lack of
opportunities for healthcare professionals to practise behavioural
techniques and negative beliefs about patient adherence to the
intervention. Enabling factors included increased self-efficacy for
effective use of behavioural change techniques and their
transferability to other patient populations.
Conclusion: Movement as Medicine for Type 2 Diabetes
represents the first programme designed, optimised and
accredited for primary care. The programme has been
demonstrated to be both acceptable and feasible in the primary
care setting.
P245
Downloaded glucose monitoring data for
insulin pump users: what matters?
E Jenkins, M Weiss, J Knott and D Kerr
Centre of Postgraduate Medical Research and Education, Bournemouth
University, Bournemouth, UK
Aims: Locally, blood glucose meters are downloaded to review

the effectiveness of self-management skills. Our aim was to assess
the value of different variables produced by downloads among
insulin pump users.
Methods: We obtained downloaded information from the
Accu-Chek 360o software on 192 patients (40% male)
established on insulin pump therapy for 3 (range 0.3 17) years,
consisting of the most recent 4 weeks of data including (a) number
of daily tests performed, (b) average blood glucose, (c) standard
deviation (SD) of measured values and (d) low (LBGI) and high
(HBGI) blood glucose indices. HbA1c was collected at the time of
download.
Results: HbA1c averaged 66.2 12.0mmol/mol (mean SD).
The average number of daily blood glucose tests was 5.28 2.53
with a daily blood glucose of 9.89 1.94mmol/l, a standard
deviation of 4.52 1.06mmol, LBGI of 1.69 1.29 and HBGI of
11.74 6.06. HbA1c correlated with HBGI (r = 0.740,
p < 0.01), average daily blood glucose levels (r = 0.739,
p < 0.01) and standard deviation of achieved glucose (r = 0.619,
p < 0.01). Using multiple linear regression analyses the strongest
predictors of HbA1c were number of daily tests performed
(p = 0.008), standard deviation of glucose levels (p = 0.032) and
female gender (p = 0.008). Overall, 61% of the variance in HbA1c
was explained by these variables.
Conclusion: Using easily accessible downloadable data on blood
glucose monitoring from insulin pump patients, the key modifiable
variables that impact on HbA1c levels are frequency of blood
testing and the standard deviation as a marker of glucose
variability. This type of software can be used to inform and
educate patients in routine clinical practice.
99
DIABETICMedicine
P246
PADDLE (a pictorial structured navigational
tool) used to deliver diabetes pregnancy
care
S Moutter1, MH Cummings1, J Head1, C Hall2 and A Going2
1
Diabetes Centre, Queen Alexandra Hosptial, Portsmouth, UK, 2Maternity
Unit, Queen Alexandra Hospital, Portsmouth, UK
Background: The number of patients with diabetes related

pregnancies is rising exponentially. Resources to deliver a
multi-healthcare professional (multi-HCP) are limited. Evidence
from previous studies suggests that group settings with a verbal
and pictorial approach offer improved educational opportunities
for patients and is more cost-efficient.
Method: The Portsmouth Diabetes Ante-Natal Team have
developed the first pictorial navigational tool (the PADDLE
Pregnancy and Diabetes Developing in a Learned
Environment) for use by multi-HCPs delivering diabetes
pregnancy care.
Results: In our audit of patients attending PADDLE
(pre-attendance
vs
post-attendance),
mothers
reported
significantly increased understanding of gestational diabetes
[scores pre 4/10 3.7 (mean ISD) vs post 9/10 1.0]. In
addition, PADDLE training resulted in 100% of mothers
understanding target glycaemia vs 0% pre PADDLE attendance;
71% of mothers were able to state which food groups affect blood
sugars vs 0% pre group attendance. Although mothers were
hesitant about attending group sessions, having completed their
PADDLE session their enthusiasm rose from 5/10 3.4 vs 9/10
0.9 with increased confidence about discussing diabetes in a group
environment (4/10 2.4 vs 6/10 2.9). Per hour, we have seen
2.5 times more patients than we would have hitherto seen on a one
to one basis.
Conclusion: We believe that group work using PADDLE (rather
than one to one consultations) can improve patient knowledge and
service efficiency, increasing the numbers of mothers with diabetes
that are seen in a timely fashion.
P247
Evaluation and utility of education
programmes and materials for those with
poor numeracy and diabetes, in eight
different countries
G Teft
Lews Castle College, University of the Highlands and Islands, Stornoway, Isle
of Lewis, UK
Introduction: Interest in poor numeracy has grown in recent

years; research demonstrates that poor numeracy skills adversely
affect health outcomes in many long-term conditions. This study
surveyed groups providing structured education for diabetes in
eight countries aiming to determine if poor numeracy was
considered during the creation, development and delivery of
structured education courses.
Methods: Groups providing structured education were identified
via the International Diabetes Federation (IDF), the Federation of
European Nurses in Diabetes (FEND) and the Diabetes Education
Network. Fourteen groups in eight countries, England, Wales,
Guernsey, USA, Sweden, Serbia, Portugal and Denmark, providing
structured education completed the survey.
Results: Thirteen groups (93%) had courses for Type 1 diabetes,
12 (85.7%) focused on carbohydrate counting. Most delivered
courses in groups (13 or 93%) and provided written materials for
100
participants. Only 50% (seven) assessed whether participants used

the material provided. No group assessed written material for
readability or numeracy skill required to understand it, assessed
participants reading or numeracy skills prior to the course, or
provided material specifically for poor numeracy skills; only one
group knew the level of poor numeracy in their population.
Summary: Structured education is provided in diabetes by many
healthcare professionals globally, with written material being provided
by most. Although poor health literacy and numeracy is common and
leads to poorer health outcomes, these results suggest that poor
numeracy is not considered or addressed when developing and
delivering structured education courses. Further research is required
to determine if addressing these issues would improve health outcomes.
P248
Type 2 diabetes education for adults with
intellectual disabilities
L Taggart1, V Coates1, MJ Davies2, B Bunting1, ME Carey2,
R Northway3, M Brown4, M Truesdale-Kennedy1,
L Martin-Stacey2 and PK Mandalia2
1
Institute of Nursing and Health Research, University of Ulster,
Newtownabbey, UK, 2Diabetes Research Group, University Hospitals of
Leicester, Leicester, UK, 3Faculty of Life Sciences and Education, University of
Glamorgan, Pontypridd, UK, 4Faculty of Health and Social Care Research,
Edinburgh Napier University, Edinburgh, UK
Aim: The aim of this study was to adapt and evaluate a selfmanagement Type 2 diabetes education programme (DESMOND) for
adults with intellectual disabilities (ID) and their carers.
Methods: Two iterations of the DESMOND education
programme were delivered in Northern Ireland. In total 13 adults
with ID with Type 2 diabetes and eight family/paid carers engaged in
this programme over a 6 week period in 2012 and 2013. Feedback
was obtained using focus groups with the adults, their carers, the
educators and three independent observers. Video recordings of the
sessions were also used to supplement this evaluation.
Results: The feedback from the first iteration of the DESMOND
programme illustrated that it needed to be adapted to accommodate
for the adults literacy skills and comprehension levels. The adults
with ID also required the support of their carers to develop and
maintain the skills to self-manage their diabetes at home, focusing on
diet, exercise and medication. The second iteration of the
programme supported the adaptations that have been made,
although other minor changes are still required. Family/paid carers
need a separate day to prepare them to engage effectively to help
support the adults to learn the objectives of the adapted DESMOND
programme. A third iteration is planned.
Conclusion: Adults with ID can be educated and supported to
develop the skills to self-manage their Type 2 diabetes. It is
important to have the support of family/paid carers. Such
self-management programmes developed for the non-disabled
population need adapting for adults with ID.
2014 The Authors.

P249
Using the internet to reduce the barriers
that prevent people with Type 2 diabetes
from being active in rural areas
J Connelly1, A Kirk2, J Masthoff3 and S MacRury1
1
Department of Diabetes and Cardiovascular Science, University of Highlands
and Islands, Inverness, UK, 2Department of Physical Activity for Health,
University of Strathclyde, Glasgow, UK, 3Department of Computing Science,
University of Aberdeen, Aberdeen, UK
Introduction: Physical activity is important in the management of

Type 2 diabetes; however, patients report low self-efficacy in this
area and this is likely to be compounded by rural location.
Aim: To use qualitative research to identify key features for a
web-based physical activity promotion tool that reduces barriers to
physical activity participation in rural areas.
Methods: Four focus groups were conducted by a trained
moderator with adults with Type 2 diabetes (n = 30) who lived
in rural Highland areas. A semi-structured interview guide was
used to explore barriers to physical activity. The sessions were
recorded, transcribed and analysed using the NVivo qualitative
analysis program and thematic analysis.
Results: Participants identified that health professionals were
unable to give structured physical activity advice and personal
trainers were unable to give diabetes-specific advice. Medication
adherence and dietary changes were thought more beneficial to
diabetes management than physical activity and all groups stated a
fear of hypoglycaemia with increased activity. Other adverse
factors expressed were lack of opportunities to be active when
living in a rural area compared with cities and the danger of being
active outside due to single track roads and poor lighting. A lack of
support systems was identified; in particular being active with
others was cited as difficult due to distance.
Conclusions: Diabetes-specific physical activity information is
needed to prioritise the importance of becoming more active
in diabetes management. A virtual support system offering
individualised advice and alternative safer options along with
activity maps to highlight local opportunities is needed.
P250
Knowledge of hyperglycaemia
management in acute coronary syndrome
P Hegde, H Shoo, N Ormsby and D Bowen-Jones
Diabetes and Endocrinology, Wirral University Teaching Hospital NHS
Foundation Trust, Wirral, UK
Aims: The management of hyperglycaemia in patients with acute

coronary syndrome (ACS) remains controversial. The inconsistent
evidence and lack of guidelines is a cause of concern. A survey was
undertaken to identify awareness of hyperglycaemia management
in ACS among hospital staff.
Methods: Multiple choice questionnaires were distributed with a
response rate of 90%. Of the 45 responders, five were consultants,
12 specialty trainee registrars, 19 were senior house officers, five
were foundation doctors and four were coronary care unit nurses.
Results: Of these, 46% would treat hyperglycaemia if the blood
glucose was above 11mmol/mol. 82% admitted that they request
blood glucose routinely in acute coronary syndrome. 22% were
aware that the target time to treat hyperglycaemia is 4h. 33% were
aware of NICE guidelines. 55% were aware of the DIGAMI trial.
58% would not routinely start insulin infusion as initial treatment.
31% were aware that the time to change to usual diabetes
medication when the blood glucose radings are stable is 48h. 73%
2014 The Authors.

DIABETICMedicine
would routinely attempt to formally diagnose diabetes in patients

with hyperglycaemia but not known to have diabetes. 31% were
aware that a formal test should be done no earlier than 4 days after
the onset of ACS. 82% were aware that they should involve the
diabetes team before discharge.
Conclusion: Lack of awareness of ACS management is
common. Education and dedicated guidelines may help address
these concerns.
P251
Prevalence of diabetes and identification of
high risk patients on Santa Cruz island in the
Galapagos archipelago
NK Tufton and T Chowdhury
Diabetes and Endocrine Department, Royal London Hospital, Barts Health
Trust, London, UK
Objectives: Many developing nations have seen a shift in the

burden of disease from malnutrition and infectious diseases to
those of obesity, diabetes and heart disease. The aim of this
survey was to offer a screening and education programme for
diabetes and hypertension on the island of Santa Cruz,
Galapagos, determine the prevalence of these disorders and
identify those at risk.
Methods: A 1 month screening programme was offered
in Puerto Ayora, Santa Cruz. The service was advertised locally,
and subjects self-presented to a free health screening
clinic. Demographic information and physical measurements
were obtained [weight, height, body mass index (BMI), fasting
capillary glucose, blood pressure, smoking status]. All subjects
were provided with education on nutrition and lifestyle changes.
Results: Of 179
patients
screened,
30
reported
a
previous diagnosis of diabetes (16.75%); of these 12 (40%) were
taking at least one medication; 85% of men and 82.5% of women
were overweight or obese (BMI > 25kg/m2). 141 patients had
fasting glucose readings taken: 30 (21.3%) had fasting glucose
readings > 7mmol/l and 26 (18.4%) had glucose readings of
5.6 6.9mmol/l. Sixteen (11.3%) and 22 (15.6%) were new
diagnosis of diabetes and impaired fasting glucose respectively.
Conclusions: This is the first reported study of diabetes and glucose
intolerance prevalence in Galapagos. A high number of patients
have suspected undiagnosed diabetes (11.3%) or are at high risk of
developing diabetes in the future with impaired fasting glucose
(15.6%). Urgent education and prevention programmes
are required to address this significant public health problem.
P252
So why do women with pre-gestational
diabetes not attend for pre-pregnancy
counselling?
A Going1, C Hall1, J Head2 and MH Cummings2
1
Maternity, Portsmouth NHS Trust, Portsmouth, UK, 2Diabetes Centre,
Portsmouth NHS Trust, Portsmouth, UK
Aims: Pre-pregnancy counselling (PPC) is associated with

improved outcome in pregnancies of mothers with
pre-gestational diabetes.However, take-up of PPC is poor. Our
survey aimed to understand in pregnant mothers with diabetes who
did not attend for PPC the reasons behind this.
Methods: Between December 2011 and August 2013 we
identified 27 women who did not attend for PPC (11 Type 1
diabetes, 16 Type 2 diabetes; 27% of mothers with pre-gestational
101
DIABETICMedicine
diabetes). They were invited to complete a questionnaire to

understand their non-attendance for PPC.
Results: A total of 85% recognised that diabetes may affect their
pregnancy. However, 41% did not understand the importance of
PPC. Only 48% of these mothers were aware that a PPC service
was available (82% Type 1 vs 31% Type 2) and, of those who
were aware, only 44% were aware how to access the PPC service.
44% of these pregnancies were unplanned.Other considerations
such as travelling distance to clinic (4%), costs to attend (4%) or
time away from work/other commitments (4%)had little bearing
on decision to attend PPC.
Conclusions: More robust methods of promoting awareness of
the PPC service are required to optimise attendance. This was a
bigger problem for mothers with Type 2 diabetes who receive their
diabetes care within the community. Greater education is also
required to highlight the risks of unplanned pregnancies in mothers
with diabetes and the benefits of PPC.
P253
Non-attendance at structured diabetes
education for Type 2 diabetes: patient
choice or poor marketing?
K Winkley1, C Evwierhoma1, SA Amiel2, A Forbes3 and
K Ismail1
1
Institute of Psychiatry, Kings College London, London, UK, 2Diabetes and

Nutritional Science, Kings College London, London, UK, 3School of Nursing,
Kings College London, London, UK
Aims: Despite widespread provision of structured diabetes

education for people with newly diagnosed Type 2 diabetes
relatively few people attend. The aim of this study was to
determine the key themes explaining non-attendance in people
with a recent diagnosis of Type 2 diabetes from three South
London boroughs.
Methods: A purposive sample of participants from the South
London Diabetes Cohort (SOUL-D) study who had not attended
the DESMOND programme was approached. The SOUL-D study
is a cohort of people with newly diagnosed Type 2 diabetes in
South London. Thirty people were interviewed using a
semi-structured interview and topic guide based on recent
literature and guidance from a local diabetes patient involvement
group. Participants were selected based on their age, gender,
ethnicity and borough of residence. Interviews were transcribed
verbatim and the data entered and managed in Nvivo 10. A
thematic framework method was used to detect themes for
non-participation, which were then grouped into meaningful
categories.
Results: Participants explanations for non-attendance at
DESMOND fitted into two central themes: whether or not their
health professional had told them about it. For the group who had
been informed about DESMOND but not attended, explanations
for non-attendance included how it was marketed; access; and a
preference for onetoone sessions.
Conclusions: One of the key explanations for non-attendance at
structured education is whether or not health professionals offer it
to their patients.
102
P254
Can group education for patients with Type
2 diabetes on multiple daily injections
improve knowledge and outcomes?
J Campbell and R McGee
Gloucestershire Foundation Trust, Cheltenham, UK
Aim: Due to the high numbers of patients with Type 2 diabetes on

basal bolus therapy and increasing one to oneconsultations, a
group education programme was developed with the aim of
improving knowledge and HbA1c in this group.
Methods: It was set up as a pilot study with a small sample size of
patients. Patients with Type 2 diabetes on basal bolus therapy were
given a letter after a clinic appointment to attend education (letters
were given out over a 3month period). Patients then opted to
attend the sessions. Two group education sessions for 4h were
devised to be delivered 1 month apart and six times per year at
both hospital sites by a diabetes specialist nurse and dietitian. The
sessions covered were carbohydrate awareness and insulin and a
diabetes refresher. Baseline measures included HbA1c and a
knowledge questionnaire pre and post course.
Results: A total of 26 patients attended session 1 over the six
sessions. Twenty-five patients attended session 2 over the six
sessions (one patient couldnot make it on the second session). 92%
stated that the course taught them new knowledge trends were
how to adjust insulin along with food and blood glucose control.
92% stated that the course had changed their understanding of
diabetes trends were adjusting insulin safely and effectively and
hypo management. 77% stated that they would make changes to
their diabetes management. 56% had a lower HbA1c 6months
post course; 4% had no change in HbA1c 6 months post course.
Conclusion: Group education has improved patients knowledge
and lowered HbA1c in a significant number of patients who
attended the two education sessions.
P255
Non-attendance in diabetes education
centres: perceptions of patients and
education providers
MO Lawal
Health and Human Science, University of West London, London, UK
Background: The NHS Outcome Framework focused on

measuring health outcomes in order to improve clinical practice
(DH 2012comment: Department of Health?) and several studies
have identified the beneficial effects of diabetes education in
promoting patient outcomes. However, non-attendance in diabetes
education clinics remains a big challenge to the implementation of
a national diabetes education policy for all newly diagnosed
diabetes patients in the UK.
Aims: To explore factors influencing non-attendance from the
perspectives of both healthcare professionals and patients.
Methods: The study used a two-staged sequential qualitative
method approach, incorporating a focus group technique, a paired
interview and one individual interview of diabetes specialist nurses,
podiatrists and dietitians (n = 10) delivering structured patient
education in a primary care trust in the southeast of England. The
second stage utilised a telephone interview of patients (n = 24) that
failed to attend the education session. Thematic analysis was used
to analyse the data.
2014 The Authors.

Results: The practitioners were challenged by complexities of the

patient and non-attendance was influenced by modifiable barriers
such as patients beliefs, culture, social norms and family influence.
The findings acknowledged the challenges posed by selfmanagement education funding and government regulations. The
psychological effects of diabetes and the impact of organisational
structure and ethos were identified as part of the barriers. In
addition, the attrition rate was influenced by non-modifiable
barriers such as bad weather and various personal intervening life
problems ranging from bereavement and work related issues to
childcare problems and ill-health.
Conclusion: Although the problem of non-compliance in clinical
practice is old, nevertheless it is complex and requires further
exploration.
P256
Impact of implementing care-planning
principles in specialist diabetes clinics
B Obale1 and M Banerjee2,3
1
Department of Medicine, University Hospitals of Morecambe Bay NHS
Foundation Trust, Kendal, UK, 2Wirral Diabetes and Endocrinology Unit,
Wirral University Teaching Hospitals NHS Foundation Trust, Wirral, UK,
3
Centre for Endocrinology and Diabetes Research Group, University of
Manchester, Manchester, UK
Objectives: Modified care-planning principles were implemented

in our specialist diabetes clinic consultations and patients response
was evaluated.
Method: Patients
anthropometric
and
blood
pressure
measurements were recorded on a sheet which included previous
and current visits biomedical information, the plan of action from
the previous visit and an area for the patients to document their
views and concerns regarding their control. They also had a
help-sheet and help from a nurse to understand the data on the
sheet. Anonymised feedback was requested after the consultation.
Results: Forty-nine completed feedback forms were returned from
52 follow-up attendees in six clinics (mean age 60.2 14.1 years,
male : female 29:15). The positive responses in the feedback forms
were as follows: Was the procedure explained by the nurse to you
in the clinic? 100%. Did you understand the results that were given
to you today? 100%. Did the doctor help you to understand the
areas you had struggled? 100%. Did you have a better
understanding of your diabetes today? 100%. Were your
concerns that you had put in been addressed today? 97.9%. Do
you have more confidence in your new plan of treatment? 100%.
Would you like to have similar consultations in future? 100%. The
doctors found the consultations to be more fluent and speedy.
Conclusions: Embedding care-planning principles in routine
secondary care diabetes clinic consultations were highly esteemed
by both patients and clinicians. The impact on the biomedical
aspects needs to be evaluated.
DIABETICMedicine
P257
How do patients and carers develop their
ability to self-manage diabetes? Evidence
from a practice-based study involving
families with a child diagnosed with Type 1
diabetes
CC Tranchant1 and D Charron2
1
School of Food Science, Nutrition and Family Studies, Faculty of Health
Sciences and Community Services, Universite de Moncton, Moncton, Canada,
2
School of Nursing, Faculty of Health Sciences and Community Services,
Universite de Moncton, Moncton, Canada
Aims: Diabetes care relies importantly on patients daily self-care

activities and care providers continuous support, but little is
known about how patients and their relatives develop the ability to
manage a chronic health situation like diabetes. This study aimed
at uncovering the processes involved in the development of
self-care in families with a child diagnosed with Type 1 diabetes
in the last 18 months.
Methods: A qualitative research with in-depth interviews was
conducted in an educational setting involving a university and
teaching hospital in Eastern Canada. Participants were 15
families (parents and children), 15 fourth-year BSc nursing
students (one/family) trained in the family systemic approach
and their instructor. Narratives, reflexive analyses and a
systemic-constructivist approach were used to provide accounts
of participants experience of self-care development from a patient
or carer perspective. Instructor and individual students met five
times; students and families met three times. Interviews were
recorded and transcribed. Data were subjected to thematic
analysis.
Results: Conceptual categories central to the experience of
self-care development were identified and used to develop an
educational theory that describes how participants co-developed
their mastery of self-care through engaging in four main interacting
processes: narration, self-dialogue, discordance, creation and
co-creation. These processes were related to principles of clinical
intervention, stages of learning and teaching strategies.
Conclusions: Findings suggest that distinct processes, engaging
both self and others in transformational change, contribute to the
development of the ability of families to manage Type 1 diabetes.
These new insights have implications for improving educational
care aimed at individuals and families.
P258
PITstop improves primary care healthcare
professionals confidence in supporting
people with Type 2 diabetes with more
complex needs
A Goodchild and A Kitt
Diabetes Ttraining, PITstop Diabetes, Kent, UK
Introduction: PITstop (Programme for Injectable Therapies) is an

accredited advanced diabetes course for primary care healthcare
professionals (HCPs). The course focuses on the practical aspects
of initiating and supporting people with Type 2 diabetes on more
complex medication regimens, including injectable therapies.
Method: Fifty-six general practitioners and practice nurses
attended the 2.5day PITstop course (September 2011 to January
2013). In September 2013 they participated in a survey focusing
primarily on their confidence, before and after the PITstop
2014 The Authors.

103
DIABETICMedicine
training, in initiating GLP-1 receptor agonists (GLP-1RA) and

insulin, and optimising treatment regimens.
Results: Out of 56 participants, 23 responded. Results are
presented as confidence post-training vs confidence
pre-training. Supporting patients in review appointments: 30.4%
excellent, 65.2% good, 4.4% satisfactory vs 13% good, 52.2%
satisfactory, 34.8% poor. Managing more complex oral
medication regimens: 34.8% excellent, 65.2% good vs 21.7%
excellent, 52.2% good, 26.1% satisfactory, 4.4% poor. GLP-1RA
initiation: 13% excellent, 82.6% good, 4.4% not initiated vs 4.4%
good, 30.4% satisfactory, 65.2% either not initiated or poor. Basal
insulin initiation: 26.1% excellent, 56.5% good, 13% satisfactory,
4.4% not initiated vs 8.7% good, 21.7% satisfactory, 69.6%
either not initiated or poor. Insulin titration: 13% excellent, 60.9%
good, 26.1% satisfactory vs 8.7% good, 13% satisfactory, 78.3%
poor or no titration. Changing insulin regimens: 13% excellent,
43.3% good, 30.4% satisfactory, 4.4% poor vs 8.7% good, 4.4%
satisfactory and 86.9% either poor or no experience at regimen
changes.
Conclusion: PITstop training improves HCP confidence in
delivering enhanced diabetes services, especially in providing
GLP-1RA and insulin initiation/titration and in changing insulin
regimens.
P259
Development of an online resource to
facilitate access to diabetes educational
resources for healthcare professionals and
patients
JRS McDowell1, A Taylor2 and S MacRury3
1
Nursing and Health Care, University of Glasgow, Glasgow, UK, 2NMAHP,
NHS Education for Scotland, Edinburgh, UK, 3Consultant Diabetologist,
Raigmore Hospital, Inverness, UK
Background: The Scottish Diabetes Action Plan identified the

need to create the post of National Education Coordinator (NEC)
to increase the quantity and quality of structured patient education
in Scotland in addition to coordinating work strands to improve
generic and diabetes-specific educational opportunities for the
workforce.
Aim: To develop a web-based resource that would host as well as
signpost both people with diabetes and professionals working in
diabetes to resources that support education within diabetes.
Method: The 2 year funded NEC postholder in collaboration
with regional diabetes managed clinical networks and the Diabetes
Patient Focus Group identified several current educational
resources applicable to patient and professional education. A
structure was determined for the website to include seven
individual pages for home; patients; professionals; key literature;
key websites; key professionals within each health board;
news. Each page was further structured to address key topics.
The overall structure was reviewed and agreed by the NEC steering
group and the Scottish Diabetes Education Advisory Group.
Results: The website was launched at the end of June 2013 and is
at www.diabeteseducationscotland.org.uk. To date (September
2013) 364 people have visited the site and viewed 1,661 pages.
Conclusions: The Diabetes Education Scotland Website is a live
resource providing access to local and regional educational
programmes for people living with diabetes and for all
healthcare professionals delivering diabetes education.
104
P260
An audit of participant evaluations
following attendance at diabetes education
self-management for ongoing and newly
diagnosed (DESMOND) sessions
H Noakes and C Tinson
Diabetes Department, Specialist Medicine, Lewisham and Greenwich NHS
Trust, London, UK
Aims: Participants are asked to complete an evaluation form

following a DESMOND session. These are checked by the
educators as part of their reflection and to see if there are any
issues to be resolved, but there has never been an audit. We aim to
audit the evaluations completed over a 6 month period to ensure
that we are meeting the needs of those attending.
Methods: The 133 evaluation forms completed at DESMOND
held from 1 October 2012 to 31 March 2013 were audited;
responses were collected on a spreadsheet and as text.
Results: In all, 85.5% of responses were positive; 100% of
respondents were clear about the key messages given. Two
questions out of 12 scored less than 90%
these
were depression and action planning.
Conclusion: DESMOND is well received by those attending and
we are meeting the needs of most. However, depression and action
planning, the two areas which scored lowest, are being addressed
by the team of educators. The aim is to improve the delivery of
DESMOND so that the attendees experience is positive and this
will aid self-management. Action planning will be introduced
earlier in the programme and discussed at several points during
the day and care will be taken to ensure that depression is
addressed in a sensitive, positive way. These results will be fed
back to the commissioners and we will also re-audit to ensure
scores have improved. If they have not, further investigation will be
required to ascertain the reasons for low scores on depression and
action planning.
P261
An exploration of personal constructs in
people with Type 2 diabetes in the
community
WP Gillibrand
Health Science, University of Huddersfield, Huddersfield, UK
Aim: To explore personal constructs in adults with Type 2

diabetes under primary care services.
Methods: We studied a convenience sample of adults with Type 2
diabetes, cared for in the community. Participants completed an
initial exploratory qualitative interview, followed by sequential
interviews using a laddered dependency grid technique, where each
participant rated constructs (aspects of living with diabetes) against
elements (people involved in the participants living with diabetes)
on a scale of 0 10. The data were analysed using a modified
principal components analysis techniqueand were plotted on
bi-plot diagrams to show the relationship strength, via angle of
association to the element and vector length from the
construct. These calculations were then triangulated against the
individual rating comments given by the participants.
Results: Nineteen participants completed the sequential
interviews and laddered dependency grids (n = 48), from a
sample of 23 who completed the initial interview. The major
construct of interdependence identified had strong or weak
association with diabetes related constructs, depending on how
2014 The Authors.

each individual participant had rated them, and the explanation

they gave of the ratings.
Summary: The findings suggest that in this sample people with
Type 2 diabetes do not always consider primary care services as
appropriate to enable good self-management and achievement of
their life goals. Each individual had developed their own level of
interdependence with some of the participants being self-reliant or
using family members/friends, more than professional health
services, for managing their diabetes. Recommendations include
further exploration of interdependence in Type 2 diabetes.
P262
Adoption of the Piecing Together Diabetes
education tool by NHS Trusts
R Herring, C Pengilley, B Tuthil, H Hopkins and D Russell-Jones
Centre for Endocrinology, Diabetes and Research, Royal Surrey County
Hospital, Guildford, UK
Objectives: The prevalence of diabetes amongst the inpatient

population is increasing. Patients are frequently admitted under
general medical and surgical specialities. Healthcare professionals
often lack the confidence and knowledge to appropriately manage
this patient group. Formalised teaching sessions are very difficult as
clinical commitments always take priority and attendance is
further confounded by shift patterns.
Methods: The admission to discharge Piecing Together
Diabetes programme was developed. It is a new education tool
which has a jigsaw design and consists of 10 laminated module
jigsaw pieces, a facilitator guide, activity cards for group activities,
and a box for storage and transportation. Diabetes specialists
working in NHS hospitals gained experience and were provided
with the education tool through a series of national and regional
training days.
Results: Six national workshops have been held throughout the
UK with 124 diabetes specialists receiving their Piecing
Together Diabetes education toolkit. A number of regional
training days have also been held to assist diabetes teams in
implementing the education programme. To date, we have
evidence that 13 NHS hospitals have fully implemented the
education programme, with 167 healthcare professionals.
This group consisted of 39 healthcare assistants, 105 trained
nurses, 22 junior doctors and one pharmacist, all of whom have
received education using the Piecing Together Diabetes education
framework. The most popular modules are those related to insulin
and insulin errors, and hypoglycaemia.
Conclusion: The education tool has been widely adopted by NHS
Trusts providing specialist healthcare professionals with the
resources to effectively educate ward staff in a flexible, relevant
and interactive way with the aim of improving diabetes inpatient
care.
DIABETICMedicine
Methods: A comparative study of individuals with Type 1

diabetes (n = 9) using basal bolus analogue insulin regimens was
undertaken. Glycaemic control before and after two 40min runs at
70% VO2 max, in both laboratory and real-life environments, was
measured across 10 points during and up to 12h after exercise.
Insulin was adjusted in all participants according to a
self-management algorithm.
Results: Pooled mean glucose concentrations at each time point
were compared. There was no statistically significant difference [F
(1, 8) = 1.489, p = 0.257] in overall mean glucose concentrations
between environments. Similarly, the exercise environment or time
point of measurement had no statistically significant effect on mean
glucose concentration [F(9, 72) = 0.499, p = 0.871]. However,
episodes of both hypoglycaemia (defined as <4.0mmol/l) (5 vs 1)
and hyperglycaemia (defined as >9.0mmol/l) (25 vs 19) during
exercise in the laboratory environment were more frequent than
during real-life exercise, with acceptable concentrations (defined as
4.0 9.0mmol/l) (24 vs 34). In the 8 12h after exercise,
hypoglycaemia (3 vs 8) and hyperglycaemia (24 vs 14), with
acceptable concentrations (29 vs 33).
Conclusions: The exercise environment does not appear to affect
overall mean blood glucose concentrations. However, it may affect
the timing and frequency of hypoglycaemia and hyperglycaemia.
P264 Abstract withdrawn

P265
Usability of a truly non-invasive glucose
monitor in home use
A Gal1, I Harman-Boehm2, A Drexler3, E Naidis1, Y Mayzel1,
N Goldstein1 and K Horman1
1
Research, Integrity Applications Ltd, Ashkelon, Israel, 2Internal Medicine and
the Diabetes Unit, Soroka University Medical Center, Beer-Sheva, Israel,
3
Division of Endocrinology, Diabetes and Hypertension, David Geffen School
of Medicine, University of California, Los Angeles, USA
Aim: To determine whether results from laboratory based

research examining glycaemic control during and after exercise
can be applied to a real-life (non-laboratory) environment.
Glucose monitoring adherence is considered essential for achieving

tight glycaemic control in diabetic patients. Non-invasive (NI)
glucose monitoring is expected to encourage frequent selfmonitoring by overcoming pain and complexity involved in
invasive measurements. To motivate contributive utilisation, an
NI device should be user friendly and simple to manage at home
and in a home-like environment. GlucoTrack is an NI, CE
marked, glucose monitoring device. GlucoTrack suitability for
home use was tested by 50 educated subjects (high-school and
higher) based on device accuracy and user feedback analyses. The
first trial day included individual calibration and brief training by a
proficient team. Forty-two subjects conducted the measurements
by themselves for three more days. Eight more participants used
the device at home for 5 7 days after calibration. Clarke error grid
analysis shows 96.2% of the points in the clinically accepted A+B
zones. A mean absolute relative difference of 30.5% was observed.
82% of all subjects expressed willingness to use the device
regularly. 78% were generally pleased with it. The GlucoTrack
display appeared clear and understandable to 89% of the
participants. The operating instructions were clear to 81% of the
high-school-educated and to 84% of the higher-educated
participants. Among high-school-educated and higher-educated,
respectively, 64% and 61% claimed the device is easy to use.
GlucoTrack yields fair accuracy and is user friendly regardless of
education level. These advantages, along with its painless nature of
measuring and competitive long-term cost of use, suggest
GlucoTrack as autilisable device for enhanced blood glucose
monitoring and tighter glycaemic control.
2014 The Authors.

105
P263
Can laboratory based research regarding
Type 1 diabetes and exercise be applied in
the real-life environment?
J Charlton1,2, L Kilbride2, R MacLean2, M Darlison2 and
J McKnight1
1
Metabolic Unit, NHS Lothian, Edinburgh, UK, 2Social Science, Edinburgh

Napier University, Edinburgh, UK
DIABETICMedicine
P266
Why are attendance rates at structured
education so low in inner city areas of
Central London? A qualitative analysis
Conclusions: The data show that the weight management

programme run by a dietetic assistant provides a cost saving to
the NHS without compromising on results. Weight loss achieved
through the programme is comparable to other studies into group
weight loss programmes.
M Jones1, J Pam2, A Devereux2 and D Gable3

1
Diabetes, Central London Community Healthcare NHS Trust (CLCH), London,
UK, 2Service Improvement, Diabetes UK, London, UK, 3Diabetes and
Endocrinology, Imperial College Healthcare NHS Trust, London, UK
Aims: In the last year attendance rates for structured education

(SE) have been 56/459 (12%) and 193/419 (46%) of those invited
in two separate inner London boroughs [249/878 (28%)
overall]. We wished to identify why attendance rates were so low.
Methods: Sixty people who did not attend the local X-PERT SE
after being invited were invited to a focus group. Initial invite was
by letter, followed by telephone contact. Participants received a
10 shopping voucher. Identical questions were asked in each focus
group.
Results: Sixteen patients attended two focus groups, one (n = 8) for
each borough. Attendance rates were 7/8 (88%) and 5/8 (63%) for
the groups (75% overall). Ethnicity was three (25%) white British,
three (25%) Afro-Caribbean, one (9%) Asian, five (41%) black
African. Themes identified were unable to commit to 6 weeks,
feeling venue and timings were unsuitable. A number did not see the
value of SE, reporting care they received at their GP practice was
sufficient. However, when X-PERT was described to them most felt
this was something important that would help them. Reward was
described as not important in encouraging attendance.
Conclusion: More needs to be done to help patients choose to
attend SE. Finding the right time and venue remains a
challenge. However, attendance rates would be higher if, at
point of referral, patients understood the purpose of SE and how
it fitted into their care. Individual care planning may go some way
to achieve this.
P267
Diabetes weight management programme
delivered by a dietetic assistant: the success
of weight reduction and the cost saving
R Fitz-Patrick and J Neale
Department of Nutrition and Dietetics, University Hospital Southampton NHS
Foundation Trust, Southampton, UK
Introduction: Southampton city delivers an 8 week weight

management programme led by a dietetic assistant, enabling
patients with or at risk of developing Type 2 diabetes to lose
weight and make positive lifestyle changes.
Methods: Patients with a body mass index >28kg/m2 with or at
risk of developing Type 2 diabetes were referred by their GP or
healthcare professional. Each programme consisted of a 2h session
held over 8 consecutive weeks. Each week patients were weighed
and asked to set an individual goal. The course actively signposted
to local support services and had input from a diabetes specialist
nurse on week 1 and a local sports centre on week 4. Patients were
followed up at 3, 6 and 12 months after the programme.
Results: A total of 51 patients completed the programme between
2011 and 2012. On average patients lost 2.08kg 2.75kg during
the programme (p < 0.0001). 36% of patients attended the 12
month review where the average weight loss had increased to
5.98kg 11.21kg (p = 0.027). Employing a dietetic assistant for
3.75h and a specialist dietitian for 3.75h a week produced a cost
saving of 1,500 per annum compared with solely employing a
specialist dietitian for 7.5h.
106
P268
Insight and engagement with patients and
the public to facilitate pathway
development: the Merseyside Impaired
Glucose Regulation (IGR) Pathway, aiming
to prevent or delay Type 2 diabetes in
accordance with NICE public health
guidelines 38
RA Du Plessis1, E Page2, A James2 and S McNulty3
1
Public Health, Sefton Council, Sefton, UK, 2Public Health, Liverpool Council,
Liverpool, UK, 3Public Health, Knowsley Council, Knowsley, UK
Aims: To utilise insight and engagement to maximise access and

to facilitate the uptake of services within the Merseyside IGR
pathway including primary care, patient education and lifestyle
intervention.
Methods: In October 2012 a sample of people aged 40 and over
was recruited. The group contained an equal split of male and
female, who were overweight and had a range of other risk factors.
Stage 1 was individual and paired interviews to explore awareness,
perceptions and attitudes towards pre-diabetes, diabetes, health
and lifestyle. Stage 2 was focus groups to explore the pathway,
identifying how it could be delivered to best meet needs and
encourage behaviour change.
Results: Sixty-five people were recruited from 11 different
localities in Merseyside. Healthy foundations segmentation
identified almost half as unconfident fatalists and a third as
live for todays. Language: participants did not understand
pre-diabetes or IGR. Borderline diabetes was the preferred
term as it indicates you can do something about it; high risk of
diabetes was also a useful term. Motivation: being identified as
borderline diabetes is a motivator for making lifestyle changes.
Access: all would access support and regular review from their GP
practice. Education: participants expressed reluctance to attend
education, preferring coaching and ongoing support.
Conclusions: The learning gained through this insight has
resulted in significant changes to the proposed pathway. It is
expected that people will access the services as outlined in the
pathway and this will motivate them to make changes to their
lifestyle. The Merseyside IGR pathway will be formally launched
in November 2013.
P269
Care planning advocates: supporting quality
care planning
J George and S Lucas
Diabetes Modernisation Initiative (DMI), Guys and St Thomas Charity,
London, UK
Aim: To use patient advocates to ensure delivery of high quality

care planning.
Method: A call to action was circulated to recruit people to act as
care planning advocates. Those interested were invited to a
day-long training workshop. This involved a comprehensive
selection process and group training programme to learn more
about care planning and to understand expectations in becoming
2014 The Authors.

an advocate. The successful advocates attended local Diabetes Eye

Complication Screening (DECS) clinics to collect patient
experience of care planning. Here, patients were asked a series of
questions to determine what they understood by care planning,
whether they were receiving care planning and to identify whether
care planning was associated with increased confidence in
managing their diabetes.
Result: Twenty-two patients expressed an interest in becoming
care planning advocates. Following the selection and training
workshop, 10 people were trained. 240 patients have been
interviewed whilst attending their DECS appointment. The
feedback collected represents 67% of GP practices in Lambeth
and Southwark. 28% of patients interviewed are receiving the
minimum standards of care planning and report significantly
greater confidence in managing their condition than those not care
planning
Conclusions: Care planning is a difficult concept to understand
and can often be a tick box exercise. Using advocates to collect
patient experience of care planning is an innovative way to ensure
it is a quality process. It facilitates knowledge transfer between
advocate and patient, so while experience is being collected the
patient is also being informed about care planning, reinforcing it as
an integral part of their care. Furthermore, the information is fed
back to general practice as an ongoing way to assess and embed
quality care planning.
P270
Is the insulin passport taking off?
AR Aiyengar1 and MA Saeed2,3
1
College of Medical and Dental Sciences, University of Birmingham,
Birmingham, UK, 2Department of Diabetes, University Hospitals Birmingham,
Birmingham, UK, 3School of Clinical and Experimental Medicine, University of
Aims: The National Patient Safety Agency (NPSA) issued an alert

in 2011 with the aim to empower people with diabetes to take an
active role in their safety after a review of 16,600 incidents
involving insulin. Insulin passports(IPs) along with patient
information booklets (PIBs) were recommended. Much smaller,
credit-card-sized insulin safety cards (ISCs) have also been agreed
as suitable alternatives. This audit aimed to assess the use of IPs
that were mainly provided by primary care (our unit provides ISCs
to new insulin starters).
Methods: Characteristics of IP use were studied through
distribution of a 28-point anonymous questionnaire to patients
attending diabetes clinics in a large teaching hospital.
Results: Out of 171 people approached,124took insulin and 40%
(n = 50) had an IP. There was no major difference between
availability of passports to English and non-English speaking
patients (40% vs 46%, respectively). Only 54% (n = 27) received
PIBs with their IPs. Two-thirds regularly carry their IPs and only
12% (n = 6) found it inconvenient to carry. 86% (n = 43) never
voluntarily present their IPs in primary/secondary care
consultations. A third reported that the recorded therapy in their
IP was not uptodate.
Conclusions: Ideally, patients should play an active role towards
their safety by carrying and presenting their IPs at consultations
across all healthcare sectors. One year after the introduction of IPs,
trends indicate that patients use of this document could be
optimised by awareness raised by healthcare professionals and
diabetes charities in regard to its vital role in patient safety.
2014 The Authors.

DIABETICMedicine
P271
Community-based Type 1 diabetes
education reduces HbA1c, total insulin
usage and insulin treatment costs
S Marden, J Murphy and KE Fayers
West Hampshire Community Diabetes Service, Southern Health NHS
Introduction: The West Hampshire Community Diabetes Service

delivers the Southern Health Intensive Insulin Education (SHINE)
programme to patients with Type 1 diabetes (4h per week over 4
weeks). Insulin dosage at baseline and course completion was
available for 123 patients. We determined how insulin usage
changes post course and the potential cost savings.
Method: The 123 patients with Type 1 diabetes (aged 46 15
years, 44% male, duration of diabetes 20 13 years) had a record
of total daily analogue insulin dose before and after education.
HbA1c measurements were available on 44 (36%) patients (aged
51 15years, 36% male, duration of diabetes 24 14years).
Results: In all, 78% of patients (n = 96) reduced their total daily
analogue insulin usage by week 4 of education. 14% (n = 17)
increased insulin usage whilst 8% (n = 10) recorded no change.
Across all subjects there was an average 11% reduction in daily
insulin usage (57 to 51 units). From the 44 (36%) patients with
HbA1c recorded at 6 months, there was an average 4mmol/mol
(0.4%) reduction in HbA1c; 36 (82%) of these subjects had
reduced their insulin usage and achieved the same average
reduction in HbA1c.
Conclusion: A community-based intensive diabetes education
course can reduce total daily insulin dose and HbA1c. This
reduction in insulin represents an average cost saving of 0.14 per
day and 52.56 per patient per year (based on an average 0.024
per unit insulin).
P272
Education of patients with diabetic foot
disease: is it effective?
E Cerrone, RA McDonnell, D Selvarajah, FM Creagh and
RA Gandhi
Department of Diabetes, Sheffield Teaching Hospitals NHS Foundation Trust,
Sheffield, UK
Background and aims: Patients attending multidisciplinary

diabetes foot clinics are given advice as part of their treatment,
seeking to improve foot outcomes and prevent re-ulceration. The
aim of the study was to see whether the advice given to patients
who had previously ulcerated had improved their footcare
knowledge and behaviour.
Methods: The Nottingham Assessment of Functional Footcare
(NAFF) questionnaire was used to assess footcare behaviours in
consecutive patients attending a diabetes foot clinic with new
ulceration. Two groups were assessed: those who had experienced
previous ulcers, and those presenting with an ulcer for the first
time.
Results: To date, data have been collected in 49 patients (19 first
ulcers and 30 with previous ulcers). The data showed no significant
difference in the NAFF scores between the two groups (mean
score SD for first ulcer group 54.3 8.7 compared with
57.0 6.7 for previous ulcer group, p = 0.22). Only 67% of
recurrent ulcer patients could recall getting advice on good
footcare; these patients did have significantly higher NAFF scores
(58.8 6.4 vs 53.1 6.5, p = 0.04). Of note, it was found that
107
DIABETICMedicine
85% of all patients had a low education level (classed as leaving

school aged 14 16 and with no further qualifications).
Conclusions: This study suggests that patients re-presenting with
a foot ulcer do not demonstrate any better knowledge of footcare
behaviour than those presenting for the first time. Further research
is required to determine the impact of patient education levels and
whether different strategies to deliver footcare advice can improve
footcare knowledge and behaviour.
P273
Barriers to self-care and diabetes
complications among patients with Type 2
diabetes in rural England
D Simmons1, XW Mei1 and J Graffy2
1
Wolfson Diabetes and Endocrinology Clinic, Cambridge University Hospitals,
Cambridge, UK, 2Primary Care Unit, Department of Public Health and Primary
Care, University of Cambridge, Cambridge, UK
Aims: Diabetes related complications are often associated with

inadequate self-care. We have investigated whether self-reported
diabetes complications are associated with particular barriers to
care.
Methods: Individuals (n = 21,961) with practice defined Type 2
diabetes from 62 general practices in rural England were posted
a barriers-to-diabetes-care survey (with stamped addressed
envelope). Two reminders were sent. A small number of
participants were recruited through community-based posters
and mail-outs to hospital clinic patients/Diabetes UK
members. Of 3,932 (17.9%) responses, 3,825 completed the
questionnaire and had Type 2 diabetes. The survey included
demographic/clinical questions and 33 close ended items covering
internal physical (IPH, e.g. comorbid conditions), external physical
(EPH, e.g. transport), educational (e.g. diabetes knowledge),
internal psychological (IPS, e.g. motivation) and external
psychological (EPS, e.g. family support) barriers. Diabetes related
complications were assessed with Have you any complications
from your diabetes? Covariate adjusted odds ratios compared
with those reporting no complications are shown.
Results: Complications were reported by 1,476 (38.6%): 697/746
with eye/foot complications respectively. Those with eye/foot
complications were more likely to report IPH [1.87(1.56 2.24)/
2.50(2.07 3.00)] and EPH [1.43(1.17 1.74)/1.28(1.04 1.56)]
barriers. Having any complication was more associated with IPS
barriers [1.42(1.07 1.90)]. Having more complications overall
was associated with more worry, fear or shame [1.59(1.24 2.05)].
Conversely, those willing to prioritise their diabetes were less likely
to report foot (but not eye) related complications [0.74
(0.56 0.98)].
Conclusions: Diabetes complications are associated with a
number of barriers to care. Although some of these may be a
consequence of having a diabetes complication, others are likely to
reflect patient issues where action might prevent long-term harm.
P274
How confident are people with diabetes in
managing their own health?
CAM Paddison, CS Saunders, GA Abel, RA Payne and
M Roland
Cambridge Centre for Health Services Research, University of Cambridge,
Cambridge, UK
Aims: Confidence in managing your own health is important for

successful self-management of diabetes. We explore levels of
108
confidence in managing health among people with diabetes in

England, and examine to what extent this varies by
socio-demographic characteristics, comorbidity, and quality of
communication with primary care clinicians.
Methods: Data from 71,488 patients with self-reported diabetes
in the 2011/12 General Practice Patient Survey (England) were
analysed using mixed effects linear regressions to investigate how
patients confidence in managing their health varies by
socio-demographic characteristics, number of comorbid
conditions, quality of life measured using five EQ5D domains,
and quality of communication with primary care clinicians.
Results: Confidence in managing your health among people with
diabetes was modest, with 12% of respondents not very or not at
all confident in managing their own health. Compared with
people of similar age, gender, ethnicity and deprivation, confidence
in managing your health among people with diabetes varied most
by the EQ5D domains of anxiety/depression, self-care and usual
activities (mean difference range 15.518.7 points on 0 100 scale,
p < 0.0001). Somewhat smaller variation was seen by age,
ethnicity and quality of doctor and nurse communication (mean
difference range 5.97.6, p < 0.0001).
Conclusions: One in 10 people with diabetes lack confidence in
managing their own health. Interventions to improve confidence
for self-management might provide support for managing
comorbidities (including pain, anxiety and depression), and focus
on older adults, patients of mixed ethnicity and those in socially
deprived areas.
P275
Effectiveness of carbohydrate counting and
insulin dose adjustment structured group
education programme for Type 2 diabetes
on basal bolus insulin regimen
I Shorrock1, J Hannah1, A Barnes1 and I Idris2
1
Diabetes and Endocrinology, Royal Derby Hospital, Derby, UK, 2Division of

Medical Sciences and Graduate Entry Medicine, Royal Derby Hospital, Derby, UK
Background: The rationale for carbohydrate counting and insulin

dose adjustment in patients with Type 2 diabetes remains a matter
of debate. This audit was conducted to determine the effectiveness
of such intervention, delivered via a structured education
programme, on glycaemic and weight control.
Methods: Data from 123 patients with Type 2 diabetes on a basal
bolus insulin regimen who participated in the intensive educational
intervention programme were analysed. The intervention was
conducted over five sessions in groups of 8 12 patients and
delivered by a dietitian and a diabetes specialist nurse. Over the five
sessions,patients were taught carbohydrate counting and dose
adjustment according to a curriculum and lesson plans. Complete
available data for HbA1c (n = 103), weight (n = 61) and insulin
doses (n = 35) were assessed at baseline and at the end of the
intervention. Mean duration of follow-up was 6.3 months.
Results: The overall intervention effect was not statistically
significant for changes in HbA1c and weight: 9.5% vs 9.6%,
p = 0.8, and 97.2kg vs 94.5kg, p = 0.4, respectively. Daily insulin
requirement,however, was significantly reduced when baseline was
compared with the end of the intervention (77.5U vs 55.1U,
p < 0.01).
Conclusions: An intensive structured education programme for
carbohydrate counting delivered in a group setting was associated
with a significant reduction in insulin dosesbut no significant effect
on HbA1c and weight. Additional research is needed to translate
2014 The Authors.

nutrition education into improved health outcomes and cost

effectiveness.
P276
The effect on glycaemic control of sharing
personalised clinical information with
people with Type 2 diabetes prior to their
consultation: an exploratory study using a
randomised controlled design
M ODonnell1, J Newell2, BE McGuire3 and SF Dinneen1,4
1
School of Medicine, National University of Ireland, Galway, Ireland, 2Health
Research Board Clinical Research Facility, National University of Ireland,
Galway, Ireland, 3School of Psychology, National University of Ireland,
Galway, Ireland, 4Department of Diabetes and Endocrinology, Galway
University Hospitals, Galway, Ireland
Aim: To measure the effect of sharing personalised clinical

information (e.g. HbA1c, lipids) with people with Type 2
diabetes prior to their consultation on glycaemic control post
intervention.
Methods: The 136 participants were randomised to one of three
groups. The intervention group (n = 45) received a booklet that
included individual medical results and targets; a general
information group (n = 44) received a similar booklet but it did
not include individual medical results; and a control group (n = 47)
received no written information prior to their consultation.
Statistical analysis was carried out by intention to treat.
Results: Post intervention HbA1c results were measured between
2 and 11 months (mean6 months) after the consultation. 130 cases
were included in the analysis. Mean HbA1c results for the control,
general information and intervention groups were respectively 61.3
(SD 15.8), 62.6 (SD 14.2) and 59.7 (SD 14.3)mmol/mol at baseline
and 59.8 (SD 14.5), 62.1(SD 13.7) and 62.8 (SD 18.6)mmol/mol
post intervention. Mean difference in HbA1c for the control,
general information and intervention groups was respectively
0.50 (SD 13.5), 0.17 (SD 9.2) and 2.91 (SD 10.7)mmol/mol.
An analysis of the change in HbA1c levels (mean 0.869mmol/mol,
SD 11.32) using an analysis of covariance, adjusting for baseline
and patient characteristics, found no evidence of a significant
difference between the three groups (p = 0.64) over time.
Conclusions: Sharing personalised clinical information with
people with Type 2 diabetes prior to their hospital outpatient
consultation does not improve glycaemic control in the short term.
DIABETICMedicine
eight specified questions: insulin not written up, name of insulin

incorrect, dose unclear, unit abbreviated to u, insulin or
prescription chart not signed, insulin given/prescribed at wrong
time. We encouraged staff to DATIX all errors. We then had a
large campaign which involved e-learning, posters, ID cards and
face/face training to promote the safer use of insulin to all
healthcare professions who prepare, prescribe or administer
insulin.
Results: The 2011 data showed that within the 500 bed acute
hospital of those treated with insulin there was an error rate of
26.8%; then, following the campaign, we carried out a further
audit and found that this was reduced to 14.3%. To date,
these results have highlighted that the main issues are still
units being abbreviated and prescriptions not signed for.
Conclusion: Although our audit figures have shown through hard
work and education that we have improved the number of insulin
errors occurring within the hospital, there are still too many. We
need to refocus on training for prescribers and further education
for staff.
P278
Acceptability of using trained lay people to
deliver diabetes self-management
education: a qualitative investigation using
data from patients and educators
PK Mandalia1, M Stone1, ME Carey1, H Daly1, R Hale2,
L Martin-Stacey1, N Taub3, K Khunti4, LJ Gray5, TC Skinner6,
S Heller7 and MJ Davies4
1
Diabetes Research Centre, University Hospitals of Leicester, Leicester, UK,

Fareham Voluntary Group, Diabetes UK, Fareham, UK, 3Department of
Health Sciences, University of Leicester, Leicester, UK, 4Diabetes Research
Centre, College of Medicine, Biological Sciences and Psychology, University of
Leicester, Leicester, UK, 5Department of Health Sciences and Diabetes
Research Centre, University of Leicester, Leicester, UK, 6School of
Psychological and Clinical Sciences, Charles Darwin University, Darwin,
Australia, 7Academic Unit of Diabetes, Endocrinology and Metabolism,
University of Sheffield, Sheffield, UK
2
Objectives: Following national guidance from the National Patient

Safety Agency (2010) for safer use of insulin which found that
between 2004 and 2009 for England and Wales there were nearly
4,000 insulin errors, we decided to audit our own data. Within our
acute hospital, we found that in 2011 our insulin errors were
26.8%. The diabetes specialist team was tasked with reducing this
number of insulin errors between 2011 and 2013.
Method: Baseline data were collected from the NaDIA audit with
focus on insulin prescription and management errors, using the
Aims: Quantitative findings from the DESMOND Lay Educator

Trial have indicated that patients obtain equivalent benefits from
self-management education delivered by one trained lay educator
and one healthcare professional educator compared with sessions
delivered by two healthcare professional educators. This sub-study
aimed to explore, qualitatively, the use of lay educators,
particularly in terms of perceived acceptability.
Methods: Prior to attending their allocated session, patients were
blinded to whether their two educators included one lay person.
Semi-structured interviews, facilitated by a topic guide, were
conducted with a purposive sample of 16 patients who attended a
session delivered by a lay/healthcare professional educator team
and 11 educators (lay and healthcare professional). Data were
analysed thematically using a constant comparative approach
including framework charting.
Results: Analysis suggested that, overall, using lay people to
deliver DESMOND education was acceptable to patients and
educators. Professional and lay educators generally described good
working relationships and enjoyable experiences. A key theme
emerged regarding educators levels of diabetes knowledge;
however, patients were positive about involving lay educators
regardless of perceptions about knowledge levels and whether or
not they became aware of the composition of their educator team.
Peer status and/or empathy were cited as beneficial characteristics
of those lay educators with diabetes themselves or in a family
member.
2014 The Authors.

109
P277
Hospital insulin prescription and
management errors: are we improving the
care to our patients and knowledge to our
staff?
LS Mills and JE Stamper
Diabetes Department, Warrington and Halton Hospitals NHS Foundation
Trust, Warrington, UK
DIABETICMedicine
Conclusions: Qualitative findings have supported and helped to

explain the positive quantitative trial results. Confirmation of
acceptability is likely to influence the feasibility of rollingout the
use of lay educators to deliver self-management education and thus
increase capacity.
P279
Development of video-based behavioural
intervention to optimise self-management
in children with poorly controlled Type 1
diabetes
CM Webster1, AC Greene2, SA Greene1,2 and VA Alexander1
1
Paediatrics, NHS Tayside, Dundee, UK, 2Child Health, University of Dundee,
Dundee, UK
Aim: To develop a validated psychology-based behavioural

modification tool, video interaction guidance VIG, for use
within a busy paediatric diabetes clinic.
Method: VIG is an evidence-based and empirically grounded
technique used to support behaviour change and sustain long-term
relationships. VIG involves video feedback of positive
interactions to enhance effective communication between young
people and their healthcare professionals (HCPs). Acceptability of
VIG to patients, parents and HCPs was explored using patient
information leaflets and questionnaires. HCPs underwent VIG
training. The technological development involved establishing
efficient methods to video record consultations with access to
instant review and immediate feedback to patients without altering
the flow of the routineclinic.Caldicott approval was
granted. Three young people aged 13 18 with Type 1 diabetes
of duration >1 year and HbA1c>80mmol/mol were recruited and
took part in three VIG sessions each. Routine clinic appointments
were videoed. Appropriate video clips were selected to play back
to the subject to enhance positively their communication
skills. Validated self-efficacy questionnaires were used. Safe and
confidential storage of video material was ensured.
Results: HCPs and young people were willing to take part in
VIG. The technological side ran smoothly and the process was
completed within 30min per session. Through working in
partnership in identifying the positive aspects of the interaction
the young people were able to explore difficult aspects of managing
their diabetes.
Conclusions: We have demonstrated that it is possible to carry out
this novel VIG intervention to support behaviour change within the
current diabetes clinic with existing staff. The young people recruited
to the study have reported positive experiences of VIG to date.
Methods: Consultation and focus groups with children, young

people and their families identified the need to develop a tool that
would support a patient-led OP consultation. The care planning
tools were designed by the local family support group for
diabetes with input from the clinical team. Two forms were
designed, for children 14 and under and children over 14.
Twenty-six patients/parents/carers were audited in total. These
were randomly selected from patients attending the diabetic OP
clinic waiting areas at the three hospitals who are working
together to improve diabetes care. The audit was undertaken on
five different days and five OP clinics attended. A verbal
explanation was given prior to the patient completing the patient
care planning tool. Following completion and after consultation,
the patients/parents were interviewed about the usefulness of the
tool. A structured questionnaire and open ended interview
technique was used.
Results: In total, 90% of patients surveyed found the draft tool to
be clear, useful and age appropriate. Overall, more than 80% liked
the idea of the tool.
Conclusions: These results suggest that the patient care planning
tool is likely to be useful and well received. For the large majority
of patients, its systematic introduction in OP clinics is likely to help
improve patient experience and reinforce patient empowerment. It
provides further evidence of the benefits of continuing to work
closely with patients and parents as partners in planning and
decision making about care.
P281
Can improvements in newly diagnosed Type
2 diabetes patients reported outcomes,
following attendance at a 2h structured
education programme, be replicated as new
educators are trained?
E Jennnings1, C OKane1, D Steele1 and J Green2
1
Nutrition and Dietetics, Betsi Cadwaladr University Health Board, Wrexham,
UK, 2Medical Physics, Betsi Cadwaladr University Health Board, Rhyl, UK
Objectives: To examine the usefulness of a patient care planning

tool in supporting a more patient-led outpatient (OP) consultation.
Aim: To investigate the effectiveness of initial and ongoing

training and peer review of dietetic educators to deliver a single,
structured 2h education programme to patients recently diagnosed
Background: A group education session was rolled out across
the health board serving the population of North Wales, Betsi
Cadwaladr University Health Board, between 2009 and 2013. In
2011,
patient
reported
outcomes
showed
significant
improvement, with no difference between four educators. From
2011, bespoke in-house training of 10 educators was piloted
including planned observation, feedback and a peer-review based
support process.
Method: This was a retrospective analysis of patient evaluations
(n = 198), pre-to-post attendance at a group learning session. The
evaluation featured 10-point analogue scales of patient confidence,
managing diabetes, learning in a group, and overall satisfaction for
eight different educators during 2012 2013. Differences in
confidence were assessed by Wilcoxon signed ranks and paired
t-test. The relationship between educator and the patient-reported
confidence was analysed using a Pearson chi-squared test.
Results: A statistically significant difference was observed
between before and after scores for confidence learning in a
group and in managing diabetes (p < 0.0005 in both cases). There
was no statistically significant association between the educator
and the improvement in confidence learning in a group (p = 0.51)
or the improvement in confidence managing diabetes (p = 0.17).
110
2014 The Authors.

P280
Developing a culture of shared decision
making between healthcare professionals
and families through collaborative care
planning: patient care planning tool to
support a more patient-led paediatric
outpatient consultation
J Cross1, L Varlow2 and L Briant3
1
Young Diabetes Connections, Evelina Childrens Hospital at Guys and St
Thomas NHS Foundation Trust, London, UK, 2Family Support Group,
Diabetes UK Lambeth Lewisham and Southwark, London, UK, 3Children and
Young People, Diabetes Modernisation Initiative, London, UK
Overall satisfaction with the group session was 8.8/10 with no

significant difference between educators (p = 0.21).
Conclusion: Improvement in patient-reported outcome is
maintained. Individual educator performance may be used to identify
strengths, to enhance peer support and improve educator training.
P282
Food for thought: do young adults with
diabetes make the right choices?
EV Iddles1, A McKillop-Smith2, JS Lemon3, R Keenan1 and
WA Watson2
1
School of Medicine and Dentistry, University of Aberdeen, Aberdeen, UK,

Department of Diabetes, JJR Macleod Care of Diabetes Centre, Aberdeen
Royal Infirmary, Aberdeen, UK, 3Directorate of Information Technology,
University of Aberdeen, Aberdeen, UK
2
Background: Scotland has an increasing prevalence of

diabetes. Last year, over 9,000 young people were affected by
Type 1 and Type 2 diabetes. During the transition from
adolescence to young adulthood there has been a reported
increase of diabetes complications thought to be due to the
omission of insulin and poor dietary choices. Education and
self-management play an important role in order to reduce both
acute and long-term complications.
Aims: To explore the dietary habits, alcohol intake and
engagement with diabetes services in young people aged 16 30
using a questionnaire.
Methods: The questionnaire was developed using Snap Survey
software with input from diabetes specialists, patients and
food and nutrition researchers. Questions examined baseline
characteristics, diet choice, educational opportunities and
diabetes services. Questionnaires were validated and paper copies
were distributed and answered anonymously in diabetes clinics
within NHS Grampian.
Results: Currently, of the 57 respondents 53% were female, 96%
had Type 1 diabetes and 12% had associated autoimmune
conditions. Many young people (82%) admitted not eating the
recommended five portions of fruit and vegetables per day and
12% were binge drinkers. The DAFNE programme had been used
by 18%. Respondents (46%) HbA1c values ranged from 39 to
114mmol/mol.
Summary: Initial results suggest an unhealthy diet and poor
lifestyle choices are common amongst young people. Further
analysis will investigate the relationship between diabetes
education, management and dietary choices. Provision of
diabetes services for young people and future public health
implications will be considered.
P283
Reaching and teaching: community based
use of bolus advisor meters improves
glycaemic control in patients with both
Type 1 and Type 2 diabetes
DIABETICMedicine
Methods: We identified two groups: carbohydrate-counting

post-Type 1 education; Type 1 plus Type 2 on fixed doses of
bolus insulin. Group 1: A 2h bolus advisor clinic was established at
the end of the Type 1 education course (week 4). Patients were
offered either Accuchek Expert (Roche) or Insulinx (Abbot) meters
the week before initiation. Follow-up: 3 and 6 months. Group 2: A
2h Insulinx Easy Mode clinic was established for those on fixed
dose regimens. Follow-up: telephone call in 1 week, with email or
face-to-face follow-up according to need.
Results: Group 1: Accuchek Expert 47 patients, Insulinx
Advanced 23
patients (n = 70,
average
HbA1c
at
baseline 77.3mmol/mol; 3 6 months n = 28, average HbA1c
68.1mmol/mol; 1 year n = 21, average HbA1c 66.4mmol/mol).
Group 2: Insulinx Easy Mode, 75 patients (n = 75, average HbA1c
at baseline 78.4mmol/mol; 3 6 months n = 37, average HbA1c
77mmol/mol; 1 year n = 34, average HbA1c 63mmol/mol). HbA1c
reduced by 14% (Group 1) and 19.6% (Group 2) at 12 months.
Conclusion: Bolus Advisor meters can be used successfully in a
community setting for both Type 1 and Type 2 patients and do not
need to be reserved for secondary care initiation.
P284
Assessment of quality of life among
patients with Type 2 diabetes
MA Aboawja, AF Alamri and T Shaqran
Family Medicine and Diabetes Clinics, North West Armed Forces Hospital,
Tabuk, Kingdom of Saudi Arabia
Objectives: To assess the quality of life and determine clinical and

soci-demographic factors affecting it among patients with Type 2
diabetes attending family medicine clinics and diabetic clinics in
North West Armed Forces Hospital in Tabuk, Saudi Arabia.
Methodology: A cross-sectional study was conducted among
adult patients with Type 2diabetes. Participants in this sample were
asked to fill out the health status questionnaire on a single occasion
while attending the family medicine centre. The self-administered
questionnaire consisted of several demographic and clinical
variables, as well as the core items from the SF-36.
Results: The study included 347 patients with Type 2 diabetes.
Their mean age was 49.3 9.2 years. More than half of them
(58.8%) were males. The highest reported mean scores of domains of
quality of life among patients with Type 2 diabetes were social
functioning (70.4) and mental health (70.4), followed by physical
functioning (69.4), while the lowest mean score was that of role
limitation due to physical health problems (43.1). Females with
lowincome (1,000 EUR/month), complicated, hospitalised, used
insulin and with comorbidity patients as well as those with poor
glycaemic control were more likely to report significantly lower
quality of life score than others.
Conclusion: This study has shown that Type 2diabetes negatively
affects quality of life in the Saudi adult population, particularly in
relation to their energy and vitality and their role limitation due to
physical health problems.
J Murphy, KE Fayers and C Atkinson

Aims: The launch of our community diabetes service has improved

access to Type 1 and Type 2 diabetes education, reducing
inequality. In the UK, initiation of bolus advisor meters is typically
restricted to hospital based intensive insulin clinics. We report the
successful use of these sophisticated meters in the community setting.
2014 The Authors.

111
DIABETICMedicine
P285
Transformation of diabetes care in Kuwait
through postgraduate healthcare
professional education
DJ Wake1, F Sukkar2, C Goddard1, N Halawa2, R Barake2,
S Chisholm1, A Morris1, K Behbehani2, J Ker1 and M Scott1
1
School of Medicine, University of Dundee, Dundee, UK, 2Dasman Diabetes
Centre, Kuwait City, Kuwait
Introduction: The KuwaitScotlandand Health InnovationNetwork

was launched in 2010 as a partnership between the University of
Dundee, Aridhia Informatics, NHS Tayside, Dasman Diabetes
Institute, Kuwait Ministry of Health. Diabetes prevalence averages
around 20% in the Gulf region. The network aims to improve chronic
disease management in Kuwait, focusing on diabetes. The programme
delivers in four areas: informatics, clinical standards, research and
education. The education programme consists of professional
development events centred around a postgraduate (PG) programme.
Methods: Launched in 2011, PG Cert/Dip/MSc Diabetes Care
and Education has recruited from throughout Kuwait (media
advertising). The programme incorporates organisational
management, education theory and practice in addition to
clinical care. It utilises blended learning, face-to-face teaching,
simulation and virtual learning environment activities and support.
Assessment is through workplace projects, which must include
elements of quality improvement/healthcare transformation.
Results: Postgraduate programme; 2011 to October 2013: 288
applications; 164 students enrolled; student background doctors
(primary and secondary care), nurses, pharmacists, public health
workers, dietitians, laboratory workers, dentists; >600 attendees
at CPD events; >400 workplace projects completed in various
healthcare settings throughout Kuwait Project examples:
Development and evaluation of an evening retinal screening clinic;
Diabetes patient foot education
local implementation; Utilising
threaded forums to analytically reflect in a community of practice;
Impact of patient centred behaviour modification in primary care.
Conclusions: Healthcare professional education programmes
have the potential to directly and rapidly impact on the
management of diabetes through project based assignments. This
model may be transferable to other countries.
P286
Increasing the uptake of structured
education: myth busting and moving
beyond basic demographic segmentation to
greater understanding of needs in Lambeth
and Southwark
RH Dale, S Lucas and J George
Diabetes Modernisation Initiative, Guys and St Thomas NHS Foundation
Trust, London, UK
Aims: To understand the barriers preventing people with Type 2

diabetes taking up the offer of structured education.
Methods: The analysis included a review of 50 case notes, referral
and activity data from service providers and qualitative interviews
with patients who had declined the offer of structured education.
The work brought together literature reviews, focus groups and
built on work from local provider teams and local university
research departments.
Results: No identifiable difference between the population taking
up the offer of structured education and those not doing so was
found. Of the referrals reviewed, 55% were men, 45% female
112
(modal age 50 59). The ethnic breakdown of the population

choosing not to attend was 37% White, 40% Black or Black
British, 8% Asian or Asian British, 5% mixed, 6% unknown and
4% other, closely reflecting the breakdown of those referred to
diabetes community services.
Conclusions: To better understand barriers to taking up
structured education analysis must move beyond population and
demographic measures. Findings suggests non-attendance is
multifactored, with individuals readiness to change, health beliefs
and quality of referral playing a role in the decision to attend
structured education. Individualised conversations are needed to
assess the appropriate support for a person living with Type 2
diabetes. To increase uptake of structured education a system is
required that is more flexible. Locally, a new tiered model of
support is required that offers not only structured education on
diagnosis but reminders and self-referral at every opportunity as
well as specific interventions on development of a complication.
P287
Introducing an effective predictor of HbA1c
for research and clinical care: the DSMQ
(Diabetes Self-Management Questionnaire)
JW Huber1, G Hood2, A Schmitt3, M Fang1, M Callender1 and
N Hermanns3
1
Institute of Health and Wellbeing, University of Northampton, Northampton,
UK, 2North East London Diabetes Research Network, Queen Mary University of
London, London, UK, 3Research Institute of the Diabetes Academy
Mergentheim, Diabetes Academy Mergentheim, Bad Mergentheim, Germany
Background: Several questionnaires on self-care are available,

with the Summary of Diabetes Self-Care Activities (SDSCA)
probably most widely used. However, correlations with blood
glucose levels tend to be weak. In order to assess self-care activities
which predict glycaemic control, the 16-item Diabetes
Self-Management Questionnaire (DSMQ) was designed. We shall
report on findings in a German sample and the adaptation of the
questionnaire for use by primary care clinicians and researchers in
England.
Methods: The DSMQ was evaluated for its psychometric quality
in 261 hospital patients with Type 1 or 2 diabetes and its predictive
power was compared with the SDSCA. Ethical approval was
granted.
Results: The DSMQ items showed appropriate characteristics
(mean difficulty 47 25; mean item-totalcorrelation 0.46 0.12;
homogeneity 0.25; mean correlation with HbA1c 0.23 0.09).
Principal components analysis indicated a four-factor structure and
confirmed the designed scale structure. Internal consistency
(Cronbachs alpha) was good (0.84), consistencies of the
subscales were acceptable (0.60 0.77). The DSMQ scales
showed significant convergent correlations with their parallel
SDSCA scales (r ranging from 0.52 to 0.58) and HbA1c (r
ranging from 0.15 to 0.40). All correlations with HbA1c were
significantly stronger than those obtained with the SDSCA.
Conclusions: These findings suggest that the DSMQ is a reliable
and valid instrument and enables an efficient assessment of
self-care behaviours associated with glycaemic control. A further
study on its psychometric properties is currently under way using a
large English sample, supported by a grant from Novo Nordisk UK
Research Foundation.
2014 The Authors.

P288
Training practice nurses to care for people
with severe mental illness to reduce the risk
of cardiovascular disease and diabetes
S Hardy1 and JW Huber2
1
University College London Partners, London, UK, 2Institute of Health and
Wellbeing, University of Northampton, Northampton, UK
Background: People with severe mental illness (SMI) have poorer

physical health and die at a younger age than the general
population, due to a high risk of cardiovascular disease and a
threefold increased risk of developing diabetes. Present contractual
guidelines (British Medical Association, NHS Employers) require
glucose testing as part of the annual screening only in patients over
40 years of age. Practice nurses at present have little contact with
people with SMI but are skilled in providing care for people with
long-term physical conditions. We report on the evaluation of a
novel 4h course for practice nurses to provide appropriate
screening and lifestyle advice for this disadvantaged group.
Method: The course designed specifically for practice nurses was
offered on 12 occasions to practice nurses in the East Midlands
(average participation n = 9). Mental health nurses were also
invited to attend. The course covered the need for glucose testing as
part of the annual screening for all patients. Practice nurses
perceived knowledge and skills were measured pre- and
post-training using questionnaires.
Results: Perceived knowledge and skills improved significantly
(pre, mean 27.2, SD 6.9, SEM 0.85; post, mean 38.5, SD 4.7, SEM
0.57, p < 0.0005). Responses to open-ended questions and an
email questionnaire around 13 months after the training indicated
that a considerable number of nurses has improved their practice.
Summary: The short course about physical health in SMI has
made a difference to practice nurses self-perceived knowledge and
skills, and facilitated changes to daily practice.
P289
Are X-PERT centres meeting audit standards
for structured patient education?
TA Deakin, SR Seed, CA Finch, BA Armitage and
I Cummingham
Structured Education, X-PERT Health, Hebden Bridge, UK
Aim: X-PERT structured education has demonstrated improved

clinical, lifestyle and psychosocial outcomes. To ensure national
implementation replicates results from the randomised controlled
trial, and standards identified in NICE guidance, continuous audit
is conducted.
Methods: X-PERT educators collect and enter patients baseline,
6 month and annual results onto the X-PERT Audit Database.
Results: In all, 79 (100%) of X-PERT centres have submitted data
for 38,299 people with new or established diabetes. Ethnicity: 84%
white and 16% BAME groups. Time to access education from
diagnosis: 8% within 1 year, 58% within 5 years, 14% within 10
years and 20% more than 10 years. Audit standards have been met
with excellent attendance scores: 95.7% attend at least one session
and 81.9% four sessions or more; patient evaluation scores 95%;
empowerment scores increased by 22.9%; HbA1c reduction
6.7mmol/mol (95% CI 6.2, 7.2) at 6 months and 6mmol/mol
(95% CI 5.5, 6.5) at 1 year; weight loss 2.1kg (95% CI 1.4, 2.6) at
6 months, 3.7kg (95% CI 2.8, 4.1) at 1 year; 2.5cm (95% CI 1.6,
3.3) waist circumference reduction at 6 and 12 months; 1.6mmHg
2014 The Authors.

DIABETICMedicine
(95% CI 1.1, 2.1) reduction in systolic and 2.3mmHg (95% CI 2.0,

2.6) reduction in diastolic blood pressure at 1 year; 0.3mmol/l
reduction in total/LDL cholesterol and 0.2mmol/l reduction in
triglycerides at 1 year.
Conclusions: National implementation of the X-PERT
programme has met audit standards. X-PERT is well attended
and evaluated and results in improved clinical and empowerment
outcomes amongst people with newly diagnosed and existing
diabetes.
P290
Are X-PERT centres meeting the NICE
structured patient education key criteria for
quality assurance?
SR Seed, TA Deakin, CA Finch, BA Armitage and
I Cummingham
Structured Education, X-PERT Health, Hebden Bridge, UK
Aim: To comply with NICE guidance, X-PERT educators are

assessed with the X-PERT Quality Assurance Programme within
their first year of delivery.
Method: The X-PERT Quality Assurance Programme contains 62
standards that are linked to learning outcomes. Each standard is
eitherfully, mostly, partly or not met. Guidance defines the criteria
required for each category and points are allocated accordingly.
Delivery to patients is assessed via direct observation or videos.
Educators receive a report and constructive feedback. If educators
score below 198 (80%) a timed action plan is prepared and the
quality assurance process is repeated within 1 year.
Results: A total of 454 X-PERT diabetes educators were eligible
to be assessed by 31 August 2013. The number of educators who
have completed the quality assurance process is 187 (41%); are in
progress 111 (24%); have not done it 148 (33%); are on long-term
leave 8 (2%). Mean score 221/248 (89%, SD 6.7%) with scores
ranging from 167 (67%) to 248 (100%). Six educators (3.2%)
required a timed action plan to increase score. Lost marks were
mainly due to lack of interpersonal/facilitation skills to support
group education and goal setting.
Conclusion: Adherence to quality standards for delivery of
X-PERT diabetes ensures that patients receive a high standard of
structured education. Two-thirds (65%) of X-PERT educators
have either been assessed or are currently going through the
process. The remaining educators will be assessed during the
following 12 months or removed from registration. A handbook
has been prepared to assist educators in developing the relevant
interpersonal/facilitation skills.
P291
Mobile technology as a tool for patient
education and self-management in the
diabetic population
IJM Campbell1, SG Cunningham2, N Conway3 and DJ Wake4
1
Medical School, College of Medicine, Dentistry and Nursing, University of
Dundee, Dundee, UK, 2Clinical Technology Centre, University of Dundee,
Dundee, UK, 3Child Health, University of Dundee, Dundee, UK, 4Medical
Research Institute, University of Dundee, Dundee, UK
Objectives: This study aims to ascertain the desire for mobile

technology, namely Smartphone apps (SA), which can support
patient education, self-management and data sharing with the
healthcare team.
Methods: This prospective questionnaire study sampled 200 patients
from My Diabetes My Way (MDMW), an interactive database for
113
DIABETICMedicine
patients in Scotland allowing access to personal online clinical

information and educational resources. The anonymous questionnaire
explored areas of current management techniques, technology literacy,
patient education, self-management and desirable features for a future
SA. Prior consent for patient contact had been agreed.
Results: A total of 200 patients from MDMW were contacted via
email; 122/200 participants responded. Results below highlight
response 1 or 2 vs response 4 or 5 on a Likert scale (1=strongly
agree and 5 =strongly disagree). Participant demographics: 56%
male, 44% female; 51%Type 1 and 48% Type 2; 58% <55 years
old and 42% >56 years old; 57% owned a smartphone. There is a
strong desire for SA development (48/68, 71%, 95%CI59 80).
Half of the participants would use SA for education (60/119, 50%,
95%CI42 59) and a similar number would prefer SA to current
methods of self-management (37/68, 54%, 95%CI42 66).
Desirable features included social media integration (70/116,
6%, 95%CI51 69). Carbohydrate tracking was undesirable (58/
119, 52%, 95%CI42 61).
Conclusions: Preliminary results from the study suggest that there is a
desire for SA,in particular offering integration with healthcare systems.
Providing patients with contemporary methods to support their care
will enable better health outcomes to be achieved and maintained.
P292
Findings from the Diabetes Attitudes,
Wishes and Needs (DAWN2TM) study
regarding education for diabetes in the UK
MJ Davies1, H Daly1, ME Carey1, K Barnard2, G Hall3,
NM Munro4 and RIG Holt2
1
Leicester Diabetes Centre Broadleaf, University of Leicester, Leicester, UK,

Faculty of Medicine, University of Southampton, Southampton, UK,
3
Portsmouth Community Diabetes Service, Solent Community Trust Diabetes
Service, Portsmouth, UK, 4Department of Health Care Management and
Policy, University of Surrey, Surrey, UK
2
Aims: This descriptive sub-analysis of UK data from the DAWN2

study examines the attitudes of people with diabetes (PWD), family
members (FMs) and healthcare professionals (HCPs) to diabetes
education.
Methods: Between 6 March and 22 May 2012, 500 adult PWD
(Type 1 diabetes, 81; Type 2 diabetes, 419), 122 FMs and 281
HCPs (GPs, 120; specialists, 81; nurses, 40; dietitians, 40) in the
UK completed the DAWN2TM survey. These data are not weighted
and therefore cannot be extrapolated across the UK.
Results: In the 12 months prior to the survey, only 15% (Type 1
diabetes) and 6% (Type 2 diabetes) of PWD and 4%7% of FMs
had attended a structured diabetes education programme. Of
available resources, PWD rely most on printed information for
help managing their diabetes (~40%). Of HCPs, 53% of GPs or
nurses/dietitians and 70% of specialists felt that the availability of
diabetes self-management education needs improvement, with
53%, 73% and 62%, respectively, agreeing that more qualified
nurse-educators should be available. Less than half GPs responding
felt that improving the availability of self-management education
would reduce the burden of diabetes.
Conclusions: Very few PWD and FMs have access to diabetes
education programmes, and many PWD rely on printed materials for
information. Whilst some HCPs agree availability of self-management
education should be improved, many do not, and remain unconvinced
of the benefits for their patients. If access to self-management
114
education is to increase, HCPs, particularly GPs, need to be made

aware of its importance and benefits for their patients.
Acknowledgement: On behalf of the DAWN2TM Study Group.
P293
Consistency of biomedical outcomes of
structured education over time: an 8 year
retrospective audit
S Jahan1, H Rogers2, P Choudhary1, SA Amiel1 and D Hopkins2
1
Diabetes Research Group, Kings College London School of Medicine,
London, UK, 2Department of Diabetic Medicine, Kings College Hospital NHS
Foundation Trust, London, UK
Background and aims: There is a risk that the efficacy of

structured education for Type 1 diabetes may be eroded due to
changes in staff and fidelity of delivery over time. To test for this we
have audited biomedical outcomes from DAFNE (Dose Adjustment
for Normal Eating) courses delivered over an 8 year period.
Method: This was a retrospective audit of 1year DAFNE
outcomes for all patients attending over 8 years at a single
centre, with division of the cohort into four sequential 2year
sub-cohorts.
Results: For the whole cohort (n = 367) mean ( SD) HbA1c
decreased from 8.32 1.3% to 7.81 1.1% (p < 0.001), with
no associated weight gain. 40.6% of the cohort achieved a
post-course HbA1c 7.5%. Prevalence of severe hypoglycaemia
over the previous year fell from 23% to 18% following
DAFNE (p < 0.001). Over the 8 years of the study there was a
fall in mean enrolment HbA1c from 8.8% to 7.9% (p < 0.005) but
mean post-course HbA1c was consistent across the four
sub-cohorts (8.1%, 7.6%, 7.9%, 7.6%, p = 0.2). No change was
observed between the sub-cohorts in the proportion of patients
achieving HbA1c 7.5%. For those with baseline HbA1c 7.5%
mean reduction in HbA1c post-DAFNE did not differ significantly
across the four sub-cohorts, ranging from 0.54% to 0.76%.
Conclusion: These data demonstrate that consistent glycaemic
outcomes can be achieved from an effective structured education
programme supported by quality assurance measures that maintain
intervention fidelity, despite significant changes in the
demographics of the population being enrolled over time.
P294
Patterns of self-monitoring of blood glucose
in Tayside, Scotland, among people with
Type 1 and Type 2 diabetes treated with
insulin
DM Cameron, J Evans and L Macaden
School of Nursing Midwivery and Health, University of Stirling, Stirling, UK
Aim: To investigate patterns and frequency of self-monitoring of

blood glucose (SMBG) over time in Tayside, Scotland (population
400,000).
Methods: Anonymised secondary data from the SCI-DC Tayside
diabetes clinical information system were used. For insulin-treated
people with Type 1 or Type 2 diabetes (13+ years), annual numbers
of SMBG strips dispensed and numbers of people receiving strips
were determined by year from 2004 to 2011. Comparison of strips
2014 The Authors.

Clinical care and other categories posters: epidemiology
dispensed by age, sex and SIMD (area measure of deprivation)

using t-tests and multiple variable ANOVA analysis was carried
out for 2009 2011.
Results: SMBG strips dispensed more than doubled from 618,950
among 1,666 people with Type 1 diabetes in 2004 to 1,269,300
among 1,936 people in 2011. The proportion of people receiving
strips increased from 58% to 69%, with mean annual strips
dispensed per person increasing from 641 to 947. In Type 2
diabetes, SMBG strips dispensed to insulin-treated people
increased from 696,441 among 3,605 people in 2004 to
740,591among 2,999 people in 2011. 48% were tested in 2004
compared with 45% in 2011, with testing frequency increasing
from 399/person/year to 553/person/year. In 2009 2011, strips
were much less commonly dispensed to older people (70+ years)
and those living in disadvantaged areas. There was a sex difference
in Type 1 diabetes only, with women receiving more strips.
Conclusion: Many insulin-treated people do not self-monitor,
particularly if they are older or living in disadvantaged areas; this
proportion has not changed substantially over time. In Tayside, the
increase in overall numbers and costs of SMBG strips is mainly due
to more frequent testing.
DIABETICMedicine
P295
My Diabetes My Way, an electronic personal
health record for diabetes care: an update at
3 years
SG Cunningham1, M Brillante1, B Allardice1, RR McAlpine1,
L Wilson2, R Kerr3, J Walker4, A Emslie-Smith5 and DJ Wake6
1
Clinical Technology Centre, University of Dundee, Dundee, UK, 2Monklands
Hospital, NHS Lanarkshire, Airdrie, UK, 3Patient Representative, NHS Fife,
Kirkcaldy, UK, 4St Johns Hospital, NHS Lothian, Livingston, UK, 5Tayside
Diabetes Managed Clinical Network, NHS Tayside, Dundee, UK, 6Medical
Research Unit, University of Dundee, Dundee, UK
Aims: My Diabetes My Way (MDMW) is NHS Scotlands

diabetes patient information portal, containing validated
educational materials, videos and interactive tools. MDMW
allows patients across Scotland access to their electronic diabetes
record. We analysed the first 33 months of usage and uptake.
Methods: Data are sourced from primary, secondary and tertiary
care via Scottish Care Information
Diabetes Collaboration
(SCI-DC).
MDMW
presents
diagnostic
information,
demographics, laboratory results, clinical parameters including
foot and eye screenings, prescribed medication and
correspondence. Explanations are given for normal values and
the reasons for assessments. Tailored information refers individuals
to facts related to their specific condition.
Results: At the end of September 2013, 5,706 individuals had
registered to access their data (59% male; 30% with Type 1
diabetes); 4,057 completed the enrolment process and 1,632
accessed the system (total logins 12,333; average eight
per patient; median 3; most logins 457). Audit trails
show 146,346 page views (90/patient); laboratory results proved
the most popular (29,721 accesses; 18/patient). The most utilised
history graph was HbA1c (6,569 accesses; four per patient).
Feedback includes: It is great to be able to view all of my results so
that I can be more in charge of my diabetes.
Conclusion: The system is now a useful aid to diabetes
self-management in Scotland. It is unique in offering access to an
entire national population, providing information from many
diabetes-related sources. Despite its development for stakeholders
in Scotland, it has the potential to connect to any electronic
medical record in the UK and beyond.
Clinical care and other categories posters: epidemiology

P296
Periodontal education: a neglected area in
diabetes care
P297
Factors contributing to mortality in Type 1
diabetes: a case control study
Z Vakil1, K Jeyaraman1, P Rihal1, C Gunpot1, E Umotong1,

I Bobboye1, S Sharma2, G Fenwick2 and R Kaushal1
K Allan1, C Garrett2, S Thomas3, SA Amiel1 and D Hopkins4
Department of Diabetes and Endocrinology, West Middlesex University

Hospital, London, UK, 2Heart of Smile, London, UK
Refer to Oral A7
1
Diabetes Research Group, Kings College London School of Medicine,
London, UK, 2Psychological Medicine, Institute of Psychiatry, Kings College
London, London, UK, 3Diabetes and Endocrinology, Guys and St Thomas
Hospitals NHS Trust, London, UK, 4Diabetic Medicine, Kings College Hospital
NHS Trust, London, UK
Aim: To identify factors associated with mortality in adults with

Type 1 diabetes.
Methods: This was a retrospective case control study including
all identified deaths occurring over 6 years (2006 2012) within a
cohort of 1,038 patients attending two specialist clinics. Each
deceased patient was matched to two survivors by age, gender,
ethnicity and socioeconomic status. Additional demographic and
2014 The Authors.

115
DIABETICMedicine
clinical data from 2002 to 2012 were collected from clinical

records and death certificates.
Results: Thirty-seven deaths occurred over 6 years (cumulative
mortality 3.6%); six deceased and three controls were excluded
due to incomplete records (31 deceased, 71 controls analysed). At
baseline (2006), those who died had higher HbA1c (mean SD,
8.8 1.8 vs 7.8 1.2%, p = 0.01), were more likely to have
microalbuminuria (58.0 vs 33.8%, p = 0.02) and to smoke (25.8
vs 5.5%, p = 0.02). They were less likely to have attended a
structured education programme (9.7 vs 35.2%, p = 0.01) and
were more likely to have a recorded history of hypoglycaemia
unawareness (32.2 vs 9.4%, p = 0.02). There were no differences
in the prevalence of psychiatric comorbidities. Cardiovascular
disease was the most commonly reported primary cause of death
(42%). Only two were attributed directly to diabetes (one DKA,
one unspecified), and diabetes was not mentioned on the death
certificate in 38% of cases.
Summary: These data demonstrate a higher prevalence of poor
glycaemic control, microalbuminuria, smoking, hypoglycaemia
unawareness and lower uptake of structured education among
patients who died. They also indicate significant under-reporting of
diabetes as a contributory factor in the certification of cause of
death.
Methods: An observational retrospective cohort design was

adopted using anonymised patient data for the complete
population of Tayside, Scotland, UK (approximately 400,000).
Patients were classified as having impaired glucose tolerance (IGT)
or impaired fasting glucose (IFG) if they had undergone two blood
glucose tests on the same date that met WHO 1999 criteria.
Patients with only fasting glucose of 6.0 6.9mmol/l were classified
as having undefined impaired glucose regulation (uIGR). A total of
50,321 patients met the criteria for either IFG (n = 2,284), IGT
(n = 1,996) or uIGR (n = 46,041) between 2003 and 2008.
Results: The total incidence of IGR across the study period was
2,720 per 100,000 person-years. Nine percent of people with IGR
progressed to Type 2 diabetes within a maximum of 9 years of
follow-up, with a mean time to diagnosis of 34 months. Male
gender, deprivation, IGT and increased age were found to confer
increased risk for progressing to Type 2 diabetes.
Conclusions: This study suggests that after an impaired glucose
regulation diagnosis there is sufficient time to intervene with
lifestyle preventative measures. Patients with impaired glucose
regulation at the highest risk of developing Type 2 diabetes are
those with IGT and from a deprived background; as such, these
patients should arguably be prioritised within prevention
programmes.
P298
Incidence of impaired glucose regulation
(IGR) and progression to Type 2 diabetes in
the Tayside region of Scotland
C Eades1, G Leese2 and J Evans1
1
2
School of Nursing, Midwifery and Health, University of Stirling, Stirling, UK,

School of Medicine, University of Dundee, Dundee, UK
Aims: To assess the incidence of IGR and to characterise

progression to Type 2 diabetes in adults in one region in
Scotland using routinely collected healthcare data.
Clinical care and other categories posters: erectile

disfunction
P299
Sexual morbidity in men with diabetes
CL Morrison1, G Morrison2, TS Purewal2 and PJ Weston2
1
Pendyffryn Medical Group, Prestatyn, UK, 2Diabetes Centre, Royal Liverpool

University Hospital, Liverpool, UK
Introduction: Physical, psychological and social problems can

engender an intricate interplay on the sexual function,
dissatisfaction and unwellness in the male with diabetes.
Addressing these needs can lead to physical benefits on
cardiovascular health and decreased pain sensitivity. Emotional
benefits include increased relaxation, decreased levels of
depression, improved quality of life and self-esteem. This study
was undertaken to determine the extent of sexual dysfunction in
men with diabetes in a primary care setting.
Methods: Through sexual history interviews with men with
diabetes, data were collected for genito-urinary conditions,
current sexual status, spousal health, medication use, presence of
erectile dysfunction and the sexual health inventory for men
(SHIM).
116
Results: In 45 men with diabetes [Type 2, 38(84.4%) (mean,SD,

range) age 66.4 years 1.8,33 91; HbA1c 64.5mmol/mol
3.1,42 132; totaltestosterone (n = 27, 60%) 11.3nmol/l 5.3,
4.1 28.8; prostate specific antigen (n = 32,71.1%) 1.61mcg/l
1.66,0.22 6.8], 32 (71%) reported erectile difficulties and 21
(46.7%) reduced libido. Last time of sexual intercourse was 4.1
years 5.96,0.01 30, 31(68.9%) were in a current relationship
(one same-sex) (duration 33.4 years 21.3,0.3 59), five (16.1%)
with a partner unable to reciprocate intimately. Nineteen (42.2%)
had sex in the last 6 months (SHIM score 15.8 8.2,5 25). Ten
(22.2%) were using phosphodiesterase type-5 inhibitors, 14
(31.1%) treated with insulin, six (13.3%) antidepressants, 19
(42.2%) beta-blockers, eight (17.8%) medication for benign
prostatic hyperplasia.
Conclusion: Regular sexual expression is associated with
improved physical and mental health. However, clinicians may
be reluctant to compile comprehensive sexual histories due to time
constraints and discomfiture. Identification of erectile dysfunction
has recently been introduced into quality of life but patients may be
disaffected if its management is not undertaken holistically.
2014 The Authors.

Clinical care and other categories posters: ethnicity
DIABETICMedicine

P300
P302
Use of anthropometric indices in predicting

prevalent Type 2 diabetes in White
European and South Asian adults in
Leicestershire
Religious difference in the prevalence of

hypertension in people with and without
diabetes in an urban adult Indian
population
BA Jamjoom, DH Bodicoat, P Carter, S Mostafa, D Webb,

K Khunti and MJ Davies
R Selvarajan1, V Kumar1 and S Rajbhandari2,3
Leicester Diabetes Centre, University of Leicester, Leicester, UK
Refer to Oral A42
P301
Designing effective culturallycompetent
diabetes care service in primary care: a
participatory research study to implement
evidence
P Zeh1,2, HK Sandhu1, AM Cannaby3 and JA Sturt4
1
Warwick Medical School, University of Warwick, Coventry, UK, 2Research
and Development, University Hospitals Coventry and Warwickshire NHS Trust,
Coventry, UK, 3Hamad Medical Corporation, Doha, Qatar, 4Florence
Nightingale School of Nursing and Midwifery, Kings College, London, UK
Aims: Systematic reviews have identified effective interventions

for improving cultural competence in diabetes healthcare. This
study aimed to explore and design an intervention to cater for the
needs of ethnic minority groups (EMGs) with diabetes in
Coventry.
Methods: A purposive participatory case study was conducted
in one Coventry inner-city practice between April 2012 and
March 2013, using weekly participant observations, monthly
participatory group meetings and one post-participatory
one-to-one semi-structured interview with three patients, four
general practice (GP) staff and two multilingual link workers
(MLWs). Data were analysed using qualitative methods. The
research explored cultural issues with staff and patients within the
GP and designed a culturally-competent diabetes service
framework.
Results: The operational activities of this GP involving staff and
patients demonstrated both helpful (e.g. some ethnic minority
patients with cultural differences being consulted by healthcare
professionals from the same ethnic backgrounds, thereby ensuring
cultural concordances) and unhelpful practices (e.g. using children
to interpret for their parents and at times patients with language
barriers consulted without the use of professional or lay
interpreters). The practice prioritised the designing of a diabetes
specialist multilingual link worker (DSMLW) model, to reduce the
inequality in diabetes service provision. Key elements of the model
were specialist training of MLWs, referral pathway, information
sharing, partnership working, and service commissioning.
Conclusions: A DSMLW framework to address deficits in GP
cultural competence is now ready for pilot testing. Involving a
broad group of stakeholders ensures that interventions to improve
EMGs access to effective diabetes care in primary care are
appropriate and feasible.
2014 The Authors.

1
3
Diabetacare, Bangalore, India, 2University of Central Lancashire, Preston, UK,

Lancashire Teaching Hospital, Chorley, UK
Aims: The aim of the study was to see if there were any differences
in characteristics of patients between predominantly Muslim and
Hindu urban Indian populations from Bangalore.
Methods: The general public were invited for a health check-up
including capillary blood glucose tests during public events
performed in the mosque and the community centre in
Bangalore. Simple history, anthropometric data, blood pressure
and random blood glucose were measured. Only those subjects
above the age of 30 were analysed.
Results: In all, 367 predominantly Muslims were screened in the
mosque of whom 326 were above the age of 30. In the community
centre 102 predominantly Hindus were screened of whom 86 were
above 30. Subjects who were known to have diabetes (DM) were
analysed separately from those who were not known to have
diabetes (non-DM). Uncontrolled blood pressure (>160/90 mm
Hg) was more prevalent amongst the Muslim population both in
the DM group (51.2% vs 27.3%; p < 0.01) and the non-DM group
(51.2% vs 27.7%; p < 0.01). Overweight (body mass
index > 25kg/m2) was common in subjects with diabetes in both
communities (71.3% vs 69%; p = NS). There were no differences
(p > 0.05) in age, body mass index and proportion of diabetes
patients but more males were screened in the mosque.
Discussion: Our data show that overweight is common in both
communities; however, uncontrolled blood pressure was twofold
higher amongst Muslims. Diabetes is often associated with
hypertension, which can be influenced by a number of factors.
Religious belief is possibly one of them and further studies are
needed to explore it.
P303
Detection of haemoglobin E on HbA1c
measurement using new software
developed for the Tosoh G8 IE HPLC analyser
SB Stilwell1, A Surj1, RA Round1, CL Mason1, J Webber2,
R Cramb1,3 and SE Manley1,3
1
Clinical Laboratory Services, University Hospitals Birmingham NHS
Foundation Trust, Birmingham, UK, 2Diabetes Centre, University Hospitals
Birmingham NHS Foundation Trust, Birmingham, UK, 3Division of Medical
Sciences, University of Birmingham, Birmingham, UK
Aim: To evaluate new software on an ion exchange high

performance liquid chromatography (IE HPLC) HbA1c analyser
to detect haemoglobin E, a rare variant haemoglobin common in
Southeast Asia, as haemoglobin E precludes the reporting of
HbA1c. The flag developed by Tosoh over 10 years is also
generated by other abnormal haemoglobins.
Methods: Flagged chromatograms were triggered by a peak and
valleys between 0.65min and 0.75min. Blood was referred
anonymously to confirm haemoglobin E on b-thalassaemia mode
on another Tosoh G8 IE HPLC analyser if A2 > 10%.
117
DIABETICMedicine
Results: Over 11 months, 79 patients had suspect E flags with 50

(63%) blood samples run on b-thalassaemia mode. This resulted in
the confirmation of haemoglobin E in 42(84%) patients. Abnormal
haemoglobins were detected in all the other eight patients
including one patient heterozygous for haemoglobin G-Coushatta
and another for haemoglobin Pyrgos that co-elutes with HbA1c.
Since the introduction of HbA1c for diagnosis, the workload in the
laboratory has doubled to about 10,000 tests per month giving a
prevalence for haemoglobin E of approximately 0.1% in this
multi-ethnic population referred for HbA1c measurement by local
hospitals and general practitioners in the West Midlands.
Conclusions: Haemoglobin E or other abnormal haemoglobins
precluding HbA1c reporting were detected in all the
chromatograms flagged as suspect E. Requestors were alerted if
HbA1c was reported before this new software was introduced.
Given the widespread use of HbA1c for diagnosis of diabetes, it is
important to identify such patients when measuring HbA1c.
P304
Current status of diabetes care in Wachet,
Rural Upper Myanmar
TT Hlaing1,2, MW Aung3, O Myint4, AP Naing3, ZM Htet4,
AT Khaing3, EK Khaing3 and KS Myint2,5
1
Department of Clinical Biochemistry, Cambridge University Hospital NHS
Foundation Trust, Cambridge, UK, 2Brighter Future Foundation Charity,
Cambridge, UK, 3Department of Medicine, North Okkalapa General Hospital,
Yangon, Myanmar, 4General Practitioner, Yangon, Myanmar, 5Department
of Diabetes and Endocrinology, Norfolk and Norwich University Hospital NHS
Foundation Trust, Norwich, UK
Aim: There is no national guideline or structured care for diabetes

in Myanmar. In this pilot project, we aimed to identify the current
diabetes care status and delivered a free, one-stop comprehensive
diabetes care in Wachet Sangha Hospital, Upper Myanmar.
Methods: Patients with diabetes were invited to attend diabetes
clinic in a national newspaper (with a capacity of 260 patients over
4 days). NICE recommended annual review screenings were
carried out. Detailed clinical consultation and management plan
were provided including 2 weeks supply of medication and
structured group education sessions.
Result: In total 260 patients were reviewed (67% female, all but
one had Type 2 diabetes) with a demographic (mean SD) of age
55.3 10.9years, body mass index 26.2 5.8kg/m2, waist
circumference 87 14.7cm, duration of diabetes 4.4 4.6years
and history of hypertension35.8%, ischaemic heart disease7.3%
and stroke1.2%. On screening, retinopathy(R) was identified in
40%(one proliferative requiring laser therapy, two grade R3, five
R2 and 28 R1). Others included cataract (n = 6) and hypertensive
retinopathy (n = 4); 36% (n = 93) had blood pressure>140/
90mmHg, 22%(n = 58) had 160/100mmHg and 10%(n = 27)
had 180/110mmHg. Only 10.8% were on a statin whereas 75%
had total cholesterol > 5mmol/l. HbA1c reported > 58mmol/mol
(78%), 86mmol/mol (43%), 97mmol/mol(21%), >151.4mmol/
mol(n = 4) whereas only 0.01%(n = 9) was on insulin therapy.
Four patients had undiagnosed chronic kidney disease stage
4(eGFR < 30ml/min/1.73m2) and 18 had stage 3, and
microalbuminuria was identified in 0.05%(n = 15). Feet
examination was abnormal in 17% cases.
Conclusion: We identified this large unmet area of diabetes care
in this pilot group, which will provide a platform for setting up
future diabetes services in Myanmar.
118
P305
Metabolic control among 15,989 Emirati
patients with diabetes
A Bakir1, N Lessan1, S Narayan2, Z Hannoun1 and MT Barakat1
1
Research, Imperial College London Diabetes Centre, Abu Dhabi, United Arab
Emirates, 2University of Birmingham, Birmingham, UK
Aim: To evaluate metabolic control among a large cohort of

Emirati patients with diabetes
Methods: ICLDC 2012 database includes 40,099 patients
including 20,852 with diabetes. Last visit data [body mass index
(BMI), HbA1c, lipid profile and blood pressure (BP)] from patients
were accessed. Emirati patients achieving American Diabetes
Association (ADA) 2012 target metabolic criteria (HbA1c
7.0%, BP 130/80mm Hg, LDL cholesterol 2.6mmol/l) were
identified.
Results: In all, 15,989/20,852 patients with diabetes (Type 2,
n = 14,999; Type 1, n = 990) were of Emirati origin
(age 52.02 14.2 years; female 52.4%; BMI = 31.0 6.4kg/
m2). In this group, target glycaemic control, BP and LDL
cholesterol were met in 49.1%, 63.9% and 48.5% of patients
respectively; 15.7% satisfied all three criteria. Patients with Type 1
diabetes had a lower BMI, worse overall glycaemic control and a
higher LDL cholesterol (BMI = 24.3 6.5kg/m2; HbA1c = 9.1
2.2%; LDL = 3.0 0.8mmol/l; BP = 115.1 1 6.8/67.5
10.1mm Hg) but a lower BPcompared with those with Type 2
diabetes (BMI = 31.5 6.1kg/m2; HbA1c = 7.5 1.6%; LDL =
2.7 0.9mmol/l; BP = 125.8 18.0/70.2 9.8mmHg).
Conclusion: A significant group of Emirati patients in this large
study group do not fulfil accepted criteria for metabolic control.
Further concerted efforts into improving glycaemic control in Type
1 diabetes and blood pressure as well as weight in patients with
Type 2 diabetes in this population are needed.
P306
South Asian adults with insulin treated
young-onset Type 2 diabetes have a lower
body mass index and better control than
other ethnic groups
S Misra, T Brabarupan, DG Johnston and NS Oliver
Diabetes and Endocrinology, Imperial College, London, UK
Aims: South Asian (SA) and African Caribbean (AC) people in the
UK are increasingly affected with young-onset Type 2 diabetes,
although incidence is also rising in white European (WE) people.
Ethnic differences are well described in adult-onset Type 2
diabetes. We aimed to determine the characteristics of people
with young-onset Type 2 diabetes according to ethnic origin and
whether insulin treated (IT).
Methods: A cross-sectional retrospective study of adults
diagnosed <35 years of age from WE, AC or SA ancestry was
undertaken from an electronic database in a London clinic.
Results: A total of 203 people were identified with an equal ethnic
split. 66.5% of all individuals were insulin treated with no ethnic
variations. In the non-IT group, body mass index (BMI) did not
vary by ethnicity (34.6kg/m2 WE, 30.6kg/m2 AC and 31.9kg/
m2 SA) and HbA1c was significantly lower in the WE group (6.9%
vs 8.7% AC and 8.5% SA, p = 0.009). On analysis of the IT
group, BMI varied significantly between each ethnic group
(34.1kg/m2 WE, 34.0kg/m2 AC and 28.4kg/m2 SA, p < 0.0001)
and HbA1c was significantly lower in the SA group (8.1% vs 9.0%
AC); however, duration of diabetes was not different across all
three groups.
2014 The Authors.

Conclusion: Despite no difference in diabetes duration significant

differences in anthropometry, treatment and control were
identified in young SA and AC people with diabetes compared
with WE. SA people with young-onset Type 2 diabetes have lower
BMIs and better control than other ethnic groups on insulin
treatment. In contrast AC people have higher BMIs, associated
with worse control. Further work is needed to investigate whether
these differences reflect ethnic variations in management or disease
process.
P307
Type 2 diabetes in young South Asian
patients leads to increased cardiovascular
comorbidity and decreased survival
R Potluri1, L Pichaipillai2, S Rayasamudra3, AK Ravi4, H Uppal5,
D Lavu6 and S Chandran7
1
School of Medical Sciences, University of Aston, Birmingham, UK,

Department of Diabetes, Pennine Acute Hospitals NHS Trust, Manchester,
UK, 3Medicine, Chesterfield Hospital, Chesterfield,UK, 4Medicine, Wayne
State University, Michigan, USA, 5Department of Psychiatry of Learning
Disability, Brooklands Hospital, Birmingham, UK, 6Obstetrics and
Gynaecology, Keele University, Stoke-on-Trent, UK, 7Acute Medicine, North
Western Deanery, Manchester, UK
2
Introduction: The prevalence of Type 2 diabetes in SouthAsian

populations is increasing significantly particularly in younger age
groups. Evidence is lacking regarding the prevalence of
cardiovascular comorbidities and their impact on health in young
SouthAsian patients with Type 2 diabetes compared with
Caucasian populations.
Methods: Anonymous information on patients with Type 2
diabetes, comorbidities and mortality attending seven hospitals in
Manchester, UK, in the period 2000 2013 was obtained from the
local health authority computerised hospital activity analysis
register. We evaluated Type 2 diabetes in SouthAsian and
Caucasian patients between the ages of 20 and 60.
Results: Of the Type 2 diabetes population of 68,194 (5.6%),
there were 19,469 patients between the ages of 20 and 60 in
South Asian and Caucasian groups. Of these, there were 4,918
South Asian patients and 14,551 Caucasian patients. Mean age in
the South Asian group (48.1 years) was lower than in the
Caucasian group (49.9 years; p < 0.05). Cardiovascular
comorbidities such as hyperlipidaemia (25.7% South Asian vs
20.8% Caucasian), hypertension (49.8% South Asian vs 48.3%
Caucasian), ischaemic heart disease (23.4% South Asian vs
21.0% Caucasian) and myocardial infarction (5.9% South
Asian vs 5.2% Caucasian) were higher in South Asian
groups. Heart failure (3.6% South Asian vs 4.4% Caucasian)
and atrial fibrillation (1.8% South Asian vs 4.3% Caucasian)
rates were higher in the Caucasian group. The average length of
hospital stay was significantly longer in the Caucasian group (4.1
days) compared with the South Asian group (3.0 days). The mean
survival from first diagnosis to death in these patients was
significantly lower in the South Asian group (2,078 days)
compared with the Caucasian group (2,102 days).
Conclusion: Type 2 diabetes in young South Asian patients leads
to increased cardiovascular comorbidity and decreased survival
compared with Caucasian counterparts in the same age group who
are actually slightly older. In the context of increasing Type 2
diabetes in patients of South Asian origin, these data suggest that
new strategies are required to prevent healthcare debacles in these
young patients.
2014 The Authors.

DIABETICMedicine
P308
Comparing outcomes of exenatide/insulin
and exenatide/oral anti-hyperglycaemic
combination therapies
R Shankland, R Oelbaum, J Simmonds and A Allen
Diabetes and Endocrinology, Southport and Ormskirk NHS Hospital Trust,
Southport, UK
Introduction: Type 2 diabetes and obesity are an increasing

problem in the western world. The synthetic exendin-4 exenatide
has been used in the management of Type 2 diabetes for several
years. Short-term benefits including weight loss and reduction in
HbA1c are noted. Although not licensed for use in combination
with insulin, this practice is commonplace within many NHS
trusts. We aim to identify which combination therapy provided
better short-term outcomes.
Methods: We performed a prospective study. All exenatide
patients were invited to attend assessment at 0, 3 and 6
months. Patient parameters: weight, HbA1c and insulin
requirements were documented.
Results: Fifty-two patients using exenatide combination therapy
were identified. Thirty-six patients completed 0, 3 and 6 month
assessments. Sixteen patients were on exenatide/insulin (EI)
therapy and 20 patients were on exenatide/oral antihyperglycaemic (EO) therapy. In the EI group average weight was
98.8kg, average weight reduction was 2.5kg,average daily insulin
requirement was 128 units and average reduction in insulin
requirement was 15 units. Average initial HbA1c was 9.9% and
average reduction in HbA1c was 0.92%. In the EO group the
average weight was 116.5kg, average weight loss was 3.5kg, average
initial HbA1c 8.7% and average reduction was 0.85%.
Conclusions: Weight loss was greater in the EO group. Reduction
in HbA1c was similar in both groups. Small study, insulin was used
safely in combination with exenatide. Both combination therapies
provided short-term improvements in baseline patient parameters.
P309
The potential impact of changing from WHO
to IADPSG criteria for screening
for gestational diabetes (GD) in a
multi-ethnic UK population
H Venkataraman1, N Sukumar1, P Pushton2, N Anderson2 and
P Saravanan1,3
1
Metabolic and Vascular Health, Warwick Medical School, Coventry, UK,

Department of Pathology, George Eliot Hospital, Nuneaton, UK, 3Diabetes
and Metabolism, George Eliot Hospital, Nuneaton, UK
2
Aims: To compare the impact of using IADPSG vs WHO criteria

for diagnosing GD.
Methods: Retrospective data was collected for all antenatal
OGTTs between Jan 2008 and 2013 across 3 centres in West
Midlands that use selective screening and WHO criteria. A
descriptive analysis was undertaken comparing WHO (FPG6.1
or 2hPG 7.8mmol/l) and IADPSG criteria (FPG 5.1 or 2-hPG
8.5mmol/l).
Results: Of 14360 women screened, the prevalence of GD was:
WHO 8.9% (n=1289) and IADPSG 11.5% (n=1667).
Meansd of FPG & 2hPG were: WHO 5.01.03 &
8.91.7mmol/l and IADPSG 5.30.82 and 7.82.2mmol/l.
Ethnicity distribution of GDM by WHO and IADPSG was, South
Asians (SA) 27.1 & 26% White Caucasians (WC) 64.3 & 65.2%.
IADPSG increased the prevalence of GDM by 29% (SA: 24% &
WC: 30%).
119
DIABETICMedicine
Conclusion: IADPSG criteria diagnose additional high risk

women with milder hyperglycaemia at the expense of missing a
third of WHO diagnosed GD, for whom treatment benefit has been
shown. Hence, we need more interventional evidence before

adopting IADPSG criteria.
Clinical care and other categories posters: foot

P310
P312
Outpatient parenteral antibiotic therapy

(OPAT) in diabetes foot infection
High rate of mortality on follow-up of

patients with diabetes and ischaemic feet
admitted to hospital with infection
U Srinivas-Shankar1, S Keigan1, P Chiran1, A Lewis2,

S OHanlon3, P Hodgkins3, P McGrail3 and KJ Hardy1
1
Department of Diabetes and Endocrinology, St Helens and Knowsley

Teaching Hospitals NHS Trust, St Helens, UK, 2Pharmacy Department, St
Helens and Knowsley Teaching Hospitals NHS Trust, St Helens, UK,
3
Community IV Antibiotic Team, Bridgewater Community Healthcare NHS
Trust, St Helens, UK
Refer to Oral A64
P311
Data from EQ-5D and SF-36 indicate reduced
quality of life in patients with acute Charcot
osteoarthropathy
T Jemmott, V Morris, M Bates, ME Edmonds, NL Petrova and
W Tang
Diabetic Foot Clinic, Kings College Hospital NHS Foundation Trust, London,
UK
Aims: To measure quality of life in patients presenting with acute

Charcot osteoarthropathy.
Methods: We studied 46 patients (15 Type 1 and 31 Type 2
diabetes; 36 males and 10 females) presenting to a single centre
with a red hot swollen foot with either abnormal foot radiographs
with typical Charcot changes or abnormal MRI scan. Quality of
life was measured using EQ-5D and SF-36 questionnaires. Data on
mobility, self-care, usual activities, pain/discomfort and anxiety/
depression together with respondents self-rated health on a
vertical, visual analogue scale (VAS) was obtained from EQ-5D.
From SF-36 we calculated the physical and mental component
scores (PCS and MCS).
Results: Data from EQ-5D showed that 70% of our patients
reported problems with usual activity and mobility, 40%
experienced pain and anxiety/depression and 20% had problems
with self-care. The mean health score (EQ-5D) was reduced at 67
(compared with 84 for the general population). Data from SF-36
showed that both PCS and MCS were reduced at 36 and 49
respectively (compared with PCS and MCS of 51 each for the
general population). There was no difference in mobility, self-care,
usual activities, pain/discomfort and anxiety/depression between
males and females. The MCS between males and females was also
similar. However, the PCS was significantly reduced in Charcot
females compared with Charcot males (28 vs 38, p = 0.02).
Conclusions: Data from EQ-5D and SF-36 indicated a severe
reduction in quality of life in patients with acute Charcot
osteoarthropathy and every healthcare professional should be
aware of this when managing patients with this condition.
120
MS Atim1 and ME Edmonds2

1
2
Diabetic Foot Clinic, Kings College London Medical School, London, UK,
Diabetic Foot Clinic, Kings College Hospital, London, UK
Aims: To conduct a retrospective study to evaluate the outcomes

of patients who had been admitted with a diabetic foot infection
and to determine whether patients with ischaemia have worse
outcomes than patients without ischaemia.
Methods: Patients that had been admitted with a diabetic foot
infection were divided into two groups, ischaemic and
non-ischaemic. Patients with ischaemia had absent foot pulses
and damped Doppler waveforms and patients without ischaemia
had palpable pulses. Baseline details recorded were age, sex, serum
creatinine levels (to reflect kidney function) and comorbidities
present. The patients were followed up for a mean of 17.6 months
(range 15 20 months). Outcomes recorded were death,
readmissions and recurrence of diabetic foot infections.
Results: In the sample population of 74 patients, 39 ischaemics
and 35 non-ischaemics, there was a high mortality rate of 20.3%
which was significantly related to ischaemia (p = 0.003) and age
(p = 0.004). Patients in the deceased group also had a significantly
higher Charlson comorbidity score (p = 0.01). Cardiovascular
causes and infection were the main causes of death. Age was
significantly higher in the ischaemic group (p = 0.003) as well as
the number of minor amputations (p = 0.008). Non-ischaemic
patients had significantly more readmissions for recurrent foot
infections (p = 0.004). The main comorbidities present were
kidney disease (28%), which was most prevalent in the
non-ischaemic group, and cardiovascular disease (30.6%), which
was most prevalent in the ischaemic group.
Conclusion: Diabetic foot infections are associated with a poor
outcome, particularly in patients with ischaemia who have high
mortality on follow-up.
P313
Outcome of osteomyelitis in patients with
diabetic foot ulceration
G Spyer1, L Davis2, R Gornall3, J Smith1, R McCarthy4 and
J Davis5
1
Department of Diabetes and Endocrinology, Torbay Hospital, Torquay, UK,

Nuffield Bursary Student, Torbay Hospital, Torquay, UK, 3Department of
Podiatry, Torbay Hospital, Torquay, UK, 4Department of Vascular Surgery,
Torbay Hospital, Torquay, UK, 5Department of Orthopaedics, Torbay Hospital,
Torquay, UK
2
Aim: To determine treatment regimes and healing rates in patients

with diabetes diagnosed with osteomyelitis resulting from a
diabetic foot ulcer.
Method: Suitable patients were identified as having attended the
diabetes multidisciplinary foot service with a diagnosis of
2014 The Authors.

osteomyelitis from January to December 2012. Case notes were

reviewed to determine the type and duration of antibiotics used, the
number of patients requiring surgery and the number of ulcers healed.
Results: Thirty-five cases in 33 patients were identified during
2012. Eighteen (51%) were managed with antibiotics alone and 17
(49%) required surgery and/or revascularisation [one (3%) below
knee amputation, 12 (34%) minor amputation or debridement,
eight (23%) angioplasty, two (6%) bypass grafts]. There were 120
swab and tissue results; 60% yielded a positive result.
Staphylococcus aureus was the most common pathogen isolated
in 18(51%) cases; 21(60%) were polymicrobial; in eight (23%) no
organism was isolated. When superficial swab results were
compared with deep tissue samples there was only 20%
concordance. On average patients required 2.4 changes to their
antibiotic regime. The mean duration of antibiotic treatment was
81 days (range 7 161 days). Twenty-two (63%) patients healed;
this was similar in both medically (61%) and surgically (65%)
treated patients. Seven (20%) patients had died and five (14%)
were requiring ongoing treatment at the time of audit.
Conclusion: Antibiotic regimes were variable and frequently
deviated from guidelines. Equivalent healing rates were obtained
in both medically and surgically treated patients, but mortality was
high. Swabs are not an accurate way of identifying a causative
organism and correlate poorly with deep tissue samples.
P314
The challenges of improving inpatient
diabetes foot care: the Severn
Deanery Project designed by diabetes
specialist registrars to better protect people
with diabetes whilst inpatients in hospitals
and raise the awareness of good foot care in
an area with high rates of amputation
C Papastathi1, G Russell1, K Gkastaris2, I Khan4, C Eboh2,
S Ahmed5, H Tan3, B Houlford3, I Ghaffar6 and S Thrower1
1
Department of Diabetes and Endocrinology, University Hospitals of Bristol,
Bristol Royal Infirmary, Bristol, UK, 2Department of Diabetes and
Endocrinology, Royal United Hospital Bath, Bath, UK, 3Department of
Diabetes and Endocrinology, Musgrove Park Hospital, Taunton, UK,
4
Department of Diabetes and Endocrinology, Weston General Hospital,
Weston, UK, 5Department of Diabetes and Endocrinology, Gloucester Royal
Hospital, Gloucester, UK, 6Department of Diabetes and Endocrinology, North
Bristol NHS, Bristol, UK
Aim: To improve the frequency and quality of foot checks in

people with diabetes on admission to hospitals in the Severn
Deanery, southwest England, by July 2013. Important skills for
specialist registrars (SpRs) include quality improvement (QI)
methodology. Twelve SpRs of seven hospitals designed a joint QI
project in 2013.
Methods: A high quality foot check was defined and baseline data
in at least three ward areas in each hospital were collated. Through
understanding the system and processes for admission with diabetes,
the SpRs created a driver diagram. This identified what work/
interventions were to be tested, in order to affect change. The
interventions were worked on collaboratively across the region. They
included an education programme, communications policy with
items to use, e.g. posters, a standard method and proforma. Regular
data collections were made to monitor progress on pilot wards up to
August 2013. The foot check data of NaDIA 2013 were reviewed as
reference to progress being maintained and spread.
Results: Baseline data showed that <6% of people with diabetes
had a foot check within 24h. Through the interventions which
2014 The Authors.

DIABETICMedicine
were trialled and shared this rose to 25% across the Severn
Deanery. NaDIA data showed a rise from 27% to 33%; whilst
pleasing, not impressive.
Conclusion: Change is hard. In the first 6 months improvement
was demonstrated and associated with a modest improvement in
NaDIA. The interventions tested were effective as demonstrated by
a fourfold rise in the data. Further work on embedding and
spreading the work continues.
Acknowledgement: Additional authors include, J Triay, M
Chen, R Gouni, F Chau and B Varupula.
P315
A systematic review examining the
prevalence, progression and outcomes of
foot ulcers and lower limb amputations in
South Asians and White Europeans with
Type 2 diabetes
AN Butt1, K Khunti1, W Crasto2, P Carter3 and MJ Davies3
1
Department of Health Sciences, University of Leicester, Leicester, UK,

Department of Diabetes and Endocrinology, University Hospitals of Leicester
NHS Trust, Leicester Royal Infirmary, Leicester, UK, 3Department of
Cardiovascular Sciences, Leicester General Hospital, Leicester, UK
2
Aim: South Asian (SA) populations compared with White

Europeans (WE) have a higher risk of developing Type 2 diabetes
and its vascular complications. However, some data in the literature
have suggested lower risk of foot ulceration in SA patients and the
perception that SAs have lower amputation rates. We sought to
conduct a systematic review to examine the prevalence, progression
and outcomes of foot ulcers and lower extremity amputations in SA
vs WE individuals with Type 2 diabetes.
Method: An electronic literature search was conducted using
MEDLINE, EMBASE and Cochrane databases.
Results: The review included two studies which showed that the
prevalence rates of foot ulcer and lower extremity amputation in WE
and SA was 4% and 2.3%, respectively. One study showed a reduced
risk of amputation in SA vs WE [odds ratio 0.37 (95%CI 0.15 0.89),
p = 0.03]. Paucity of studies and considerable heterogeneity between
the two studies precluded further meta-analysis. None of the studies
compared the progression and outcomes of foot ulcer and amputation
in SA and WE with Type 2 diabetes.
Conclusion: The review identified only two studies. There were
no differences in prevalence rates; however, foot ulcer and
amputation rate estimates were higher in WE compared with SA.
Overall prevalence rates were not computed due to lack of study
data. This review did not identify any outcome studies comparing
clinical outcomes such as cardiovascular comorbidities, survival,
metabolic control or quality of life outcomes. Further research in
this area is warranted.
P316
Diabetes led or vascular led: does it affect
root cause? A root cause analysis of major
foot amputations in patients with diabetes
at Newcastle Primary Care Trust (NPCT)
A Santhakumar1, N Coates1,2, S Manohar1 and N Leech1
1
Newcastle Diabetes Centre, Newcastle-upon-Tyne Hospitals NHS Foundation
Trust, Newcastle-upon-Tyne, UK, 2Newcastle Hospitals Community Health
Podiatry Department, Newcastleupon-Tyne Hospitals NHS Foundation Trust,
Aim: Disintegrated primary and secondary care was highlighted as

a contributor in root cause analysis (RCA) of diabetes related
121
DIABETICMedicine
amputations in London. The National Diabetes Audit 2012

reported 1.14 major amputations per 1,000 people with diabetes
in NPCT. We aim to identify and apply the RCA tool to the 38
major amputations in NPCT and map individual foot care
pathways.
Method: All diabetes related major amputations in NPCT
between April 2008 and March 2011 were identified from
surgical lists and RCA was applied to retrospective data collected
from podiatry and hospital records.
Results: There were 38 major amputations in 34 patients. Mean
age was 67 years (43 92). Risk factors included peripheral vascular
disease (91%), neuropathy (85%) and previous amputations
(25%.). Thirty-one (91%) patients received podiatry care
including 21(61%) in a multidisciplinary foot clinic (MDT). Mean
interval between podiatry review and admission for amputation was
29 days (0 161). Mean length of stay was 47 days. Twenty-two
(65%) patients have died post amputation, 11 (50%) within 2 years.
Foot care was suboptimal in 17 (45%) patients due to poor patient
concordance and gaps in care including lack of podiatry input post
vascular referral and non-referrals into MDT.
Conclusions: 91% of patients were known to podiatry compared
with 54% in the London group. Gaps in care post vascular referral
and patient concordance contributed to amputations in our high
risk group. Early MDT referrals, joint diabetes vascular clinics and
a foot care coordinator may reduce our amputations.
P317
What are the key factors in avoidable
amputations?
RA McDonnell, D Selvarajah, FM Creagh and RA Gandhi
Department of Diabetes, Sheffield Teaching Hospitals NHS Foundation Trust,
Sheffield, UK
Background and aims: Although it is recognised that many

amputations are avoidable, it is not clear what factors in the
complex pathway to amputation are important in avoidable
amputations. Better understanding of these factors is critical if
the numbers of amputations are to be reduced. The aim of this
study was to carry out a detailed review of all amputations in
patients with diabetes in 2012 within our service.
Methods: Detailed root cause analyses were carried out for 146
amputation episodes (identified from the surgical database). The
determination of whether an amputation was potentially avoidable
and the contributory factors was carried out by two expert
assessors (diabetes consultants with expertise in foot care).
Results: 21% of amputations were potentially avoidable. Key
factors involved in cases of avoidable amputation were delayed
referral by primary care (26%), delayed referral to vascular
surgery (13%), delayed vascular investigations or intervention
(17%), inappropriate antibiotics (39%), antibiotics discontinued
inappropriately (22%) and patient noncompliance (4%). Only
34% of patients underwent revascularisation before or at the time
of amputation.
Conclusions: This study showed that at least a fifth of all
amputations were potentially avoidable. By far the most
important contributory factor was problems with antibiotic
administration, but delays in referral, investigations and
interventions also played an important role. The care of people
with diabetic foot problems is complex, involving multiple
disciplines across primary and secondary care. Services need to
allow patients to navigate efficiently through the care pathway.
The findings of this study have important implications for policy
development and service redesign.
122
P318
Successful conservative management of
calcaneal osteomyelitis with avulsion
fracture in a diabetic cohort
V Bravis1,2, WM Gedroyc3, P Rosenfeld4, RGJ Gibbs5,
DG Samarasinghe6, J Valabhji1, S Robinson1 and R Fikri1
1
Department of Diabetes and Endocrinology, St Marys Hospital, Imperial
College Healthcare NHS Trust, London, UK, 2Department of Investigative
Sciences, Imperial College, London, UK, 3Department of Radiology, St Marys
Hospital, Imperial College Healthcare NHS Trust, London, UK, 4Department of
Orthopaedic Surgery, St Marys Hospital, Imperial College Healthcare NHS
Trust, London, UK, 5Department of Vascular Surgery, St Marys Hospital,
Imperial College Healthcare NHS Trust, London, UK, 6Department of
Microbiology, St Marys Hospital, Imperial College Healthcare NHS Trust,
London, UK
Objectives: To evaluate the conservative management of

calcaneal osteomyelitis with avulsion fracture in three patients
with diabetes and peripheral neuropathy.
Methods: In this prospective cohort study, management included
antibiotic therapy, debridement, vascular optimisation, vacuum
pump therapy and heel offloading. Healing progress, ankle
function preservation and quality of life were assessed.
Results: Three patients were included [mean age 56.33 (SD 6.34),
mean HbA1c 81mmol/mol (SD 25.35)]. All had Type 2 diabetes
(mean duration 19.67, SD 11.59) and were treated with insulin. All
patients had documented peripheral neuropathy and presented
with infected, deep heel ulcers. Imaging confirmed calcaneal
osteomyelitis with avulsion fracture. All patients had peripheral
vascular evaluation and optimisation. Two patients underwent
surgical debridement. All required sharp debridement.
Microbiology guided antibiotic therapy, which continues after 5
months. Rest, split removable cast/volar slab use and PRAFO
boots offloaded the wounds. Vacuum pump therapy was applied.
All patients attend podiatry weekly and our multidisciplinary
diabetes foot clinic monthly. Continuous, uncomplicated healing
progress, with healthy granulation tissue formation, has been
observed in all wounds at 5 months. Ankle function is preserved in
all patients. Graded remobilisation has been established. Quality of
life assessment is underway.
Conclusions: Calcaneal osteomyelitis with avulsion fracture in
patients with diabetes and peripheral neuropathy can occur
following injury or spontaneously and has deleterious effects on
locomotion and biomechanics. Currently there is no consensus on
whether to manage these patients conservatively or surgically.
These patients often undergo various level amputations.
Conservative management is a clinically effective method in
achieving wound healing and preservation of function.
P319
Amputation rates in diabetes patients: can
we decrease our rates?
N Sargazi, M Shah, R Zaidi, D Kalathil, D Sharma, P Weston,
T Purewal and A Ahmad
Diabetes and Endocrinology, Royal Liverpool and Broadgreen University
Hospitals, Liverpool, UK
Background and aim: Amputation rates in patients with diabetes

are variable across England. Liverpool was noted to have a high
amputation rate in this cohort. The aim of this audit was to assess
the pre-amputation management of these patients against NICE
guidelines.
Method: All diabetes patients who had an elective amputation
between April 2012 and 2013 were identified. Data were collected
2014 The Authors.

from 36 randomly selected patients. A proforma based on NICE

guidelines (CG87 and CG10) was produced and data were
analysed against NICE standards.
Results: 94.4% of amputations were below knee, of which 75.8%
were single toe and 18.2% were mid-foot amputations. 56% with
foot ulcers were unknown to podiatry. 48.4% with at least one risk
factor (peripheral vascular disease or neuropathy) were not known
to podiatry. HbA1c was measured in 55.6% in the 4 months
leading up to the amputation (mean HbA1c 75 23mmol/mol).
36.1% of subjects were under follow-up with the diabetes team
pre-amputation. Inpatient diabetes reviews occurred in 13.8%
pre-amputation.
Conclusion: More rigorous community-based reviews of glycaemic
control are required, with an early referral to secondary care in high
risk patients. All foot ulcer patients should be under the care of
podiatry and have their HbA1c checked regularly. Inpatients with
foot ulcers should be reviewed by the diabetes team and foot MDT.
Based on the results, an inpatient foot MDT has been established,
consisting of all relevant specialities.
P320
Inpatient diabetes foot screening: a
different approach for the non-specialist
M Pass, K Duncan, E Ricci and R Nayar
Department of Diabetes and Endocrinology, City Hospitals Sunderland NHS
Introduction: Increasing numbers of acute admissions and

complexity of diagnoses lead to prioritising of assessments.
National audits confirm poor neurovascular assessment for
inpatients with diabetes, resulting in a missed opportunity.
Aims: (1) To develop a simple screening tool to identify foot risk
suitable for electronic patient records. (2) To audit the
effectiveness of the screening tool against the gold standard.
Methods: We constructed a four-point assessment tool that could
be completed at first contact by any healthcare professional.
Questions were based on previous history of foot ulcer or
bandaged foot, contact with a vascular surgeon, and
comorbidities (renal disease). We undertook a prospective audit
of current inpatients across a large acute hospital comparing the
gold standard neurovascular assessment with our assessment tool
establishing its validity in identifying low risk patients.
Results: Forty-five patients were audited, average age 69 years,
male 21, insulin treated 19; gold standard identified 26
patients as low risk, 14 patients as increased risk and four
patients as high risk. Our screening tool identified 19 patients as
low risk with five patients scoring yes to one further question
(renal) in isolation. It identified three out of the four high risk
patients (failed to identify age related neuropathy). Previous
vascular surgeon contact was a high predictor of risk rather than
nephrology, with 92% sensitivity 89% specificity in identifying
low risk patients.
Conclusion: Our small audit has demonstrated that a question
based screening tool can identify low risk patients effectively;
however, further work is needed to develop an effective algorithm.
2014 The Authors.

DIABETICMedicine
P321
Benefits and risks of peripherally inserted
central catheters (PICC lines) in the
management of acute and chronic diabetic
foot infection
M Bates, I Alejandro, T Jemmott and ME Edmonds
Diabetic Foot Clinic, Kings College Hospital NHS Foundation Trust, London, UK
Aim: To report the benefits and risks of the use of PICC lines in
the management of diabetic foot infection.
Patients: We studied 31 patients with diabetes(22 males and 9
females; 8 Type 1 and 23 Type 2 diabetes; mean age 59
years 10.6 and duration of diabetes18 years 11.4) who
were treated with intravenous (IV) antibiotics. The reasons for
treatment with PICC line were (1) acute onset infections in the foot
(72%) and (2) long-term treatment of persistent infection (28%).
Osteomyelitis was observed in 24% of the total. PICC lines were
monitored routinely in the diabetic foot clinic for complications.
Results: The median duration of treatment with IV antibiotics
was 9 weeks (2 65 weeks) and the median duration of PICC line
in situ was 9 weeks (range 3 67 weeks). Despite the long
treatment in some cases, this therapy was well tolerated by all
patients with positive feedback. Main advantages of PICC line
therapy reported were being painless and having no
nausea. Common complications noted were PICC line blockage,
skin dermatitis and cellulitis. However, these were overcome and
the overall management of diabetic foot infection with PICC lines
resulted in avoidance of hospital admission in 30% of cases. In
46% of the cases who were admitted to hospital, PICC line
insertion facilitated early discharge. Ulcers healed in 51%,
improved in 43% and 6% remained unhealed.
Conclusions: We believe the use of the PICC line to be a
significant advance in the management of diabetic foot infection.
P322
Non-removable plastic tag: a psychological
tool in improving compliance with DH
Walker
JE Spragg1, J Spratt1, J Armstrong1, G Steel2 and
S Rajbhandari2
1
Podiatry, Lancashire Care NHS Foundation Trust, Preston, UK, 2Medicine,
Lancashire Teaching Hospitals Foundation Trust, Preston, UK
Aim: To study the feasibility of use of a non-removable plastic tag

to improve compliance with the DH Walker. A proportion of
ulcers healed completely within a 12 week study period whilst
using such a device.
Method: This was a prospective study undertaken in a
multidisciplinary foot clinic with non-infected active diabetic foot
ulcers and palpable pulses. After debridement, subjects were fitted
with a DH Walker made irremovable with signed and dated plastic
tag. Subjects were given an emergency action leaflet. Subjects had
foot check and tab resealed within 72h and reviewed in clinic
weekly for 12 weeks or until healing.
Results: Eight subjects, mean age 56 years, were recruited. Foot
deformity was present in four (50%) subjects, two due to Charcot
neuroarthropathy. Two (25%) withdrew, could not tolerate
wearing the walker in bed. One withdrew, infection. Of the
remaining five subjects, two (40%) with chronic ulcers of longer
than 6 months failed to heal by 12 weeks;three (60%) with short
duration ulcers healed within the study period; mean healing time
2 weeks ( 1 week). The tag was intact for all subjects at each visit
123
DIABETICMedicine
apart from two where the walker had been removed due to an
adverse event including infection and abrasion.
Conclusion: Total Contact Cast is the gold standard for
offloading diabetic foot ulcers but disadvantages include skills,
equipment and cost needed for casting and removal difficulties as
opposed to the benefits of a removable walker. The study suggests
that use of plastic tag may help improve compliance by acting as a
psychological barrier to patient removal.
P323
The pathway to amputation in patients with
diabetes; missed opportunities for early
intervention and implications on clinical
management
MA Ali1, KS Higgins2, DR Webb2,3, MJ Davies2,3, E Choke1,
RD Sayers1 and MJ Bown1
1
Department of Cardiovascular Sciences, University of Leicester, Leicester,
UK, 2Department of Diabetes and Endocrine Medicine, University Hospitals of
Leicester, Leicester, UK, 3Diabetes Research Centre, University of Leicester,
Leicester, UK
Objectives: To determine whether there was a window of

opportunity to intervene towards preventing major lower limb
amputation in patients with diabetic foot disease.
Methods: This was a retrospective study of 213 patients with
diabetes who underwent major lower limb amputation in a tertiary
vascular unit between 2000 and 2012.
Results: A total of 234 amputations were performed, 87 (37.2%)
for isolated diabetic foot sepsis without arterial insufficiency (75
below knee, 12 above knee). The median time from first symptom
to amputation in this group was 95 days (range 47 138 days).
Only 12.6% of patients had been seen by their primary care team
prior to their admission that led to amputation. The median time
from onset of symptoms to primary care review was 74 days (range
65119 days), and the median time between first primary care
review and amputations was 24 days (range 7 51 days).
Twenty-three patients (26.4%) with isolated foot sepsis had been
seen in the multidisciplinary foot clinic during the 6 months prior
to amputation and only four (4.6%) since the onset of their latest
symptoms. In those seen within the 6 months prior to amputation,
65.2% had presented at a point where their limb had become
unsalvageable and 30.4% refused treatment from their primary
care team (referral/admission, antibiotics, debridement) at earlier
stages of their clinical course.
Conclusions: There is a need to improve diabetic foot services
such as early identification and adequate antimicrobial treatment
or early surgical intervention, which may prevent progression to
major limb amputations.
P324
Case series of patients with diabetic foot
ulceration in Galway University Hospitals
R Mannion1, E Young2, D Gallagher3 and SF Dinneen1,4
1
School of Medicine, National University of Ireland, Galway, Ireland, 2Merlin
Park Podiatry Clinic, Galway University Hospitals, Galway, Ireland, 3Medical
Assessment Unit, Galway University Hospitals, Galway, Ireland, 4Department
of Diabetes and Endocrinology, Galway University Hospitals, Galway, Ireland
Aim: This audit aimed to assemble and describe all patients with
active diabetic foot ulceration seen in Galway University Hospitals
(GUH) between April 2012 and March 2013.
Methods: We undertook a retrospective review of clinical and
laboratory records among patients attending the Diabetes Day
124
Centre, Merlin Park Podiatry Clinic (MPPC), and hospital

inpatients over the 2 years prior to ulceration. Information was
sourced from the DIAMOND Clinical Information System, the
MPPC chart, hospital inpatient enquiry records and
electronic-discharge summaries.
Results: Of 129 patients, 16 (12%) had Type 1 diabetes and 113
(88%) Type 2 diabetes. Ulcer type was neuropathic (41%),
neuro-ischaemic (19%) or ischaemic (16%). Foot deformity and
neuropathy was present in 22% and 59% respectively. Peripheral
vascular disease was documented in 32%. 12% were current
smokers and 34% past smokers. 33% had body mass index >30kg/
m2. HbA1c was not documented in 46% and 41% had >53mmol/
mol. Blood pressure was >140/90mm Hg in 20%. 70% were
taking antihypertensive medication. 35% had retinopathy since
last examined. Stage 3 or further stages of chronic kidney disease
were 22%. Microalbuminuria was reported in 19 patients. 27%
had ischaemic heart disease. 48% of patients were taking insulin,
and 66% oral hypogylcaemics, 70% antiplatelet drugs, 8%
warfarin and 73% statins.
Conclusion: This is the first report detailing diabetic foot
ulceration in GUH. The amount of missing data highlights the
need for better documentation in this patient group. This report
will serve as a baseline for future activity in this area of practice.
P325
Evaluation of 18 month mortality for
patients seen within a multidisciplinary
diabetic foot clinic
CA Manu1, OG Mustafa2, NL Petrova1, M Simmgen1 and
ME Edmonds1
1
Diabetes Foot Department, Kings College Hospital, London, UK, 2Diabetes
Department, Kings College Hospital, London, UK
Introduction and objective: Despite the improvement in limb

salvage within the multidisciplinary diabetic foot clinic (DFC),
patients continue to have high mortality rates. We evaluated the
mortality rate for our DFC cohort.
Method: This was a retrospective evaluation of mortality, with
demographics analysed with SPSS and descriptive data from death
certificates.
Results: Clinical notes of 563 patients presenting over a 6 week
period were reviewed: Type 1 diabetes 19.6%; Type 2 diabetes
74.6%, and patients without diabetes 5.8%. There were 56 (9.4%)
deaths over the subsequent 18 months, all in patients with diabetes.
The deceased patients were significantly older compared with the
alive cohort [71 years (64 78)vs 64 years (55 72.8), p < 0.001]
and more likely to have Type 2 diabetes (89.3% vs 78%,
p = 0.031). Duration of diabetes was 21 years (13.5 29.5) vs
18 years (10 27), p = 0.051. There was no difference in gender
distribution (p = 0.269) and diabetes control (p = 0.935). Analysis
of 24 death certificates available showed 21 (87%) deaths in
hospital. Primary causes of death were split almost evenly between
cardiovascular diseases and infections (mainly pneumonia). All
cases requiring post-mortem examination (n = 5) confirmed
myocardial infarction. Overall, diabetic foot disease as a
contributing factor was only documented in 8%.
Conclusion: There is significant under-reporting of diabetic foot
disease. A significant proportion of deaths occurred in hospital.
Nearly 50% of deaths were due to infections; thus the need to treat
all forms of infection aggressively in these patients. The high
incidence of death from cardiovascular disease should lead
to changes of care within the DFC to target cardiovascular
risk optimisation.
2014 The Authors.

P326
Delays in referral of patients with diabetes
with active foot disease to secondary care
MA Ali1, A Aber1, R Berrington2, M Quinn2, KS Higgins2,
MJ Davies2,3, E Choke1, RD Sayers1 and MJ Bown1
1
Leicester, UK
Objectives: This study investigated the time course of delayed

referrals for patients with diabetic foot disease to a regional
secondary care service and determined if delayed referrals were
associated with poor outcomes.
Methods: We did a1 year prospective audit of patients with
diabetes with active foot disease referred late to our
multidisciplinary foot clinic.
Results: Over a period of a year, 20 patients with active foot disease
were identified as delayed new referrals. The median age at time of
referral was 74 years (range 63 85 years), with 10 (50%) being
males. Duration of symptoms varied from 4 to 36 weeks (median 18
weeks). The aetiologies at the time of diagnosis were as follows:
non-infected ulcer (n = 4), infected ulcer and foot sepsis (n = 8),
gangrene (n = 3), osteomyelitis (n = 4) and Charcots arthropathy
(n = 1). In 16 patients (80%) the median time from first symptom to
referral was 25 weeks and this was due to patients reluctance at
attending the clinic or lack of understanding of the significance of
their foot disease. A minor amputation was required in five (25%)
patients due to foot sepsis or osteomyelitis, while one patient needed
an above knee amputation for a non-revascularisable limb that
presented with toe gangrene and rest pain.
Conclusions: Delayed presentations of patients with diabetic foot
disease were common and were associated with poor outcomes.
Public health education on the importance of seeking early treatment
for diabetic foot disease may improve outcomes in these groups.
P327
The impact of a streamlined pathway on
vascular referrals for patients with diabetic
foot disease
MA Ali1, KS Higgins2, M Quinn2, MJ Davies3, E Choke1,
RD Sayers1 and MJ Bown1
1
Leicester, UK
Objectives: Delays in vascular referrals for patients with diabetic

foot disease are a frequent cause of major limb amputations. This
study assessed the impact of a streamlined pathway on vascular
referrals in patients with diabetic foot disease.
Methods: Between October 2012 and July 2013, all vascular
referral for patients with diabetic foot disease from the
multidisciplinary clinic were allocated to either a streamlined or
standard referral pathway according to clinic of referral. In the
former group, fast-tracking protocol was used to reduce
administrative delays in patient referral. Intervals between clinic
appointments and the outcomes were analysed.
Results: Twenty-three patients were recruited, with 13 (57%)
allocated to the streamlined pathway. There was no difference in
age, sex, pre-existing peripheral neuropathy or median duration of
2014 The Authors.

DIABETICMedicine
symptoms between the groups. There was a substantial reduction

in the interval between foot clinic appointment and the vascular
outpatient appointment (median 16 days, range 1027) compared
with the control (median 29 days, range 24 45). In the control
group, three patients were discharged after initial appointment,
four underwent endovascular intervention, and one patient
required a major amputation for non-revascularisable disease,
while two patients were allowed conservative management and
follow-up. In comparison, in the streamlined group, four were
discharged, six underwent endovascular intervention, one patient
required a major amputation, and the remainder were kept under
surveillance.
Conclusions: The introduction of a streamlined pathway
significantly reduces delays in vascular referrals for patients with
diabetic foot disease. Whether this translates to improved
outcomes remains to be evaluated.
P328
Follow-up audit: inpatient management of
diabetic foot problems in patients who have
subsequently undergone a major lower limb
amputation and implications for the
multidisciplinary diabetic foot team
C Rice, KS Higgins and A Gallagher
Diabetes and Endocrinology, University Hospital Leicester, Leicester, UK
Aims: The following changes were made to our service in response

to the previous audit (2010 2012): appointment of vascular surgeon
(specialist interest in diabetes); shared learning for the
multidisciplinary team (MDT); weekly MDT diabetic foot ward
rounds; weekly diabetes review of vascular ward patients; diabetic
foot antimicrobial guidelines. We aimed to complete the audit cycle.
Method: Twenty-two patients were audited between 2012 and
2013.
Results: Mean age 70years (53 86years), 16 male (73%), six
female (27%). Risk factors: neuropathy n = 3(14%), peripheral
vascular disease n = 7(32%), both n = 4(18%), not documented
n = 8(36%). All cases were admitted with active foot disease as the
primary cause for admission. Fifteen were admitted with infected
foot ulcers (68%), seven with sepsis (32%). All were classified as
severe infections. In four cases there was no documented
assessment for sepsis. Complete assessment for sepsis was evident
in 12 cases. All cases were referred to the specialist team, 18
reviewed within 24h. Fifty percent were managed exclusively by
the vascular team. Four cases documented that both feet were fully
exposed on examination (29%); six cases assessed for neuropathy
(27%); 18 assessed for PVD (82%). Seventeen (77%) received
empirical antibiotic therapy. Fifteen (68%) received antibiotics
according to guidelines. Fifteen (68%) with severe infections
received immediate intravenous antibiotics. Antibiotic therapy
and duration of treatment was informed by microbiological results
and clinical assessment was documented in four patients (18%).
Four patients were reviewed for pressure-relieving devices. Two
patients were upset by prosthetic review arranged prior to consent
for amputation. Community follow-up was arranged for three
patients (14%).
Conclusions: This audit identifies a significant improvement in
care, compared with previous results. There remains compelling
evidence to continue to review and improve our service.
125
DIABETICMedicine
Clinical care and other categories posters: hypoglycaemia

P329
P333
Temporal patterns of hypoglycaemia in 11

UK NHS Trusts reveals more frequent
nocturnal hypoglycaemia particularly in
sulphonylurea treated patients: the RHAID
(Reduction of Hypoglycaemia Amongst
Inpatients with Diabetes) study
Time to repeat capillary blood glucose

measurement following inpatient
hypoglycaemia as an indicator of quality of
care shows poor concordance with
guidelines, and can be improved by an
intervention (NHS Scotland Quality
Improvement Hub Think Glucose pilot):
improvement persists beyond cessation of
intervention
R Rajendran, C Kerry and G Rayman

Diabetes Centre, Ipswich Hospital NHS Trust, Ipswich, UK
Refer to Oral A69
P330
CAR Sainsbury1, A Mohammed1, A Monaghan2, B Kennon3,

C Perry4 and GC Jones1
1
Department
Department
3
Department
4
Department
of
of
of
of
Diabetes, Gartnavel General Hospital, Glasgow, UK,

Vascular Surgery, Western Infirmary, Glasgow, UK,
Diabetes, Southern General Hospital, Glasgow, UK,
Endocrinology, Western Infirmary, Glasgow, UK
Diabetes patients requiring ambulance

assistance for severe hypoglycaemia have
twice the mortality rate of the general
diabetes clinic: identification of possible
risk factors
J Elliott1,2, S Beveridge1, V Gordon1, C James3, T Dodd1 and

A Scott1
P334
Coefficient of variation from continuous
glucose monitoring is associated with the
frequency of hypoglycaemia in Type 1
diabetes complicated by impaired
awareness of hypoglycaemia: baseline
analysis of the HypoCOMPaSS study group
Department of Diabetes and Endocrinology, Sheffield Teaching Hospitals,

Sheffield, UK, 2Department of Human Metabolism, University of Sheffield,
Sheffield, UK, 3Department of Clinical Pathways, Yorkshire Ambulance
Service, Wakefield, UK
Refer to Oral A70
P331
Impact of hypoglycaemia on length of stay,
time to readmission, and mortality in
inpatients with diabetes
GC Jones and CAR Sainsbury
Department of Diabetes, Gartnavel General Hospital, Glasgow, UK
Refer to Oral A71
P332
One year mortality in individuals with
diabetes following severe hypoglycaemia
requiring emergency services intervention
FR Elwen1, AC Huskinson1, MJ Bottomley1, S Heller2, C James3,
PD Baxter4, L Clapham5 and RA Ajjan1
1
Leeds Institute of Genetics, Health and Therapeutics, University of Leeds,
Leeds, UK, 2Academic Unit of Diabetes, Endocrinology and Metabolism,
University of Sheffield, Sheffield, UK, 3Yorkshire Ambulance Service,
Yorkshire, UK, 4Division of Biostatistics, Leeds Institute of Genetics, Health
and Therapeutics, Leeds, UK, 5Leeds Teaching Hospital Trust, Leeds, UK
Refer to Oral A72
126
Refer to Oral A73
HK Tan1, SA Little2, L Leelarathna3, E Walkinshaw4,

A Lubina-Solomon4, D Kerr5, S Heller4, ML Evans3, JA Shaw2
and D Flanagan1
1
Diabetes and Endocrinology, Derriford Hospital, Plymouth, UK, 2Institute of
Cellular Medicine, Newcastle University, Newcastle-upon-Tyne, UK,
3
Wellcome Trust, Medical Research Council, Institute of Metabolic Science,
Metabolic Research Laboratories, University of Cambridge, Cambridge, UK,
4
School of Medicine and Biomedical Sciences, University of Sheffield,
Sheffield, UK, 5Centre of Postgraduate Medical Research and Education,
Bournemouth University, Bournemouth, UK
Refer to Oral A74
P335
A prospective study of hospital admissions
among patients with Type 2 diabetes
admitted with hypoglycaemia
M Munu, C Jairam and A Dornhorst
Department of Endocrinology and Diabetes, Imperial College Healthcare NHS
Trust, London, UK
Objective: To examine the contributing risk factors for hospital

admission from hypoglycaemia among patients with diabetes.
Methods: Between April 2012 and February 2013 the diabetic
nursing team saw all hospital admissions from hypoglycaemia,
recoding demographic details, comorbidities and medications.
Results: During this 10 month prospective study, 25 patients were
admitted with hypoglycaemia, two with Type 1 and 23 with Type
2 diabetes. Of the 23 patients with Type 2 diabetes, 12 were
treated with sulphonylurea gliclazide without any insulin and 11
were taking insulin, two with gliclazide and one with
2014 The Authors.

pioglitazone. The median age (79 years, range 20 93) of patients

on gliclazide without insulin was similar (74 years,range 55 91) to
those taking insulin. Of the nine patients above 80 years old six
were taking gliclazide without insulin. All 23 patients with Type 2
diabetes had comorbidities that contributed to their hypoglycaemia
and length of stay in hospital. Comorbidities included seven with
dementia (five on gliclazide without insulin), four with difficulties
around eating, (two on gliclazide without insulin), four took
alcohol in excess (two on gliclazide without insulin) and two had
renal disease (one on gliclazide without insulin). The median
length of inpatient stay was 6 (1 42) days among those on
gliclazide without insulin vs 3 (1 26) days for those on insulin.
Conclusions: If decreases in hospital admissions from
hypoglycaemia are to be achieved there needs to be greater
awareness of hypoglycaemia with gliclazide prescribing in elderly
vulnerable patients with Type 2 diabetes.
P336
Hospitalisation costs arising from
hypoglycaemic episodes in England
ML Evans1, ML Wolden2, BL Thorsted2, P McEwan3 and
JL Jacobsen4
1
Llandough Hospital, Cardiff, UK, 2Novo Nordisk A/S, Sborg, Denmark,

Swansea Centre for Health Economics, Swansea University, Swansea, UK,
4
Statcon ApS, Kokkedal, Denmark
3
Aims: Hypoglycaemia is one of the most common and serious side

effects of insulin therapy, leading to hospitalisation in some cases.
This analysis aimed to estimate the mean length of stay in hospital,
and total per patient expenditure resulting per hypoglycaemic
episode requiring hospitalisation in patients with Type 1 or Type 2
diabetes.
Methods: The cohort consisted of all Type 1 diabetes and Type 2
diabetes patients, aged 18 years, with a diagnosis of diabetes
between 1 January 2002 and 30 October 2011 in the CPRD
database, who had initiated insulin treatment and had a recording
of hypoglycaemia in the UK Hospital Episodes Statistics database
in the same period. The in-hospital expenses are based on the cost
per bed-day and emergency ambulance transportation. The unit
cost per bed-day, based on a weighted average, was 158. The cost
of ambulance transportation was 246; as 66% were admitted via
ambulance, a fixed unit cost of 162 was added to each admission.
Results: In total, 1,131 (Type 1 diabetes 409, Type 2 diabetes
722) patients contributed to the analysis. Baseline demographics
(mean SD) were age 71.2 17.1 years, body mass index
27.5 5.6kg/m2, HbA1c 9.5 2.3%. The mean length of stay
for patients with Type 1 diabetes and Type 2 diabetes was
10.1 19.9 and 12.0 20.8 days respectively. Therefore, mean
total expenditure per hospital admission for hypoglycaemia was
1,758 for patients with Type 1 diabetes and 2,058 for patients
Summary: Hospitalisation following a hypoglycaemic episode
leads to significant expenditure and use of healthcare resources.
Reducing the incidence of hypoglycaemic episodes is important
from both a medical and economic perspective.
2014 The Authors.

DIABETICMedicine
P337
Are hypoglycaemia treatment boxes
(hypoboxes) being used correctly to
improve hypoglycaemia treatment?
EA Hackett1, S Rotimi1,2 and K Higgins3
1
Pharmacy Department, University Hospitals Leicester, Leicester, UK,

Pharmacy School, University of Manchester, Manchester, UK, 3Diabetes and
Endocrinology, University Hospitals Leicester, Leicester, UK
2
Aims: Following the introduction of hypoboxes across our Trust,

the aims were to undertake an audit of use and evaluate whether
their use had resulted in improved care.
Methods: All wards (excluding paediatrics and intensive care)
were audited. Hypoboxes were located and the contents examined.
Patients medical records were screened for capillary blood glucose
readings of less than 4.0mml/l. Documentation of hypoglycaemia
management was reviewed.
Results: Hypoboxes were available on 52/60 wards (87%). Of
these only 3/52 (6%) contained the correct contents. One hundred
and eighty-one inpatients with diabetes were identified. Of these,
40/181 (22%) patients had at least one hypoglycaemic episode
(range 1 7 hypos per patient). Alarmingly 7.5% had experienced
7 hypos during their stay. One hundred and seventeen episodes
of hypoglycaemia were recorded. Hypoglycaemia treatment
was documented as administered in 65/117 (56%) with only
2/117 (1.7%) treatments in accordance with guidance. The
administration of longer-acting carbohydrates after initial
treatment with rapid-acting carbohydrates was extremely poorly
recorded.
Conclusion: The replenishing of hypoboxes was poor. There was
improvement in the documentation of treatment of hypoglycaemia
compared with audit data from 2008. Documentation of treatment
using long-acting carbohydrates was poor although we believe this
step is actually being administered as a meal, but not being
recorded as a treatment. Audit recommendations are to re-educate
staff on correct replenishment of hypoboxes and stepwise
documentation of treatment and to audit the administration of
long-acting carbohydrates separately.
P338
Admissions to emergency department with
hypoglycaemia are associated with
increased mortality in patients with Type 2
diabetes
R Rajendran1, D Hodgkinson2, K Lentini3 and G Rayman1
1
Diabetes Centre, Ipswich Hospital NHS Trust, Ipswich, UK, 2Accident and
Emergency Department, Ipswich Hospital NHS Trust, Ipswich, UK,
3
Information and Development, East of England Ambulance Service NHS
Trust, Cambourne, UK
Aim: To report characteristics of patients with diabetes who

require emergency department (ED) services with hypoglycaemia
and their outcomes.
Method: Electronic data on all attendances with hypoglycaemia
to a single district hospitals ED was obtained from three sources:
ambulance records, hospital episode statistics and patient
administration records between 1 April 2012 and 31 March 2013.
Results: There were 165 hypoglycaemia attendances amongst 132
patients with diabetes [Type 159 episodes in 43 patients, Type 2
106 episodes in 89 patients (therapy: 54 insulin, 35 sulphonylureas,
11 both, 6 others)]. Patients with Type 2 vs Type 1 diabetes were
significantly older (median age 79 vs 61years, p = 0.0001) with no
significant difference in HbA1c (median 62 vs 68mmol/mol,
127
DIABETICMedicine
p = 0.065). 102 (61.8%) attendances required admission. In

atleast 17(10.3%) attendances, hypoglycaemia resulted in serious
harm. 36(27.3%) patients died since the start of study (median
time to death after discharge 75 days, n = 24), of whom 12 died
during hospitalisation. Mortality following first ED attendance
during the study period was significantly higher in patients with
Type 2 vs Type 1 diabetes (36% vs 9.3%, p = 0.0015). There was
no significant difference in post discharge survival in patients with
Type 2 diabetes with or without coexisting reasons for
hospitalisation (78.8% vs 75.9%, p = 0.7915).
Conclusions: More patients with Type 2 diabetes present to ED
with hypoglycaemia and have greater mortality than those with
Type 1 diabetes. This increased mortality is irrespective of whether
or not they have coexistent reasons for hospitalisation. Of note
33% of hypoglycaemia attendances in patients with Type 2
diabetes are related to sulphonylureas.
P339
Serious harm from inpatient
hypoglycaemia: a survey of hospitals in the
UK
G Rayman and R Rajendran
Aim: To estimate the incidence of severe harm amongst inpatients

with diabetes as a consequence of inpatient hypoglycaemia.
Method: An anonymised questionnaire was emailed to lead
organisers for the 142 acute NHS Trusts that contributed to
NaDIA2012. ABCD and Diabetes Inpatient Specialist Nurse
Group also helped publicise the survey. Each diabetes team was
asked to collectively recall and report any serious harm from
inpatient hypoglycaemia in the previous year. A single member
returned the questionnaire to prevent duplicate responses.
Eighty-threetrusts agreed to participate. Serious harm was
defined as death, cardiac or cerebral event or fall resulting in
permanent physical injury or fracture.
Results: Disappointingly, only 41 of 83 trusts replied. Of these,
only 28 (68.3%) were confident that robust methods existed in
their trust to ensure all such events were reported and only 23
(56.1%) were confident that they were reported to the diabetes
team. Despite these reporting concerns, the retrospective nature
and reliance on recall, 12 serious events were reported from nine
trusts: three deaths, two permanent cerebral damage, two
successfully resuscitated cardiac arrests, three seizuresand two
undefined events. Insulin therapy was implicated in 10.
Importantly three events with two deaths occurred in patients
who had received insulin/dextrose to correct hyperkalaemia, only
one of whom had diabetes.
Conclusions: Twelve patients in 41 trusts annually suffering
serious harm from inpatient hypoglycaemia is alarming. This may
be
the
tip
of
the
iceberg
considering
potential
under-reporting. Robust reporting mechanisms are required to
determine the full extent of this serious preventable harm.
128
P340
Impact of hypoglycaemia on people with
diabetes, family members and healthcare
professionals in the UK: results from the
Diabetes Attitudes, Wishes and Needs
(DAWN2TM) study
NM Munro1, K Barnard2, G Hall3 and RIG Holt2
1
Postgraduate Medical School, University of Surrey, Guildford, UK, 2Faculty of

Medicine, University of Southampton, Southampton, UK, 3Primary Care
Diabetes Service, Solent Community Trust Diabetes Service, Portsmouth, UK
Aims: To analyse the UK data from the DAWN2 study to examine

the impact of hypoglycaemia on people with diabetes (PWD),
family members (FMs) and healthcare professionals (HCPs).
Methods: Between 6 March 2012 and 22 May 2012, 500 PWD
(Type 1 diabetes, 81; Type 2 diabetes, 419), 122 FMs (of people
receiving insulin, n = 46; receiving non-insulin medication,
n = 76) and 281 HCPs in the UK completed the DAWN2 survey.
Results: Severe hypoglycaemic episodes were reported with a
mean frequency of 1.63.7 (PWD) and 1.24.5 (FMs) in the 12
months prior to the survey. The largest proportion (45%61%) of
FMs assisted with hypoglycaemic episodes less often than once a
month/once a month, with 33%36% reporting they felt
extremely/very confident in doing so. While 32%58% of
PWD were very worried about the risk of hypoglycaemia, 14%
54% experienced self-treated episodes several times a month or
at least once a week in the 12months prior to the survey; for FMs
thinking about the PWD they live with, these proportions were
33%70% and 25%48%. Of PWD, 32%61% fully/mainly
agreed that taking insulin increases the risk of hypoglycaemia.
Overall, 78%91% of HCPs said reduced risk of hypoglycaemia
would be a helpful improvement in current diabetes medications.
Conclusions: Similar proportions of PWD and FMs worried
about hypoglycaemia; a notable percentage of PWD had
experienced self-treated hypoglycaemia and severe episodes
occurred relatively frequently. While a third of FMs felt
confident assisting with severe hypoglycaemia, more support
could be beneficial. Both PWD and HCPs acknowledged the risk
of hypoglycaemia associated with insulin treatment.
P341
Decreased hypoglycaemia and improved
wellbeing in adults treated with continuous
subcutaneous insulin infusion (CSII)
A Al-Janabi1, MA Jones1, S Ng1, K Davenport1, C Ward1 and
ML Evans1,2
1
Wolfson Diabetes Endocrine Clinic, Cambridge University Hospitals NHS
Foundation Trust, Cambridge, UK, 2Wellcome Trust , Medical Research Council,
Institute of Metabolic Science, University of Cambridge, Cambridge, UK
Aims: When using insulin pumps (CSII), clinical success is often

judged by HbA1c. For patients themselves, the burden of
hypoglycaemia and wellbeing are often more important
outcomes. We examined whether these non-HbA1c outcomes
improved in adults starting on CSII.
Methods: As part of an ongoing clinical audit, adult patients
complete validated questionnaires prior to commencing CSII and
annually thereafter. We assessed the burden of hypoglycaemia
using the Hypoglycaemia Fear Survey (HFS) and the Clarke
hypoglycaemia awareness questionnaire (which also includes an
assessment of frequency) and wellbeing by the Problem Areas in
Diabetes (PAID) and Hospital Anxiety and Depression Score
2014 The Authors.

(HADS). Shown here, we analysed each questionnaire

separately where paired data were completed at baseline and at
6 18 months after commencing CSII. Data are shown as
mean SEM.
Results: There were significant improvements in all parameters
measured. Annual frequency of severe hypoglycaemia (4.6 0.6
to 2.0 0.5, p < 0.001, n = 79), fear of hypoglycaemia (HFS
68 2 to 58 2, p < 0.001, n = 73) and hypoglycaemia
awareness [Clarke score (higher score indicates unawareness)
1.9 0.2 to 1.2 0.2, p < 0.05, n = 77] improved. Similarly
wellbeing scores improved (PAID 24 2 to 13 2, p < 0.001,
n = 45; HADS total 11.0 0.9 to 8.6 0.9, p < 0.001, n = 50).
Conclusion: In addition to previouslydocumented improvements
in HbA1c, CSII therapy in adults is associated with marked
improvements in outcomes important for patients emotional
wellbeing, awareness of hypoglycaemia and frequency of
hypoglycaemic episodes.
P342
A comparison of methods used to assess
impaired awareness of hypoglycaemia in
Type 1 Diabetes
SS Hussain, K Ghandi and A Dornhorst
Department of Diabetes, Endocrinology and Metabolism, Imperial College
London, London, UK
Aims: We report a prospective outpatient study comparing three

validated methods to assess prevalence and characteristics of
patients with impaired awareness of hypoglycaemia (IAH) in an
adult outpatient cohort with Type 1 diabetes.
Methods: IAH and normal awareness (NA) was assessed as per
the Clarke, Gold and Pedersen questionnaire based scoring
methods. Edinburgh hypoglycaemia symptom score was used to
assess symptom severity. Baseline demographic and clinical
information was extracted from clinical records.
Results: A hundred adult patients with Type 1 diabetes completed
the study. Mean SEM for age (years) 44 1.4; duration
of diabetes 21.8 1.4; glycated haemoglobin (HbA1c)
69.6 2mmol/mol. The prevalence of IAH identified by the
Clarke, Gold and Pedersen methods was 18%, 19% and 61%
respectively. Spearmans correlation coefficients between Pedersen
and Clarke and Pedersen and Gold were Rs = 0.555 (p < 0.001)
and Rs = 0.645 (p < 0.001) respectively. Only a moderate
association between Clarke and Gold scoring methods was
observed. Further analysis highlighted that the three methods
defined IAH in different overlapping subsets of patients. Detailed
findings of this study revealed differences in established clinical
features of IAH between patients defined as having IAH by the
Clarke, Gold and Pedersen methods.
Conclusion: In summary this study highlights non-equivalence of
the Clarke, Gold and Pedersen methods in defining IAH and
emphasises caution in their utility in the clinical and research
settings.
DIABETICMedicine
P343
Non-specific symptoms associated with
hypoglycaemia in the elderly are common in
patients treated with metformin only, as
well as those treated with insulin or
sulphonylureas
SV Hope1, PJ Taylor2, BM Shields1, RA Oram1, AJ Chakera1,
WD Strain1 and AT Hattersley1
1
National Institute for Health Research, Exeter Clinical Research Facility,
University of Exeter, Exeter, UK, 2Axminster Medical Practice, NHS,
Axminster, UK
Background: Hypoglycaemia recognition is difficult in older

patients. We aimed to explore if non-specific symptoms
associated with hypoglycaemia in older patients were associated
with known risk factors for hypoglycaemia (treatment type,
HbA1c and recognised hypoglycaemia episodes).
Methods: We collected data from a general practice database on
patients aged over 65, treated with insulin (n = 65),
sulphonylureas (n = 54) or metformin only (n = 41) between
2012 and 2013. We recorded episodes of hypoglycaemia,
documented hypo clues from consultation records, and assessed
association with treatment type, HbA1c and recognised
hypoglycaemia episodes.
Results: Patients in the three treatment groups were of similar age
(74.5 6.7 years in insulin-treated patients, 76.9 7.5 years in
sulphonylurea-treated patients and 77.1 7.9 years in
metformin-only-treated patients (p = 0.45). Age was not associated
with frequency of documented hypoglycaemia episodes (p = 0.09)
but was associated with frequency of hypo clue consultations
(p = 0.03). Numbers of hypo clue consultations were similar
between treatment groups: 1.94 consultations/insulin-treated
patient/year, 1.46/sulphonylurea-treated patient/year and 1.54/
metformin-only-treated patient/year (p = 0.7). In contrast there was
a difference in frequency of recognised hypoglycaemia episodes
between treatment groups: 1.0 episode/ insulin-treated patient/
year, 0.2/sulphonylurea-treated patient/yearand 0.1/metforminonly-treated patient/year (p = 0.001). Neither frequency of hypo
clue (p = 0.7) or documented hypoglycaemia episodes (p = 0.4)
differed across HbA1c quintiles within treatment groups.
Conclusions: Elderly patients who have hypoglycaemia may
present with non-specific hypo clue symptoms, but a high
frequency of these symptoms is seen in patients treated with
metformin only, who are unlikely to have significant
hypoglycaemia. Therefore in elderly patients non-specific features
of hypoglycaemia are non-discriminatory for flagging up potential
hypoglycaemia.
P344
The Portsmouth hypoglycaemic pathway:
The Hypo Hotline
JL Buchanan, A Morcombe, S Daltry and PS Kar
Diabetes Centre, Portsmouth Hospitals Trust, Portsmouth, UK
Aim: The diabetes inpatient specialist team (DIST) had noticed

significant recurrent hypoglycemia admissions to the Trust.
Analysis suggested some admissions due to inappropriate
application of quality of framework targets, as well as lack of
cascading of information from ambulance crew to primary care or
specialist teams.
Method: A pathway (The Hypo Hotline)was set up which would
involve in summary a three-pronged approach: education at (a)
each individual GP surgery level using the Super Six model and
2014 The Authors.

129
DIABETICMedicine
(b) local residential homes pathway for paramedics to highlight all

patients having hypos to the DIST; DIST to give specific advice on
any patients flagged by the pathway (admitted or not).
Results: Since launch of the Super Six model coupled with the
hypoglycaemia hotline pathway, there has been a reduction of
hypoglycaemic admissions by 16% (116 to 97). Using the
pathway, there have been 130 contacts from the South Central
Ambulance Trust to the DIST. 100% have been answered within
one working day. All users of the system (paramedics and patients)
have expressed high satisfaction with the pathway allowing easy

access to specialist care.
Summary: The joint working of the community model, which hinges
on education and support for primary care (where specific sessions on
hypoglycaemia have been conducted) along with an easy access for
paramedics for patients needing treatment for severe hypoglycaemia,
has helped to reduce inappropriate admissions. This is a pathway
which has been easy to set up and can be emulated in other centres to
aid in reduction of such debilitating events for patients.
Clinical care and other categories posters: inpatient care

P345
A shorter length of stay, but at what
benefit? Changes in the management and
characteristics of patients presenting with
diabetic ketoacidosis to Auckland City
Hospital between 1988 and 2011
G Braatvedt1,2,3, J Wallace2, I Wallace1,2, H Britton2,
A Tekiteki2 and M Khanolkar1,2
1
Auckland Diabetes Centre, Greenlane Clinical Centre, Auckland, New
Zealand, 2Department of Medicine, Auckland City Hospital, Auckland, New
Zealand, 3School of Medicine, University of Auckland, Auckland, New
Zealand
Refer to Oral A8
P346
An audit on inpatient insulin prescription
and administration at Mater Dei Teaching
Hospital, Malta
doses. Complications from errors in insulin administration were

found in six patients and these all involved hyperglycaemia.
Conclusion: It is evident from this audit that more awareness
needs to be done with regard to safe insulin prescription and
administration. This will be implemented by giving lectures to
doctors and nurses with the aim of issuing local hospital guidelines.
P347
Improving the quality of inpatient diabetes
care: the importance of understanding the
whole care system and its complexity
A Cox1, R Lim2, J Anderson3 and R Malik1
1
Kings Diabetes Centre, Kings College Hospital, London, UK, 2Reading
School of Pharmacy, University of Reading, Reading, UK, 3Florence
Nightingale School of Nursing, Kings College Hospital, London, UK
Background: Errors in insulin prescription and administration

can lead to considerable patient morbidity and mortality. The
commonest errors include incorrect insulin type, dose, timing in
relation to meals, use of abbreviations and illegible handwriting.
Aim: This audit assesses whether insulin is being prescribed
accurately by doctors and administered safely by nursing staff.
Method: The medical notes and treatment charts of all patients
on insulin in the medical wards over a period of 4 weeks were
reviewed.
Results: A total of 105 patients were recruited. The mean age was
69 years. 91% of patients had the name of the insulin written
legibly but only 24% of patients had units written in full. Of
these two patients had U resembling a zero and in five patients it
was not written at all. Only 10 patients had the insulin prescribed
in relation to meals and nine patients had appropriate prescription
of supplementary actrapid in relation to meals. Blood glucose
monitoring was being carried out in all patients but only 39% of
these had readings checked pre-meals and at 10pm. 31% of
patients had appropriate administration of insulin in relation to
meals and four patients had inappropriate omission of fixed insulin
Objectives: Improving the quality of inpatient diabetes care is a

priority for many organisations. Most improvement efforts are not
underpinned by an analysis of the whole care system and so risk
only targeting part of the problem. The study objective was to
model the relationship between inpatient diabetes care delivery and
the wider general medical work system to guide future quality
improvement work.
Methods: Semi-structured interviews were conducted with 11
diabetes specialists and 19 non-specialists in a large urban teaching
hospital in the UK. The model was produced by analysis of
qualitative interview data and refined in diabetes specialist
workshops. The model shows the purpose of the work system,
its values and performance measures, functions and infrastructure.
Results: The work systems purpose is to provide quality inpatient
diabetes care which is patient-centred, effective, efficient, safe,
timely and equitable. The analysis shows that there are 14 major
functions that are needed to achieve quality care. The model shows
interactions between these functions and the complexity of work
organisation. The discrete functions identified, e.g. nutrition,
glucose monitoring, can become the focus of quality
improvement modules. The model can be used to trace how an
improvement in one area could affect others. It thus provides a
map for understanding interdependences between functions of the
system.
Conclusions: Holistic understanding of the whole care system is
essential for effective quality improvement. A system model can
identify linked functions that are potential targets for improvement
and can guide system-wide quality improvement efforts.
130
2014 The Authors.

R Agius, A Moore, C Attard, L Mercieca, S Azzopardi, F Uman,

R Shaban and M Cachia
Diabetes and Endocrine Centre, Mater Dei Hosptial, Msida, Malta
P348
National Diabetes Inpatient Audit 2012
(NaDIA 2012) cohort of a large teaching
hospital: one year follow-up
KD Dawoud1, RH Heath1, IR Ray2, AM Murphy3 and SG Ghosh3
1
Medical School, University of Birmingham, Birmingham, UK, 2Department of
Economics, University of Birmingham, Birmingham, UK, 3Department of
Diabetes, University Hospital Birmingham NHS Foundation Trust,
Birmingham, UK
Aims: NaDIA 2012 cohort was used (1) to assess mortality in

diabetics and possible associations of mortality until 31January
2013 and (2) to examine, in a years time, the prevalence of
diabetes and outcome of the cohort from hospital informatics.
Method: In all, 154 patients with diabetes were identified in
NaDIA 2012. Complete records were available for 145 who were
followed up for the duration of their stay until discharge, death or
31 January 2013. Data were collected on a variety of factors that
influence mortality. Logistical regression was employed for
analysis of the independent variables and ANOVA tests were
then applied to the variables that were found to have a difference in
their mean values. We also collected data on prevalence of
diabetes, mortality and readmission rates of the NaDIA 2012
cohort on 6 September 2013 from hospital informatics.
Results: Fifteen of the145 patients with diabetes died within the
spell in the initial assessment until 31 January. In multivariate
analysis history of cancer increased the risk of inpatient mortality
by 30%. On longer follow-up 197 patients with diabetes were
identified in the cohort. Mortality within spell and readmission
rates in patients with diabetes were 9.13% and 16.24% compared
withpatients without diabetes of 5.77% and 11.67% respectively.
Conclusion: History of cancer is the only variable positively
associated with inpatient diabetic mortality in our cohort. A
snapshot audit like NaDIA may underestimate the prevalence of
inpatient diabetes. Longer follow-up data of this cohort can shed
more light on prevalence of diabetes and important outcome
indicators of care which can be used to benchmark different Trusts.
P349
Management of adults with diabetes
undergoing surgery and elective procedures
JA Worthington, D Flanagan, L Mole and E Moore
Diabetes Centre, Plymouth Hospitals NHS Trust, Plymouth, UK
Aim: Elective surgery at Derriford Hospital is supported by a

full-time diabetes specialist nurse and support assistant. We report
service delivery and outcomes for the last 10 months.
Method: Two pre-assessment wards are visited Monday to Friday.
Patients with diabetes who have an HbA1c > 69mmol/mol and all
patients on insulin are referred. Their medication is titrated via the
service prior to surgery. A diabetes management summary including
the pre-operative and postoperative medication plan is documented
on a diabetes clinical record communication sheet. Referrals to the
diabetes multidisciplinary team are made as required, i.e. dietitian,
podiatrist, diabetes consultants. Emails update surgeons when the
patients diabetes management has improved for surgery. Patients
with diabetes are identified via an alert on the computer system.
Patients are requested to be first on the theatre list. The day of surgery
wards and surgical wards are visited Monday to Friday to give
support. The diabetes specialist nurse liaises with general
practitioners, practice nurses, pre-assessment staff, day of surgery
staff, surgical teams, theatre and recovery staff and surgical ward
staff. The patient is involved in any decisions made.
2014 The Authors.

DIABETICMedicine
Results: 461 patients referred; 89 Type 1 plus nine insulin pump

patients, 357 Type 2. Six not known diabetes. 360 have completed
surgery. Only one admitted the previous night due to diabetes and
other concerns. 101 are still waiting. Mean HbA1c 71mmol/mol
(SD 20) (range 33 169mmol/mol). Total number of contacts so
far 1,773.
Conclusion: Care of patients with diabetes has improved with the
establishment of a dedicated elective surgical support team.
P350
An electronic Think Glucose referral system
to facilitate and monitor inpatient diabetic
care
TM Kapadi, RP Warwick, BC Lee and A Basu
Diabetes and Endocrinology, City Hospital, Birmingham, UK
Aims: To assess admitting clinicians compliance with trust

diagnostic guidelines and to assess our diabetes service response
to patients referred with diabetic ketoacidosis (DKA) or
hyperosmolar hyperglycaemic state (HHS) via an electronic
Think Glucose referral system, which also allowed
determination of inpatient duration.
Method: This was a retrospective audit of all patient records
(electronic and paper) referred via Think Glucose with diagnoses of
DKA or HHS for the period 8 August 2010 to 13 July 2012. The
accuracy of diagnosis, appropriate use of investigations, mortality,
inpatient duration and Think Glucose diabetes service response
times were assessed.
Results: A total of 242 referrals were made; 166 cases (69%) were
reviewed. Of these, 75 (45%) were diagnosed as DKA, 38 (23%)
as HHS and 53 (32%) as hyperglycaemia. 94% of patients referred
with DKA and 89% with HHS met the respective diagnostic
criteria. Five patients diagnosed with hyperglycaemia actually met
criteria for HHS. Laboratory venous glucose testing was performed
in 71.6% cases. Mean DKA admission duration was 4.6 days vs
12.4 days for HHS. 92% of referrals were seen with 84% (DKA)
and 77% (HHS) seen within 1 day. 44% and 41% of DKA and
HHS patients respectively were reviewed by consultant
diabetologists.
Conclusion: An electronic referral system facilitates prompt
review of patients by an appropriate clinician and aids the
collection of data for audit and service evaluation. Most patients
referred met the appropriate diagnostic criteria. Some with HHS
were misdiagnosed as having hyperglycaemia. This may be
explained by poor compliance with laboratory venous glucose
testing or a lack of clinician awareness.
P351
Clinical incidents relating to insulin and oral
hypoglycaemics in hospital
HK Tan, A Baggott, E Moore and D Flanagan
Diabetes and Endocrinology, Derriford Hospital, Plymouth, UK
Introduction: Reducing diabetes related treatment errors has

been a priority of the inpatient diabetes team for the past 2
years. We have analysed the frequency of incidents related to
insulin and oral hypoglycaemics in a tertiary hospital.
Method: Incidents reported on the DATIX risk management
system were searched from 1 September 2012 to 31 August 2013
using keywords diabetes, insulin and hypoglycaemia. Incidents
were categorised as prescription error; administration error; device
error; error in handover; lack of knowledge; and needle stick
131
DIABETICMedicine
injury. We also included the number of incidents related to the use

of variable rate intravenous insulin infusion (VRIII). All incidents
are rated by the person filling in the forms into No harms, Near
miss and Harm.
Results: A total of 107 incidents were reported during the period.
Twenty-one of these incidents were related to the use of VRIII. The
majority of incidents are related to administration error (32%) and
prescription error (28%). Of the administration errors, 47% were
related to a wrongly omitted or missed dose. Incorrect prescription
(60%) and failing to prescribe insulin (37%) appeared to be the
commonest subgroup of prescription error. 63% of the incidents
are rated as Near miss or Harm.
Conclusion: Despite specific work by the inpatient team, incidents
surrounding the usage of insulin continue to be a significant
problem. Further awareness among healthcare professionals is
required to improve the situation. This may include more robust
training on the usage of insulin rather than e-learning and
organised training sessions for healthcare professionals.
P352
Inpatient diabetes: do-it-yourself electronic
referral system to support and enhance
Think Glucose
REJ Ryder, W Burbridge, LS Braycotton, RJ Ryder, ML Cull,
PH Davies, P De, A Basu and BC Lee
Diabetes, Sandwell and West Birmingham Hospitals NHS Trust, Birmingham,
UK
Background: Think Glucose is a national initiative to improve

inpatient diabetes care including effective use of the inpatient
diabetes specialist team(IPDST). It utilises a comprehensive
traffic-light system to guide which patients should be referred to
IPDST. The traffic-lights include 31 cases: 16 always refer cases
(red), eight sometimes refer (amber) and seven rarely refer
(green). It is difficult for busy clinical staff to remember all these
possibilities and, even with reminder cards/leaflets/posters, these
are readily not to hand when needed. Previously referral to IPDST
required phone, fax or internalpost with their built-in delays and
patients being missed.
Aims: Most NHS hospitaltrusts have an internal, online,
electronic investigationordering system. Depending on the test
being ordered, this brings up a form which when completed is sent
electronically to the laboratory or imaging department concerned.
We aimed to use this approach for Think Glucose referral.
Methods: We used our hospitaltrusts investigationordering
system to develop an electronic Think Glucose Assessment form
and mandated all clinical staff to ensure the assessment on all
diabetes patients as soon as possible after admission. The quick
and easy form incorporates the 31 traffic-light cases with simple
tick boxes, and safety data on glucose and feet assessment within
4h of admission. Once submitted the form appears instantly in a
generic NHSemailaccount accessible by IPDST. This facilitates
daily IPDST ward round lists.
Results: Over the year following implementation the number of
patients seen by IPDST increased from 83/month to 452/month,
whilst at the same time ensuring those seen complied with the
national traffic-light criteria.
Conclusion: All hospitals with electronic investigation ordering
systems could easily emulate our system.
132
P353
An audit of mortality of inpatients with
diabetes as a comorbidity
R Zaidi, D Kalathil, M Shah, Y Mahgoub, D Sharma, A Ahmad,
J Vora, TS Purewal and PJ Weston
Diabetes and Endocrinology, Royal Liverpool University Hospital, Liverpool, UK
Aim: Mortality rate of diabetes inpatients was noted to be high at

our centre, as revealed by the National Diabetes Information
Service for 2011 2012. The aim of this audit was to identify
reasons for this high mortality rate.
Methods: We conducted a retrospective audit, analysing case
notes of final admission of 30 inpatient deaths with diabetes (10%
of total mortality) between April 2010 and March 2012.
Results: Mean age of patients was 70.5 10.6 years. Admissions
were under surgical (52%), medical (42%) and ITU (6%) care. All
patients had Type 2 diabetes. HbA1c was not documented in any
patient. Capillary glucose readings were documented in 47%. 76%
of patients did not have any change in antidiabetic medications.
88% were not referred to the diabetes team. All of those referred
were to consultants or registrars and 50% were also to diabetes
specialist nurses. Referral time ranged between 2 and 12 days after
admission and all were seen on the same day of the referral.
Respiratory (24%), gastrointestinal (24%) and cardiovascular
complications (12%) were the main causes of death. It was felt
that diabetes team intervention would have made a difference in
59% of patients.
Conclusions: High mortality rate is contributed to by the
complexity of inpatients, poor documentation of glycaemic
control and low referral to the diabetes team. Optimisation of
glycaemic control is paramount in all inpatients with diabetes.
Automatic notification of all diabetic inpatients to the diabetes
team is advised, as per the NICE Quality Standards Advice.
P354
Improving inpatient insulin prescribing and
administration by introduction of a
standard prescribing chart: results of a
prospective audit
EJ McCracken, PC Johnston, L Hegarty and AG Nugent
Diabetes Education Unit, Belfast City Hospital, Belfast, UK
Aim: Errors in prescribing and administering insulin are common

and can result in serious harm. In total, 3,811 wrong insulin dose
incidents were reported to the National Patient Safety Agency
(NPSA) in 2004 2009. NPSAs 2010 rapid response report Safer
administration of insulin highlighted abbreviations such as U for
units as a cause for dose administration errors. Our trust has
developed an insulin prescribing chart designed to reduce error. We
present audit findings regarding practice in our hospital.
Methods: Using a standard proforma, information was obtained
from insulin prescription charts and medicine kardexes from 19
medical inpatients in our hospital in May June 2012.
Results: With the word units pre-printed, all patients insulin
doses were clearly specified. All had capillary blood glucose levels
recorded appropriately and nurses had signed for all administered
insulin. Despite space to fill in the patients target blood glucose,
this was not completed on any assessed charts. 10/19 home insulin
doses were recorded. 16/19 doses were appropriately increased
with hyperglycaemia. 8/19 patients had a hypoglycaemic episode,
but only six had insulin down-titrated. 11/19 insulin prescriptions
were referenced in the main kardex.
2014 The Authors.

Conclusion: Insulin prescription charts were successful in

avoiding use of abbreviations. Usual home insulin doses and
target blood glucose were poorly recorded. There was suboptimal
referencing of the insulin chart on the drug kardex,
highlighting the risk of the insulin chart being overlooked.
Insulin was not always adjusted for hyperglycaemiaor
hypoglycaemia. We recommend ongoing staff education in order
to minimise risk of error with re-audit to assess achievement of
standards.
P355
The challenges of maintaining an effective
inpatient diabetes service: demonstrating
the need for supporting diabetes resources
A Bray1, B Mumford2, J Jones2, H Rasmussen1, R Griffiths1,
M Fogarsy1, H Husband1, K Hamden2 and N Agarwal2
1
Diabetes Centre, Cwm Taf Local Health Board, Llantrisant, UK, 2Diabetes
Centre, Cwm Taf Local Health Board, Merthyr Tydfil, UK
Objective: The National Diabetes Inpatient Audit (NaDIA)

highlighted elements where ward clinical practices could be
improved to reduce adverse outcomes for hospitalised
patients. Management of hypoglycaemia and errors related to
prescription/management of insulin/oral hypoglycaemic agents
were prioritised to improve patient safety. Dedicated inpatient
diabetes teams and adaptation of the NHS Institute of Innovation
and Improvement educational toolkit, Think Glucose, addressed
staff training requirements. However, the rising prevalence of
diabetes in hospitalised patients accompanied by diminishing
resources challenges the delivery of effective clinical care.
Method: NaDIA 2010 data were used as baseline. Secondary
care medical/nursing/pharmacy staff received Think Glucose
training
sessions
in
insulin/OHA
administration
and
hypoglycaemiamanagement.
Insulin-specific
prescriptions,
bloodglucose monitoring documentation and Hypo-boxes were
introduced. Re-audit data were collected regularly until September
2013 to provide quality assurance and evaluate the impact of
intervention. Non-parametric statistical analyses determined the
significance of resulting differences.
Results: 19% of local acute beds were occupied by patients with
diabetes in 2010, rising to 23% in 2013. Baseline errors in insulin
prescribing occurred in 48% of inpatients, with consequent
management errors in 31%. Post-intervention data (September
2012) demonstrated sustained reductions in prescription
(<10%;p < 0.01) and management errors (<15%;p < 0.05).
Hypoglycaemia management at baseline was appropriate in only
50% of inpatient episodes, subsequently rising to >90%
(p < 0.001). However, with rising demands on inpatient services
and subsequent reductions in educational opportunities to continue
to deliver Think Glucose, the rates of errorsin prescribing (29%;
p < 0.01) and management (24%;p < 0.01) were once again
observed to increase.
Summary: These results support the importance of effective
multidisciplinary education and the need for dedicated inpatient
diabetes teams to raise diabetes awareness amongst healthcare
professionals, reduce adverse outcomes and improve safety for
patients with diabetes when admitted to hospital. However,
sustaining improvement requires continuing education and
support from inpatient diabetes teams and this must be
adequately resourced to remain effective.
2014 The Authors.

DIABETICMedicine
P356
Clinicians opinions of a new inpatient care
pathway, the Diabetes Inpatient Care and
Education (DICE) chart, containing a unique
referral system, the Diabetes Patient at Risk
(DPAR) score
RM Round, C Kerry, R Rajendran, N Anscomb, S Laflin and
G Rayman
Aims: To obtain opinions of a new comprehensive diabetes

inpatient care pathway incorporating a novel referral system.
Methods: As part of our Diabetes Inpatient Care and Education
(DICE) project, addressing issues identified in NaDIA, we devised a
new care pathway contained in an eight page booklet (DICE
chart). Included are an admission check list, patient self-insulin
administration assessment tool, foot assessment tool, combined
glucose/insulin chart, insulin infusion chart, safe discharge check list
and simple guidance for managing hyperglycaemia/hypoglycaemia.
Key is a novel referral system, the Diabetes Patient at Risk (DPAR)
score, that occurs on admission and whenever the patients
condition changes. This empowers any clinician but particularly
nurses to obtain appropriate, timely specialist advice related to the
score, e.g. 8 triggers immediate consultant review.
Results: Ninety-five clinicians were surveyed (23% medical, 77%
nursing). Responses to five statements are summarised in three
categories: strongly agree/agree, neither agree nor disagree,
disagree/strongly disagree. Asked if viewed helpful in improving
inpatient care, responses were 97%, 1%, 2% respectively. Asked
whether the chart length was appropriate, responses were 60%,
19% and 21%; whether user friendly, 64%, 23%, 13%; whether the
DPAR score improves access to specialist advice, 77%, 19%, 4%;
whether an advance on previously used charts, 71%, 22% and 7%.
Conclusions: Within a month of introduction the DICE chart and
DPAR score have been well received by the majority of clinical
staff surveyed. Their effect on referrals and overall patient care
continue to be assessed as part of the DICE project.
P357
Characteristics of patients who develop
hyperglycaemia and require insulin
treatment in hospitalised patients receiving
prednisolone
R Bhatti, N Gayle, S Hopkins, S Thomas and D Kariyawasam
Diabetes and Endocrinology, Guys and St Thomas, London, UK
Aims: To describe the characteristics of patients who develop

hyperglycaemia after receiving prednisolone.
Methods: This was a prospective intervention study from January
to March 2013. Demographic data and blood glucose readings
were collected from Vector and patient notes. An initiation and
titration protocol of insulin was used in eight patients who
developed hyperglycaemia.
Results: Of the 22 patients recruited, n = 8 had Type 2 diabetes,
n = 14 no history of diabetes. 55% of all patients and 35.7% of
the patients with no diabetes developed hyperglycaemia. Patients
with Type 2 diabetes had HbA1c 7.65%compared withthe group
without pre-existing diabetes 5.7% (p = 0.005).The group with
Type 2 diabetes were older and had lower hip waist circumference
compared with those without diabetes. Baseline data: mean(SD)
age 72 14.92 vs 60 21.33 (p = 0.12), hip waist
circumference 109.00 7.98 vs 115.25 7.36 (p = 0.74),
133
DIABETICMedicine
blood glucose monitoring each day 2.84 0.69 vs 1.87 0.83.

N = 9 normoglycaemic
(group
1);
n=5
developed
hyperglycaemia (group 2). Baseline characteristics group 1 vs
group 2, 53 23 vs 74 5.1 (p = 0.041),weight 93.16 33.64
(p = 0.02), body mass index 32.91 11.64 vs 21.76 2.47
(p = 0.03). Mean(SD) length of hospital stay for the patients
who remained normoglycaemic 5.7 3.77compared with
hyperglycaemic patients 7.83 4.84.
Conclusions: 55% of patients developed hyperglycaemia with
prednisolone. Age was significantly higher(and was an independent
predictor) in the group of patients that developed hyperglycaemia
whilst on prednisolone.
P358
Demographics of inpatients presenting with
hypoglycaemia to a district general hospital
HA Bajwa1, MM Alam1, AW Safi2, WB Leong3, K Crowley1,
A Tahrani2 and H Siddique1,2
1
Endocrine and Diabetes, Dudley Group NHS Foundation Trust, Dudley, UK,
Division of Medical Sciences, University of Birmingham, Birmingham, UK,
Endocrine and Diabetes, Heart of England NHS Foundation Trust,
Birmingham, UK
2
3
Aim: To demonstrate characteristics and glycaemic control for

patients admitted to secondary care with an acute hypoglycaemic
attack.
Method: All patients seen by the specialist diabetes team (SDT)
with admission diagnosis of hypoglycaemia from June 2007 to
December 2010 were studied.
Results: Of 2,002 patients seen by the SDT, 11.6% (n = 233)
presented with hypoglycaemia as the primary diagnosis. The mean
age was 71.98 13.64 years, 56% (n = 130) were males and 94%
(n = 219) had Type 2 diabetes. The mean baseline HbA1c was
58mmol/mol (7.48% 1.59). Follow-up HbA1c after treatment
relaxation was marginally but not significantly higher at 59mmol/
mol (7.59% 1.57, p = 0.2). 66.8% of patients in this cohort
were above 70 years of age. The HbA1c increased in those aged
70 years 58mmol/mol (7.50% 1.75) vs 60mmol/mol
(7.64% 1.87, p = 0.4) and for aged >70 years 58mmol/mol
(7.47% 1.51) vs 59mmol/mol (7.56% 1.40, p = 0.4).
Conclusion: Nearly 60% of patients in this cohort were over the
age of 70 years. Tight glycaemic control in elderly patients may
sometimes be detrimental. Poor nutritional status, associated
comorbidities such as dementia, cerebrovascular accident, poor
cognitive status, worsening renal and liver functions and
polypharmacy may predispose them to increased risk of
hypoglycaemia.Management should therefore be tailor-made
taking the above factors into consideration as well as their
ability to recognise and treat hypoglycaemia. SDTreview during
their hospital stay may provide a unique opportunity to empower
patients and address such issues.
P359
An audit of the management of diabetic
ketoacidosis in a district general hospital
after introduction of the Joint British
Diabetes Societies guidelines
H Kurdi, LP Pinto and FJ Smeeton
Diabetes and Endocrinology, Nevill Hall Hospital, Abergavenny, UK
Aim: The Joint British Diabetes Societies (JBDS) published

guidelines for the management of diabetic ketoacidosis (DKA) in
2010. Admissions with a diagnosis of DKA were audited before
134
and after introduction of the new protocol in a district general

hospital.
Methods: Twelve months of admissions coded for DKA were
retrospectively audited in 2010 and 2013.
Results: Thirty-three episodes with a correct diagnosis of DKA
were identified in 2013 with a median stay of 3 days; 36% (n = 12)
were managed on ITU. The new protocol was implemented in 31
cases (94%). In the first hour intravenous fluids were started in
70% (n = 23) and intravenous insulin infusion (IVII) in 54%
(n = 18). Fixed rate insulin infusion (FRII) was used in 90%
(n = 30) in 2013 compared with an equivalent number on variable
rate insulin infusion (VRII) in 2010. Blood ketones were measured
in 27 cases (90%), potassium 20 cases (67%) hypokalaemia
occurred in 30% (n = 9). Long acting insulin was continued in
43% (n = 13). The rate of hypoglycaemia was 10% (n = 3) on
FRII compared with 47% (n = 12) on VRII in 2010. Ketosis
resolved within 24h in 97% (n = 29), median time 11h.
Subcutaneous insulin was restarted within 24h in 60% (n = 18).
Mean time on IVII was 11h in 2013 vs 35h in 2010.
Conclusions/summary: The
number
of
admissions,
demographics, length of stay, timely use of intravenous insulin
and fluid resuscitation were similar in both audits. There were
lower recorded rates of hypoglycaemia on FRII and a reduced time
on IVII following adoption of the new guidelines.
P360
Diabetes inpatient consultant input
improves follow-up of patients admitted
with hypoglycaemia
K Clements, V Connor, T Williams, S Wynne, R Lim, M Banerjee
and KS Leong
Wirral University Teaching Hospital, Wirral, UK
Aim: To improve diabetes care, a consultant with an interest in

inpatient diabetes care was appointed in April 2013. Our aim was
to determine if inpatient consultant input increases review and
follow-up of patients admitted with hypoglycaemia.
Method: Electronic records of all patients admitted with
hypoglycaemia over the past 12 months (September 2012 to
August 2013) were reviewed. The records were examined to
determine if patients were reviewed by the diabetes team, whether
a hypoglycaemia review form was completed, if there was
post-discharge follow-up and if readmission occurred. We
compared the data pre and post May (when change in practice
commenced).
Results: Sixty-seven (31 male, 36 female) patients (12
readmissions) were admitted in total; 42 patients (seven
readmissions) from September to May and 25 patients (five
readmissions) from May to August 2013. The mean ages were
similar (67.8 vs 65.2 years). 66% of patients were on insulin and
30% on a sulphonylurea. From May, more patients were seen by
the specialist diabetes team [22 (71%) vs 30 (61%), p = 0.27]; and
an increased number of electronic hypoglycaemia review forms
were completed [11 (50%) vs10 (33%), p = 0.22]. There were
fewer readmissions with hypoglycaemia within 1 month of
discharge, 2 (7%) vs 8(16%), p = 0.1316. However, significantly
more patients were followed up by the specialist diabetic team
following an admission: 15 (50%) vs 11 (22%) (p = 0.0114).
Conclusions: Following the appointment of a consultant to
improve inpatient care, significantly more patients with
hypoglycaemia have been followed up. It is too early to
determine if this intervention will result in reduced readmissions.
2014 The Authors.

P361
Never too late for diabetic ketoacidosis
TS Shekaraiah and PJ Evans
Department of Endocrine and Diabetes, Royal Gwent Hospital, Newport, UK
Background: An extensive review of the literature yielded limited

reports of new onset Type 1 diabetes in elderly patients. We
present one such case in a 77-year-old female.
Case history: A 77-year-old Caucasian female presented with
acute delirium. History from husband revealed that the patient had
been lethargic for 2 weeks, had developed polydipsia and polyuria
over 2 days and became drowsy on the day of admission. There
was no history of significant comorbidities and no regular
medications. Patient did not drink alcohol but had been a heavy
smoker. On examination, the patient was confused, dehydrated
and in hypovolemic shock. The rest of the examination was
unremarkable with no signs of sepsis. Investigations revealed blood
glucose of 47.7mmol/l, pH 6.98, pCO2 2.4kpa, HCO3 4.7mmol/l,
base excess of 27mmol/l, blood ketones 10.5mmol/l and serum
osmolality 373mmol/kg. The patient was diagnosed with diabetic
ketoacidosis (DKA) and treatment was started without delay. She
improved clinically and insulin was stopped, but the patient
reverted back to DKA. She was then commenced on a twice daily
insulin regime. Further investigation revealed anti-glutamic acid
decarboxylase antibodies of >2,000U/ml.
Discussion: New onset diabetes in the elderly is usually classified
as Type 2 diabetes. However, this patient with ketoacidosis, insulin
dependence and antibody positivity makes Type 1 diabetes more
likely. The differential diagnosis includes latent autoimmune
diabetes in adults. This is unlikely in view of the rapid
progression to insulin dependence.
Conclusion: Clinicians should be aware of the possibility of
late-onset Type 1 diabetes in elderly patients.
P362
Two years on, how well are we doing?
Management of diabetic ketoacidosis (DKA)
in acute medicine
RWF Lee, A Abiodun, CL Wong, R Buka and P Rehsi
DIABETICMedicine
performance. Following presentation of these findings in both the

diabetes journal club and the acute medicine forum, a one page
simplified guideline adapted from the JBDS guidelines and bedside
ketone meters have been introduced in QEHB. Hospitals across the
country should review their local guidelines and ensure they are up
to date and user friendly to junior doctors.
P363
An evaluation of a proactive inpatient
diabetes team initiative on a liver unit
GM Noble-Bell and A Forbes
Diabetes, Kings College Hospital, London, UK
Background: The proactive diabetes inpatient team initiative is a

multi-focal innovation that includesdaily inpatient review; electronic
monitoring of hypoglycaemia and hyperglycaemia; therapy
modulation; staff training and support; and patients education.
Aim: The aim of this study was to evaluate the impact of the
proactive inpatient diabetes team on medication management and
blood glucose control in a liver ward setting.
Method: We did a prospective audit of 69 patients post liver
transplantation or other hepatobillary diseases with coexisting
diabetes on a liver ward over a 3 month period. Data were
collected over 3 day periods at three 6 weekly intervals following
the introduction of proactive care. Data were collected on good
glucose days based on national standards: blood glucose levels 4
and 11mmol/l in a 24h period with not more than one blood
glucose level outside this range (NaDIA 2011) .
Results: Most patients (91%) were on insulin therapy. During
observation 1 (pre innovation) only 22% of cases had a good
glucose day compared with 61% at observation 3 (post
innovation). The data showed that insulin dose adjustment was a
significant factor for good blood glucose days (p = 0.003). A
reduction in medication errors was also observed, with no errors
for wrong timing of insulin at observation 3 compared with 3% in
observation 1.
Conclusion: The proactive diabetes team initiative led to a
significant improvement in increasing the number of patients
with good glycaemic control and improved medicines management
for liver inpatients.
Acute Medicine, Queen Elizabeth Hospital Birmingham, Birmingham, UK
Background: DKA is often initially managed by junior doctors

who are inexperienced in managing diabetic crises. The Joint
British Diabetes Societies (JBDS) inpatient care group published
guidelines in March 2010 which provide a framework in
management of DKA nationally.
Aim: To determine whether the initial management of DKA in
Queen Elizabeth Hospital Birmingham (QEHB) adhered to the
JBDS guidelines.
Methods: We analysed records of patients admitted to QEHB
with DKA between October 2010 and October 2012. The initial
investigations performed and treatments commenced were assessed
and compared with the recommendations in the JBDS guideline.
Results: Eighty-six patients met our inclusion criteria. All patients
had capillary glucose and venous blood gas performed, of which
33% had not had a laboratory glucose performed. 95% of patients
had a urine dip stick test with subsequent microscopy and culture.
33.3% of patients were not given an appropriate fluid regime and
24.4% of patients were not given insulin in the fixed rate infusion.
Only 5.8% of patients had ketones, either urinary or blood,
monitored hourly in the first 6h when treatment had commenced.
Conclusion: Local guidelines for management of DKA in QEHB
were outdated and this may explain the below standard
2014 The Authors.

P364
An insight into nursing staff knowledge on
managing hypoglycaemia
K Leong, P Joseph and KS Leong
Diabetes Department, Clatterbridge Hospital, Wirral, UK
Background: In 2011 hypoglycaemia treatment boxes (hypo

boxes) which also contained the local trust protocol for
managing hypoglycaemia were distributed to all clinical areas.
All nurses were given education on managing hypoglycaemia as
part of this initiative. However, some recent ward referrals
demonstrated that this knowledge was suboptimal.
Aim: To determine the nurses current knowledge on
hypoglycaemia management.
Method: A questionnaire was devised to assess staff knowledge.
The questionnaires were distributed by the diabetes nurses to five
clinical areas (Outpatients, Dermatology, Stroke and two surgical
wards). Both qualified and non-qualified staff completed the
questionnaires.
Results: Sixty-five (64%) nurses completed the questionnaires, 43
(66%) qualified and 22 (34%) non-qualified. All staff recognised
135
DIABETICMedicine
what hypoglycaemia was and the associated symptoms. However,

only 25 (38%) staff followed the local trust protocol and 42 (65%)
were aware of the availability of hypo boxes and their use. 11
(17%) non-qualified members of staff felt it was not their
responsibility to check or treat patients for hypoglycaemia. This
was worrying as some clinical areas have a higher non-qualified to
qualified staff ratio.
Conclusion: A significant proportion of nursing staff were not
familiar with the local protocol for treating hypoglycaemia or the
availability of the hypo boxes. There was some confusion amongst

some members of staff about their roles and responsibilities in
assessing and treating patients with hypoglycaemia. Regular
education updates are required to maintain nurses knowledge of
hypoglycaemia management. Since identifying these findings a
programme of education has been devised and this is being
implemented.
Clinical care and other categories posters: insulin: actions,

metabolism and therapy
P365
Predictors of response to biphasic insulin
(Humalog Mix50) following conversion from
other insulin regimens: a retrospective
cohort study
J Mamza1, R Mehta2 and I Idris1
1
Division of Medical Sciences and Graduate Entry Medicine, University of
Nottingham, Derby, UK, 2Trent Research Design Services, Nottingham
University Hospitals, Nottingham, UK
Background: Intensification of insulin regimen is paramount to

achieving and maintaining glycaemic control in patients with
diabetes. We examined clinical parameters that might predict
glycaemic response to Humalog Mix50 and tested its effectiveness
in reducing HbA1c levels among patients with suboptimal glucose
control.
Methods: A cohort of 2,183 patients 18 years old with diabetes,
newly treated with Humalog Mix50 between January 2000 and
May 2012, were sourced from UK General Practices via the Health
Initiative Network (THIN) database. Baseline parameters of 1,538
patients with suboptimal HbA1c >7.5% who had received other
insulin regimens for at least 6 months before the index date were
compared against 12months follow-up data. Response was
defined as HbA1c < 7.5% and/or HbA1c reduction by >1% at
12 months. A subgroup of 233 patients initiating insulin therapy
with Humalog Mix50 was also assessed. Associations were
examined using multivariate logistic regression techniques.
Results: At 12 months, patients on existing insulin therapy who
were converted to Humalog Mix50 showed a mean HbA1c
reduction of 2.1% (p < 0.001). Adjusted odds ratios (OR) show
glycaemic response to intensification was significantly associated
with baseline HbA1c [OR1.72(1.58 1.87)], increased body mass
index [OR1.09(1.03 1.15)], reduced cholesterol [OR0.86
(0.76 0.98)] and sulphonylurea use [OR1.32(1.04 1.68)].
Patients in the subgroup showed a significant HbA1c reduction
of 3% (p < 0.001). Response in this subgroup was associated with
gliptin use [OR4.59(1.26 16.7)].
Conclusion: Humalog Mix50 is effective for achieving glycaemic
control in some patients with diabetes with suboptimal glucose
control on other insulin regimens. Higher baseline HbA1c,
previous sulphonylurea or gliptin use appear to be independent
predictors of response to this insulin regimen.
Clinical care and other categories posters: lesser known

complications
P366
P367
Glycated haemoglobin and all-cause and

cause-specific mortality: a cohort study
utilising data from the Health Survey for
England and Scottish Health Survey linked
to mortality records
Glucose characteristics predict mortality in

inpatients with diabetes
CAR Sainsbury and GC Jones
Department of Diabetes, Gartnavel General Hospital, Glasgow, UK
Refer to Oral A68
VLZ Gordon-Dseagu, J Mindell and N Shelton

Department of Epidemiology and Public Health, University College London,
London, UK
Refer to Oral A65
136
2014 The Authors.

Clinical care and other categories posters: lesser known complications
P368
Should we routinely ask our patients with
diabetes about gastrointestinal symptoms
and do we think about pancreatic exocrine
insufficiency?
PL Chong, DR Meeking, ICP Cranston, PS Kar, V Hunter and
MH Cummings
Academic Department of Diabetes and Endocrinology, Portsmouth Hospitals
NHS Trust, Portsmouth, UK
Background: Anecdotal evidence suggests that healthcare

professionals do not routinely ask patients with diabetes about
gastrointestinal (GI) symptoms nor consider screening for
pancreatic exocrine insufficiency (PEI) other than in those
patients with secondary diabetes. We aimed to assess the
presence of GI symptoms in patients with diabetes and the
presence of possible PEI in symptomatic patients.
Methods: We evaluated the presence of GI related symptoms
[incorporating the Bristol Stool Chart score (BSCs)] in consecutive
diabetes patients attending our clinics. In symptomatic patients, we
ascertained how many were screened for PEI using a faecal
elastase-1 stool measurement (FE-1).
Results: In all, 156 patients were screened. Of these, 34 patients
(22%) described abnormal stools (BSCs > 5), steatorrhoea or
unexplained weight loss. FE-1 samples were requested in all 34
patients but only 19 patients (56%) elected to provide a faecal
sample. 7/19 (37%) symptomatic patients had low FE-1 levels
either severe ( < 100mcg/g) 2/7 (29%), or moderate
(100 200mcg/g ) 5/7 (71%). Three patients had Type 1
diabetes, three had Type 2 diabetes and one had secondary
diabetes. 6/7 patients (86%) had HbA1c 69mmol/mol, diabetes
duration > 5 years and at least one other diabetes complication.
Conclusion: These preliminary data suggest that GI related
symptoms are frequently observed in patients with diabetes and
should be sought. In addition PEI may be a common cause of GI
symptoms in all forms of diabetes.
DIABETICMedicine
Results: Amongst the 51,898 inpatient episodes for 36,503

individuals, a diagnosis of diabetes was recorded in 7,377
episodes (14.2%) for 4,792 individuals. The overall inpatient
mortality was 4.1% (2,140 deaths). The mortality was significantly
higher in people with diabetes (6.1% vs 3.8%, p < 0.001).
Diabetes remained an independent predictor of inpatient
mortality [odds ratio (OR) 1.31 (1.17 1.46), p < 0.001] after
correction for confounding factors (age, gender and ethnicity) by
logistic regression analysis. Diabetes was recorded in the death
certificate only in 29.6% of those with diabetes. Of the 452
deceased individuals who had diabetes, the main causes of death
were sepsis [44%,the majority chest sepsis (74% of all infections)],
vascular (coronary or cerebrovascular events) (33%), malignancy
(13.5%) and acute kidney injury (4.6%).
Conclusion: Diabetes should be consistently documented in death
certificates to improve coding accuracy. The coexistence of
diabetes is associated with significantly higher mortality in
hospitalised patients. The contribution of diabetes to mortality
could be multifactorial; infection and vascular causes remain the
leading cause of mortality.
P370
A preliminary survey on awareness of
periodontal risk among people with
diabetes in a district general hospital
K Jeyaraman1, Z Vakil1, P Rihal1, C Gunpot1, E Umotong1,
R Kaushal1, C Cotzias2, A Dixit2 and R Kaushal1
1
Department of Diabetes and Endocrinology, West Middlesex University
Hospital, London, UK 2Department of Obstetric and Gynaecology, West
Middlesex University Hospital, London, UK
Aim: To study the impact of diabetes on inpatient mortality

during emergency admissions at a large tertiary centre and explore
the causes of in-hospital death in individuals with diabetes.
Methods: All admissions to the hospital in 2012 were identified
from the hospital information management system. Day-case,
elective, maternity and paediatric hospital admissions were
excluded. Individuals with diabetes were identified using the ICD
codes. The causes of death for inpatients with diabetes were
obtained from the hospital bereavement services or the
histopathology department for patients who had undergone
autopsy.
Background: There is growing evidence linking periodontal

disease and diabetes. Active periodontal disease in people with
diabetes may contribute to worsening glycaemic control and
increases the risk of ischaemic heart disease and stroke.
Aims: The aim of this study was to evaluate the awareness and
perception of the above link among people with diabetes in a
district general hospital.
Methods: The study was conducted among people with diabetes
(n = 200) attending specialist clinics between March and May
2013. A patient questionnaire created with the help of a validated
risk assessment tool developed by the American Academy of
Periodontology was used to determine diabetes and dental history.
Results: The response rate was 100%. 58% had Type 2 diabetes,
33% had Type 1 diabetes and 1% had latent autoimmune diabetes
of adulthood. 43% had diabetes for more than 10 years. 33% had
bleeding gums, 18% had loose teeth, 15% had receding gums and
3% had pus visible on gums. 69% brushed twice or more per day
and only 18% flossed daily. 16% have had teeth extraction due to
gum disease in the past. Only 70.5% have seen a dentist in the last
2 years. 90% were never asked about gum health by the healthcare
provider.
Conclusions: A high number of people with diabetes attending
the hospital had an active gum disease and were unaware of the
mutual relation between diabetes and periodontitis which
demonstrates that this topic is neglected in diabetic care. This
study highlights the importance of having appropriate guidelines
for the management of gum disease in people with diabetes.
2014 The Authors.

137
P369
The adverse impact of diabetes on
in-hospital mortality for non-elective
admissions
G Gill, B Jose, L Varadhan, GI Varughese, A Walker and
AU Nayak
Diabetes and Endocrinology, University Hospital of North Staffordshire NHS
Trust, Stoke-on-Trent, UK
DIABETICMedicine
Clinical care and other categories posters: lipids and

fatty liver
P371
Biochemical and radiological factors
associated with the development of
advanced liver disease: the Edinburgh Type
2 Diabetes Study (ET2DS)
JR Morling1, JA Fallowfield2, RM Williamson3, CM Robertson1,
LD Nee4, S Glancy4, IN Guha5, MWJ Strachan2 and JF Price1
1
UK, 2Queens Medical Research Institute, University of Edinburgh, Edinburgh,
UK, 3Metabolic Department, Western General Hospital, Edinburgh, UK,
4
Radiology Department, Western General Hospital, Edinburgh, UK, 5Digestive
Diseases Biomedical Research Unit, University of Nottingham, Nottingham,
UK
Objective: Advanced chronic liver disease (CLD) is more

common in people with Type 2 diabetes. There is no consensus
on the best way to identify people with diabetes at risk of
CLD. We report on the development of CLD in a representative
population-based cohort with Type 2 diabetes aged 60 75 years,
and its association with baseline liver blood tests and hepatic
steatosis.
Methods: Altogether 922 participants in the ET2DS had liver
blood tests measured [alanine aminotransferase (ALT), aspartate
aminotransferase (AST) and gamma-glutamyltransferase (GGT),

hyaluronic acid (HA)] and an ultrasound assessment of hepatic
steatosis. They were followed up for incident advanced CLD
[hepatocellular carcinoma (HCC), oesophageal/gastric varicesand
hepatic fibrosis/cirrhosis].
Results: After a mean 6 years follow-up there were 25 (2.7%)
patients with incident advanced CLD (HCC n = 6, varices n = 10
and fibrosis/cirrhosis n = 24). The presence at baseline of ALT,
AST or GGT above the upper limit of normal (ULN) was
associated with a significantly higher development of advanced
CLD [normal 1.0% (8/771), >ULN 7.7% (9/117), >2 9 ULN
25.8% (8/31), p < 0.001]. Development of advanced CLD was
also associated with higher baseline HA levels (<50U/l 0.4%,
50 100U/l 1.4%, >100U/l 25.0%, p < 0.001). The presence of
hepatic steatosis (any cause) was not associated with advanced
CLD development (normal 2.3%, steatosis 3.1%, p > 0.005).
Conclusions: Abnormal tests of liver function and structure
identify people with Type 2 diabetes at increased risk of CLD,
especially if the blood test results are substantially elevated.
Hepatic steatosis per se is not associated with the development of
CLD.
Clinical care and other categories posters: MODY

P372
MODY is uncommonly diagnosed in
non-white ethnic groups as a result of
difficulty in differentiating young-onset
Type 2 diabetes from MODY
S Misra1, B Shields2, K Colclough3, DG Johnston1, NS Oliver1,
S Ellard3 and AT Hattersley2,4
1
Diabetes, Endocrinology and Metabolism, Department of Medicine, Imperial
College London, London, UK, 2National Institute for Health Research, Exeter
Clinical Research Facility, University of Exeter, Exeter, UK, 3Molecular
Genetics, Royal Devon and Exeter NHS Foundation Trust, Exeter, UK,
4
Research and Development, Royal Devon and Exeter NHS Foundation Trust,
Exeter, UK
Background: Maturity onset diabetes of the young (MODY) has

been extensively described in White-European (WE) populations;
however, comparatively few cases have been found in other ethnic
groups. It is unclear if this reflects a lower prevalence or a disparity
in testingrates. We examined non-white (NW) referrals for MODY
testing in the UK.
Methods: A retrospective cross-sectional analysis of UK patients
referred for MODY genetic testing was undertaken. Ethnic origin,
138
results of genetic analyses and clinical characteristics were

examined in all referrals.
Results: In all, 305 of 3,674(8.3%) referrals were of NW ethnic
origin. South Asian people represented the largest proportion
(5.7% of all referrals), matching UK census data (5.3%
population). The mutation pickup rate was lower in the NW
group than the WE (13.8%vs 27.3%,p < 0.001). HNF1-a
mutations were the commonest in both groups. NW referrals
were diagnosed younger and had higher HbA1c values (p < 0001),
although similar body mass index to the WE group. NW people
without mutations were less likely to be insulintreated than WE
people(39%vs 48%,p < 0.05), although those with mutations had
similar proportions. In people without mutations, parental diabetes
was more common in the NW group (75% vs 57.7% WE,
p < 0.001), but similar in those with MODY.
Conclusions: The reduced UK prevalence of MODY in NW
peoplereflects a lower pickup rate. The reported characteristics of
the NW referrals suggest a higher prevalence of Type 2 diabetes in
this group. Differentiating MODY from young-onset Type 2
diabetes in NW people is harder than in WE people. Further
work is required to develop targeted referral criteria for NW ethnic
groups.
2014 The Authors.

Clinical care and other categories posters: monitoring
DIABETICMedicine
Clinical care and other categories posters: monitoring

P373
Equivalent glycaemic cut-points for the
onset of microalbuminuria in White
Europeans and South Asians: a
cross-sectional analysis of a population
based cohort
SA Mostafa, DH Morris, MJ Davies, BT Srinivasan, D Webb and
K Khunti
Diabetes Research Centre, Leicester Diabetes Centre, Leicester, UK
Aims: The presence of microalbuminuria (MA) is associated with

adverse long-term outcomes with evidence suggesting that early
MA detection and multifactorial risk factor management reduces
mortality rates. The onset of MA may begin at lower levels of
glycaemia in ethnic minority groups, but less is known between
White Europeans (WE) and South Asians (SA). We analysed this in
a UK multi-ethnic cohort.
Methods: We did a cross-sectional analysis of 6,372 participants
(1,684 SA) aged 40 75 in the ADDITION-Leicester study who
were screened between 2004 and 2009. Within each ethnic group,
the cut-point of fasting plasma glucose (FPG), 2h plasma glucose
(2hPG) and HbA1c associated with MA (defined as ACR2.5 and
3.5 in men and women, respectively) was estimated by fitting
fractional polynomial models.
Results: In WEs, the onset of MA began at FPG, 2hPG and
HbA1c of 8.5mmol/l, 11.8mmol/l and 6.7% respectively in men
and 23.1mmol/l, 21.4mmol/l and 8.7% respectivelyin women. The
same analysis in SAs corresponded to 5.7mmol/l, 10.0mmol/l and
6.2% in men and 10.7mmol/l, 12.0mmol/l and 7.4% in women
respectively. The corresponding ethnic difference for MA onset
was 2.8mmol/l, 1.8mmol/l and 0.5% lower in SA men and
12.4mmol/l, 9.4mmol/l and 1.3% in SA women. ACR was
independently higher in SA compared with WE in men, mean
difference 1.18mg/mmol (95% CI 1.09 1.27, p = 0.031), after
adjusting for HbA1c, FPG, 2hPG, systolic/diastolic blood pressure,
anti-hypertensive medication, age, creatinine, body mass index and
smoking status, but not in women, 1.15mg/mmol (1.06 1.24,
p = 0.515).
Conclusion: MA begins at lower levels of glycaemia in SA. Future
research should ascertain whether earlier detection of MA leads to
better long-term outcomes.
P374
Development of electrochemical
non-esterified fatty acid (NEFA) biosensor
for patient management of Type 2 diabetes
AT Hussain1, M Catt2 and EH Yu1
1
Chemical Engineering and Advanced Materials, Newcastle University,
Newcastle-upon-Tyne, UK, 2Institute of Ageing and Health, Newcastle
University, Newcastle-upon-Tyne, UK
Background: Once a patient is diagnosed with Type 2 diabetes

there is simultaneous elevation of blood glucose and non-esterified
fatty acid (NEFA) levels in patients blood.
Aims: To develop a biosensor that will determine the NEFA
concentrations using an electrochemical method, which enables
further development of multiplex biosensors to detect glucose and
NEFA as well as other biomarkers from a single sample.
2014 The Authors.

Methods: The enzymes involved are based on the two-step

enzymatic reactions for NEFA. Acyl-coenzyme A synthetase
(ACS) step 1: NEFA + coenzyme A + ATP ? acyl-coenzyme A +
AMP + pyrophosphate. Acyl-coenzyme A oxidase (ACOD) step 2:
acyl-coenzyme A + O2? enoyl-coenzyme A + H2O2. The fatty acid
investigated is oleic acid, one of the most abundant plasma NEFAs.
As H2O2 production is proportional to the concentration of NEFA,
H2O2 was electrochemically detected via cyclic voltammetry,
linear sweep voltammetry and chronoamperometry.
Results: Oxidation of H2O2 product was subsequently detected in
solution using different types of screen printed electrodes. Current
response showed linearity up to 3mM NEFA concentrations.
Summary: The electrochemical detection of H2O2 from the
NEFA enzymatic reactions produces a platform for development
of an electrochemical biosensor with linearity, sensitivity and
reproducibility using oleic acid as the reference NEFA. Having an
efficient method of NEFA detection will give both the clinician and
patient added information on metabolism, its management and
control. The device could be used for lifestyle intervention and
management.
P375
Glucometabolic abnormalities and their
association with novel biomarkers in
patients with acute coronary syndrome
MA Karamat1,2, AA Tahrani1,2, UY Raja3 and MJ Stevens1,2
1
Diabetes, Heartlands Hospital, Birmingham, UK, 2Centre for Endocrinology,
Diabetes and Metabolism, University of Birmingham, Birmingham, UK,
3
Diabetes, Warwick Hospital, Warwick, UK
Aims: To explore the potential of a panel of biomarkers to

enhance the power of our screening tests to detect diabetes in
patients with acute coronary syndrome.
Methods: This was a prospective 3 year study carried out in two
large inner city hospitals in the UK.
Results: In all, 118 patients were included in the analysis. At
baseline we noticed a clear positive association between fasting
plasma glucose (FPG) and c-peptide, pro-insulin, HOMA IR and
HOMA b. In contrast HOMA IS was negatively associated with
baseline FPG. The post-load glucose concentrations were positively
associated with most variables including c-peptide, pro-insulin,
glucagon, leptin, leptin adiponectin ratio, HOMA IR, TIMP-1
and IL-1ra. HOMA IS was negatively associated with post-load
glucose. The HbA1c was positively associated with c-peptide,
pro-insulin, HOMA IR, HOMA b and IL-1ra. At baseline mean
c-peptide, glucagon, intact pro-insulin and HOMA IR were higher
in the diabetic group compared with normal and IGS groups.
HOMA IS was lower in the diabetic and IGS groups compared
withthe normal cohort. At 3 months mean c-peptide, IL-1ra,
TIMP-2 and intact pro-insulin were higher in the diabetic group
compared withthe NGT and IGS groups.
Conclusion: These findings suggest we can in future potentially
look at utilising some of the markers on insulin production and
function as helpful in predicting glycaemic status. One of our
future aspirations is to look at comparing this with long-term
cardiovascular morbidity and mortality data in those with normal
as well as impaired glycaemic status or Type 2 diabetes.
139
DIABETICMedicine
P376
Biochemical monitoring in patients with
diabetes treated with variable rate insulin
infusion
P377
Z Banu and D Flanagan
JL Griffin1, RSJ Williams1, C Nwoguh1, S Wylie2 and P Grewal1
Endocrinology and Diabetes, Derriford Hospital, NHS, Plymouth, UK
1
Department of Vascular Surgery, Queen Alexandra Hospital, Portsmouth,
UK, 2Department of Microbiology, Queen Alexandra Hospital, Portsmouth,
UK
Aim: There is concern that a commonly used variable rate

intravenous insulin infusion regimen (insulin with 5% dextrose)
(VRIII) may be associated with a high risk of hyponatraemia
although there are few published data to support this. The aim of
this study was to determine the frequency of hyponatraemia in
patients with diabetes managed with VRIII.
Methods: This was a prospective observational study. Twentythree patients with diabetes were enrolled in the study using the
standard Derriford Hospital VRIII protocol. Insulin was infused
between 0.5 and 6 units/h with a fixed rate 5% dextrose at 1ml/kg/h.
Biochemistry was monitored pre, during and post VRIII use.
Results: Eighteen patients were male. Mean age was 66.5years.
Nineteen patients were from the medical ward and four were from
the surgical ward. Duration of insulin infusion was <24h in 13
patients, 24 48h in six patients, 48 72h in two patients and more
than 72h in two patients. Sodium level prior to infusion was low in
nine patients (<135mmol/l) and normalised in five on VRIII.
During VRIII four remained hyponatraemic; 24h post infusion
all but one had normal sodium (a patient with chronic
hyponatraemia). Potassium was low in two patients pre VRIII
and became normal post VRIII. One patient with normal
potassium became hypokalaemic post VRIII. Capillary glucose
during the infusion ranged from 4.0 to 33.2 with two episodes of
hypoglycaemia.
Conclusion: Hyponatraemia is relatively common in patients
with unstable diabetes but appears to be an unusual problem in
individuals using VRIII with insulin and 5% dextrose.
Systemic inflammatory response syndrome

(SIRS) score is a poor predictor of infective
organism in diabetic foot infection
Objective: A key aspect of the initial management of diabetic foot

infection (DFI) is the prescription of empirical antibiotics. This
study investigated whether the SIRS score could be used to identify
patients who went on to grow gram-negative or anaerobic isolates
from their ulcers and could therefore be used to better direct
empirical antibiotic therapy.
Methods: Patients admitted with DFI between January and
December 2012 were included in the study. Clinical records and
microbiological results were reviewed. Patients were divided into
two groups SIRS positive and SIRS negative on the basis of a
SIRS score of 2 or more.
Results: Microbiology results were available for 73 patients (SIRS
positive 32, SIRS negative 41). There was no significant difference
in the mean number of isolates grown by each group (SIRS positive
3.3 vs SIRS negative 2.9, p = 0.29). There was no significant
association between having a SIRS score of 2 or more and growing
a gram-negative or anaerobic isolate (p = 0.5, odds ratio 1.15).
This lack of association remained when only specimens obtained in
theatre were used in a subgroup analysis.
Conclusions: The SIRS score does not adequately predict the type
of isolate that patients with DFI will grow. As such, its use in
directing empirical antibiotic therapy cannot be recommended.
Clinical care and other categories posters: nephropathy

P378
Improving quality in care: chronic kidney
disease in diabetes. Implementation of a
continuous audit tool to facilitate focused
review and optimal medicines management
of patients with diabetes and chronic
kidney disease (CKD) stage 3 and above
within the primary care setting
N Milne
Primary Care, Northenden Group Practice, Manchester, UK
Aims: A significant event review within a general practice

highlighted that, when there is a decline in renal function outside
a patients formal diabetes review, medication optimisation is not
always undertaken. The development of an audit tool would
facilitate the continuous and focused review of CKD stage 3 and
above within the practice diabetes population which, in
addition, would also enhance appropriate medicines management.
Methods: An automated weekly search engine using the
practice information technology system was developed and
implemented to identify and invite patients within three
population subgroups [estimated glomerular filtration rates (ml/
min/1.73m) 46 60, 31 45, and <30] that were failing to
140
meet practice established protocol targets to include optimal

medication.
Results: Within the practice diabetes population of 616
patients, the prevalence of CKD stage 3 and above was 22%.
During the first 6 months, eight patients (6%) required adjustment
of their diabetes medication therapy and four patients
(3%) required alternative medication to their non-steroidal
anti-inflammatory drugs. Those patients without a recorded
HbA1c fell from 8% to 3% and at the end of the 6 months all
patients with an HbA1c over 54mmol/mol had a review plan.
Those without a blood pressure measurement fell from 3% to 2%.
Patients who were not prescribed an angiotensin-converting
enzyme inhibitor or angiotensin-II receptor antagonist lowered
from 6% to 1%. Those patients without a measurement of
microalbuminuria rose from 12% to 13% thus highlighting an
area for improvement. Three patients were identified as requiring
referral to secondary care.
Conclusion: This audit tool, in combination with subsequent focused review, is effective in reducing variation and in
improving quality of patient care. In addition, it affords timely
intervention to ensure appropriate and optimal medicines
management within this complex patient population.
2014 The Authors.

Clinical care and other categories posters: nephropathy
P379
Improving uptake of annual urine
microalbuminuria screening in people with
diabetes attending the Newcastle Diabetes
Centre
AL Mitchell, R Wright, N Leech and SM Marshall
Newcastle Diabetes Centre, Newcastle-upon-Tyne NHS Trusts,
Background: NICE state that people with diabetes should

provide a urine specimen annually for microalbumin estimation.
In Newcastle, people attending for annual review are sent an
appointment reminder letter asking them to bring a urine sample;
however, in 2012 only 48% had a urine test. Aims: We aimed to
determine why individuals do not provide a urine sample and
whether sending an information leaflet with the appointment letter
might improve screening uptake.
Methods: The 102 patients attending for annual review over 2
weeks were studied; patient demographics and whether the patient
provided a urine sample were recorded. Patients who did not were
invited to give a reason. A short information leaflet explaining the
purpose of urine screening was then designed and sent to all
patients prior to their annual review. After 3 months, 125 patients
attending for annual review over 3 weeks were studied and the
same information was gathered. The results were compared by v2
test.
Results: Of 102 people (54% female; mean age 57, range 20 91
years), 70 (69%) brought a urine sample and 32 (31%) did not
(81% of these had forgotten). At the 3 month review, of the 103
people who had received the information leaflet (49% female:
mean age 59, range 24 93), 84 (82%) had brought a urine sample
and 19 (18%) had not (84% of these had forgotten). The patient
information leaflet had significantly improved microalbuminuria
screening uptake (p = 0.032).
Conclusions: An information leaflet can be used to significantly
improve the uptake of urine ACR screening.
P380
Are feet at risk in a diabetic renal clinic
population?
P Singh1, J Hogg2, E Guest2, D Corlis-Jones1 and P De1
1
Department of Diabetes and Endocrinology, Sandwell and West Birmingham
NHS Trust, Birmingham, UK, 2Department of Chiropody, Sandwell and West
Birmingham NHS Trust, Birmingham, UK
Aims: To look at the relationship between various foot ulcer risk

factors in patients with diabetic nephropathy.
Methods: Information was gathered from the chiropodist foot
assessment proforma and electronic data records on all patients
attending our diabetic renal clinic (January 2012 2013).
Results: 48% were Asian, 30% Caucasian and 16%
Afro-Caribbean
103 (57%) males. 84 (51%) patients had
stage 3, 46 (26%) stage 4 and 12 (7%) had stage 5 chronic kidney
disease (CKD). All feet (N = 181) were assessed for risk
low/
intermediate/high. 13% (22/166) patients with CKD 3 vs 23% (26/
112) with CKD 4/5 had absent pulses (p = 0.037). 24% (43/174)
2014 The Authors.

DIABETICMedicine
patients with CKD 3, 17% (29/174) with CKD 4 and 7% (12/174)

with CKD 5 had loss of monofilament sensation. 7% (6/84)
patients with CKD 3 vs 20% (12/58) with CKD 4/5 had callus
(p = 0.017). Foot ulcer was found in three patients with CKD 3 5
and in two patients with any proteinuria. 18% (33/174) patients
with CKD 3, 14% (25/174) with CKD 4 and 3% (6/174) with
CKD 5 were intermediate/high risk. 31% (41/131) patients with
any proteinuria [16% (21/131) with macroalbuminuria] were
intermediate/high risk. There was no relationship of proteinuria
with any foot ulcer risk.
Conclusions: Our findings suggest that CKD stages 4/5 is
significantly more associated with callus formation and absent
pedal pulse (than any other risk factors) in diabetic nephropathy
patients. A thorough feet assessment is required in diabetic renal
clinics due to the potential risk of both neuropathic and vascular
foot ulcer development, especially in those with more advanced
CKD.
P381
Population based survey of
macroalbuminuria in a diabetic screening
programme
A Lawrence-Ball, J Littlewood, PA Meadows and JN Harvey
Wrexham Academic Unit, Bangor University, Wrexham, UK
Aims: Macroalbuminuria (established diabetic nephropathy) is a

serious prognostic risk factor in diabetes. We assessed prevalence,
cause and management of macroalbuminuria identified in a
district-wide diabetic microalbuminuria screening programme.
Methods: Macroalbuminuria was albumin creatinine ratio
(ACR) > 28(men), > 44mg/mmol (women). Adequate blood
pressure (BP) control was defined as BP < 125/80. Inadequate
was >135/85. Measurements between or with only one of systolic
BP/diastolic BP adequate were deemed borderline.
Results: Of 8,768 ACR samples in 2009, 239 patients had
macroalbuminuria (2.7%). Only 191 of these had diabetes: 22
Type 1, 18 having proteinuria due to diabetes (13male, 5female).
169 had Type 2 diabetes with 119 diabetic nephropathy (79male,
40female). Chronic kidney disease stage 3 was present in 4/18
with Type 1 diabetes and 55/119 in Type 2 diabetes. Of the 191
proteinuric patients, 84 were not under an endocrinologist or
nephrologist (2/22 with Type 1 diabetes, 82/169 Type 2 diabetes).
97 were on ACEI or ARB. A further 17 were on both. 50 were on
neither but only 14 for stated clinical reasons. In patients with
diabetic nephropathy BP was good in 26/139, borderline 69,
inadequate 23, unknown 21. Further annual screening showed
improvement in albuminuria in the majority. The mortality rate
was high.
Conclusions: Microalbuminuria screening programmes attract
samples from individuals without diabetes. In diabetes not all
macroproteinuria is due to diabetic nephropathy. Significant
numbers of Type 2 diabetes proteinuric patients were not under
appropriate specialist care. While some were not on ACEI/ARB
therapy for appropriate reasons there is scope for wider usage and
better BP control. A centralised programme can identify individual
cases for more active management.
141
DIABETICMedicine
Clinical care and other categories posters: neuropathy

P382
Hearing loss and diabetes: is there a link?
CL Morrison1, P Morar2, G Morrison3, TS Purewal3 and
PJ Weston3
1
Central Surgery, Pendyffryn Medical Group, Prestatyn, UK, 2Ear, Nose and
Throat Department, Royal Blackburn Hospital, Blackburn, UK, 3Diabetes
Centre, Royal Liverpool University Hospital, Liverpool, UK
Introduction: Studies have shown that diabetes affects hearing

function and the incidence of hearing loss can range between 44%
and 69.7%. Auditory brainstem response wave V latencies are also
statistically significantly longer in people with diabetes. Severity of
hearing loss has been associated in some reports with duration of
diabetes, poor glycaemic control and microvascular complications.
Improvements in metabolic control do not return normal auditory
function.
Method: This study examined the outcomes in patients with Type
2 diabetes aged over 40, referred for audiological assessment due to
self-identified hearing difficulties.
Results: From a large general practice with a diabetes register of
1,428, 107(7.5%) reported hearing problems, nearly twice the rate
seen in the non-diabetic population (499/12,422,4%). Eighty-two
underwent an audiological assessment (mean age 74.4 years, SD
8.5, range 44 94; HbA1c 54.8mmol/mol 1.7,31.2 119.7; male
61%). Otoscopy showed defects in 16 (19.5%), tympanometry
revealed abnormal compliance in 17 (20.1%) and middle
ear pressure in 20(24.4%). Pure tone audiometry identified
sensori-neural hearing loss in 69(84.1%) (mild 13%, moderate
63.8%, severe 18.8%, profound 4.3%). Eight had mixed/
conductive hearing loss and five were normal. A history of
occupational noise exposure was present in 39(47.6%) and
tinnitus in 25(30.5%).
Conclusion: This study has identified a substantial incidence of
hearing loss in patients with diabetes. Hearing impairment can
have a negative impact on the social function of an affected
individual. As patients with diabetes are at a greater risk of
developing hearing loss, audiological tests to monitor auditory
function should be introduced into comprehensive packages of care
provided by diabetes services.
corneal nerve fibre density (CNFD) and length (CNFL) at baseline

and at 2 years. Neuropathy was defined by the Toronto criteria.
Results: At follow-up, neuropathy had developed in eight of 38
(21%) patients without neuropathy at baseline. Only duration of
diabetes (p < 0.001) was significantly greater in those who
developed neuropathy. Age,body mass index, systolic blood
pressure, HbA1c, albumin creatinine ratio (ACR), cholesterol
and triglyceride levels did not differ between the two groups.
PMNCV (p = 0.002) was reduced and VPT (p = 0.008) was
elevated with no difference in NDS (p = 0.13), whilst DBHRV (p
= 0.03), IENFD (p = 0.02), CNFD (p = 0.03), CNFL (p = 0.002)
but not neuropad (p = 0.29) were significantly reduced at baseline
in those who developed diabetic neuropathy.
Conclusions: This prospective study shows that only duration of
diabetes, but not baseline HbA1c, lipids or blood pressure, predicts
the development of neuropathy. Furthermore, PMNCV, VPT,
IENFD, CNFD, CNFL and DBHRV are significant predictors for
the development of diabetic neuropathy. CCM is not only an
important diagnostic test but is also a good prognostic test for
diabetic neuropathy.
P384
Changes in interstitial glucose are not
associated with exacerbation of pain in
subjects with diabetes with painful
neuropathy
R Gouni1,2, M Errikson3, L Gribble3, D Kerr1 and D Coppini2
1
Bournemouth Diabetes Endocrine Centre, Royal Bournemouth Hospital,
Bournemouth, UK, 2Diabetes Department, Poole Hospital NHS Trust, Poole,
UK, 3Clinical Research Centre, University of Surrey, Guildford, UK
Objective: To determine the predictive validity of existing

diagnostic measures of large and small fibre neuropathy in
patients with Type 1 diabetes over a 2 year follow-up.
Methods: Thirty-eight patients with Type 1 diabetes underwent
neuropathy assessment at baseline and after 2 years using
neuropathy disability score (NDS), peroneal motor nerve
conduction velocity (PMNCV), vibration perception threshold
(VPT), deep breathing heart rate variability (DBHRV), neuropad
(%), intra-epidermal nerve fibre density (IENFD) and the novel
technique of corneal confocal microscopy (CCM) to quantify
Aim: Very little is known about the relation between glucose

variability and pain, the impact on daily activities and sleep
disturbance among people living with painful diabetic neuropathy.
Method: Sixty-five subjects with painful diabetic neuropathy
(mean age 65years, mean body mass index 32kg/m2; the majority
have Type 2 diabetes) were recruited as part of a clinical trial
comparing three agents for the treatment of painful neuropathy.
During the drug washout period, regular pain medication was
withdrawn. During the final 2 days of the placebo treatment (days
7 and 8) subjects wore a continuous glucose monitor and
measurements were taken of brief pain inventory, sleep dairies
and sleep quality (Leeds Sleep Quality questionnaire).
Results: There was no relationship between overall glycaemic
burden area (AUC) and severity of pain (r = 0.06, p = 0.6), pain
interfering with activities(r = 0.18, p = 0.14), pain interfering with
sleep(r = 0.09, p = 0.45) and subjective quality of sleep(r = 0.01,
p = 0.8). Similarly glucose variability including (a) mean of glucose
excursions (MAGE), (b) area under the glucose measurement
curve >5.6mol/l and (c) standard deviation of nocturnal
measurements were also not significantly correlated. There was,
however, a higher percentage of nocturnal hypoglycaemia
associated with faster onset of sleep (r = 0.33, p = 0.02).
Conclusion: Neuropathic pain, pain interference with activities
and subjective sleep are not correlated with glucose variability in
this cohort and these are both interesting and unexpected findings.
Hypoglycaemia is also associated with quicker onset of sleep.
These findings highlight the importance of specialised
pharmacological therapies for this condition in addition
to glycaemic improvements.
142
2014 The Authors.

P383
Corneal confocal microscopy: a prognostic
test for the development of diabetic
neuropathy
G Ponirakis, I Petropoulos, HF Fadavi, AU Uazman, OA Asghar,
M Jeziorska, MF Ferdousi, AM Marshall, M Tavakoli and
RAM Malik
Centre for Endocrinology and Diabetes, University of Manchester,
Manchester, UK
Clinical care and other categories posters: neuropathy
P385
Factors influencing diabetes
polyneuropathy: assessment using methods
of small fibre function and structure
S Sharma, P Vas and G Rayman
Diabetes Research Unit, Ipswich Hospital NHS Trust, Ipswich, UK
Aims: A number of factors influence the development of diabetes

polyneuropathy (DPN) including poor glycaemic control, body
mass index (BMI) and triglycerides (TGs). In this study we examine
the relationship between these factors and small fibre function
using the laser Doppler imager (LDI) flare technique as well as
small fibre structure using the corneal confocal microscopy
(CCM).
Methods: In all, 135 subjects with diabetes (DM) and 65 healthy
volunteers (HV) were studied. All participants underwent clinical
evaluation and fasting biochemical testing. CCM was performed
using previously described methods and corneal nerve fibre density
(CNFD) was derived using ACCMetrics software. Small fibre
function was assessed using the modified LDIflare which measures
the size of the axon mediated neurovascular response to foot-skin
heating.
Results: Within the DM cohort, both LDIflare and CCM (CNFD)
significantly correlated with BMI (p < 0.0001 and p < 0.0001
respectively), HbA1c (p = 0.029 and p < 0.0001) and TGs
(p < 0.0001 and p < 0.0001). However, within the HV group,
the correlations were only with BMI (p = 0.015 and p = 0.005)
and TGs (p = 0.008 and p = 0.004) but not with HbA1c (p = 0.50
and p = 0.112). In both groups, no relationship was observed with
total cholesterol and in the diabetic group with duration of
diabetes.
Conclusions: Our study demonstrates that despite assessing
different aspects of neural integrity and in different anatomical
sites, the LDIflare and CCM methods demonstrate excellent
correlations with the same factors influencing DPN. Although
this suggests that they may be interchangeable, longitudinal studies
are needed to determine whether short-term changes of these
factors influence both techniques to the same degree.
P386
Socioeconomic deprivation independently
predicts symptomatic treated painful
diabetic neuropathy in Type 2 diabetes
AH Heald1,3, SG Anderson2, RP Narayanan1,3, N Malipatil1,3,
H Roberts1 and G Dunn4
1
Department of Medicine, Leighton Hospital, Crewe, UK, 2School of Medicine

and Manchester Academic Health Sciences Centre, University of Manchester,
Manchester, UK, 3Cardiovascular Sciences Research Group, Core Technology
Facility, University of Manchester, Manchester, UK, 4Department of Podiatry,
East Cheshire NHS Trust, Macclesfield, UK
Aims: A greater understanding of predisposing factors and

prescribing can facilitate more effective resource allocation for
diabetic neuropathy.
Methods: The Townsend index of deprivation (numerically
higher for greater disadvantage) was examined in the
pseudonymised records of 15,387 individuals (6,770 females,
44%) with Type 2 diabetes and related to prevalence of drug
treated severe diabetes related neuropathic pain. Age range
16 101 years.
Results: Treatment for neuropathic pain was initiated in 3,232
(21%) of patients, who were older at 68.2 (95% CI 67.8 68.7) vs
66.6 (66.4 66.8) years and had greater body mass index[32.2
2014 The Authors.

DIABETICMedicine
(32.0 32.4) vs 31.1 (30.9 31.2)kg/m2; p < 0.0001] than patients
not on treatment for neuropathic pain. There was no difference in
HbA1c at 57mmol/mol. eGFR was lower in the neuropathic pain
group at 70.0 (69.4 70.6) vs 73.4 (73.1 73.7)ml/min/1.73m2.
Amitryptiline was the most commonly prescribed agent (1,719/
3,232). Duloxetine (60mg daily) was prescribed in 157/3,232 of
cases. There were significant differences between groups for the
Townsend index, with a greater proportion (30.6% vs 22.8%) of
patients with treated neuropathic pain having a score of 1
(v2 = 83.9, p < 0.0001). Multivariate logistic regression analyses
indicated that each unit increment in the Townsend index was
associated with a 16% increased odds of requiring neuropathic
pain treatment (odds ratio 1.16, 95%CI 1.12 1.20, p < 0.001).
This was independent of age, HbA1c, gender, diabetes duration,
systolic blood pressure and eGFR.
Conclusion: Amitriptyline was the most commonly used agent for
treatment of diabetes related neuropathic pain with duloxetine
much less used. A higher level of socioeconomic deprivation may
predispose to severe neuropathic pain in diabetes. Differential
allocation of resources may benefit this group.
P387
Abnormal small nerve fibre function
demonstrated by the laser Doppler imaging
(LDI) flare technique in children and young
people with Type 1 diabetes
J Hassler-Hurst, S Sharma and G Rayman
Objective: To modify the LDIflare method for use in children and

to use it to compare small fibre function in children with and
without diabetes.
Methods: The modified LDIflare method involves stepwise
heating of foot skin (45C2min, 46C2min, 47C30s (compared
with 3min in adults) and 46C2min) to evoke the neurogenic flare
and measuring its size with a laser Doppler imager (LDI). Young
people aged 12 16years in the following groups were studied:
healthy controls (HC) n = 18; Type 1 diabetes n = 26, 18 with and
eight without microvascular complications.
Results: Children with diabetes had significantly smaller LDIflares
(5.9 2.6vs 9.3 4.0cm2; p = 0.002). Due to small numbers
there was no difference between the diabetes groups. LDIflare size
inversely correlated with all measurements of HbA1c, i.e. mean
HbA1c in last 12 months, r = 0.60, p = 0.001; mean HbA1c
over total diabetes duration, r = 0.51, p = 0.008; and most
recent HbA1c, r = 0.473, p = 0.015. LDIflare size correlated
with HbA1c percentage change over 2 years, i.e. the mean of all
results for the previous year to the mean of all results for the most
recent year, r = 0.49, p = 0.015. The procedure was rated as
tolerable and the majority of the children found the study
interesting.
Conclusions: The LDIflare method is suitable for use in children
and demonstrates reduced small nerve function in this group. Its
relationship with HbA1c and HbA1c change at this early stage of
diabetes suggests a degree of plasticity (reversibility). Of potential
importance, children were engaged by the images suggesting that it
could be a potential motivational tool to improve glycaemic
control.
143
DIABETICMedicine
P388
P389
Sudoscan: a simple, rapid and objective

method of screening of diabetic neuropathy
A new hand-held nerve conduction

measuring device for the accurate diagnosis
of diabetic peripheral neuropathy
P Shillo1, M Greig1, G Rao3, RA Gandhi1, S Tesfaye1 and

D Selvarajah1,2
1
Department of Diabetes, Sheffield Teaching Hospitals NHS Trust, Sheffield,

UK, 2Human Metabolism, University of Sheffield, Sheffield, UK,
3
Neurophysiology, Sheffield Teaching Hospitals NHS Trust, Sheffield, UK
Aims: To evaluate if Sudoscan, a new method developed to

provide a quick, non-invasive and quantitative assessment of
sudomotor function, can reliably screen for diabetic peripheral
neuropathy (DN).
Methods: Thirty-six subjects with Type 1 diabetes underwent
peripheral
nerve
assessments
including
clinical
and
neurophysiological tests (vibration detection thresholds, sural,
common peroneal and tibial nerve conduction studies). Based on
these the overall neuropathy composite score (NCS) was calculated
and subjects were divided into DN and No-DN. Sudomotor
function was assessed with measurement of hand and foot
electrochemical sweat conductance (ESC).
Results: There was no significant difference in age and duration of
diabetes between the two groups. DN subjects [2.1(1.4)] had
significantly lower NCS compared with No-DN subjects [11.1
(6.5); p < 0.001]. ESC hands [mS, 49.0(21.1) vs 64.6(14.5);
p = 0.03] and feet [53.8(30.0) vs 72.3(15.8); p = 0.05) were
significantly lower in the DN group compared with No-DN. Both
ESC hands and feet significantly correlated with overall NCS
but only ESC feet demonstrated significant correlation with all
nerve conduction attributes measured (sural amplitude r = 0.55,
p = 0.01; common peroneal amplitude r = 0.44, p = 0.04; velocity
r = 0.54, p = 0.01, and latency r = 0.50, p = 0.03; and tibial
latency r = 0.58, p = 0.01). Receiver operating characteristic
(ROC) area under the curve values for ESC feet and hands were
similar [0.76 (0.56 0.96) vs 0.78 (0.58 0.98)].
Conclusion: Sudoscan, a non-invasive and quick test, could be
used as a screening test for DN in the context of a busy diabetic
clinic. This preliminary study demonstrates good concurrent
validity of Sudoscan against the current gold standard for
diagnosing DN. It also highlights the ability of Sudoscan to
classify patients with DN.
M Greig, P Shillo, G Rao, RA Gandhi, S Tesfaye and D Selvarajah

Diabetes and Endocrinology, Sheffield Teaching Hospitals, Sheffield, UK
Aims: The diagnosis of diabetic neuropathy (DPN) relies on

peripheral neurological examination using the tuning fork, the 10g
monofilament and the assessment of pin prick sensation. A
hand-held device, the NC-STATDPN-CheckTM(NeuroMetrix,
Waltham,Massachusetts, USA) was compared with the gold
standard conventional electrophysiology equipment (CEE).
Methods: Thirty-one subjects [11 with diabetes (8 DPN and 3
No-DPN) and 20 healthy volunteers(HV)] underwent peripheral
nerve assessments including vibration detection thresholds using
CASE IV and sural, common peroneal and tibial nerve conduction
studies using CEE. The DPN was diagnosed using the neuropathy
composite score(NCS). Each subject also had SSCV and SA measured
antidromically using the NC-STAT DPN-Check. The performance of
NC-STAT was compared with the 10g monofilament test.
Results: NCS was significantly higher in the diabetes group[6.55
(5.0)] compared with HV[0.8(1.4);p < 0.0001]. There was a
positive correlation between NC-STAT DPN-Check measured
SSCV(r = 0.82;p < 0.0001) and SA(r = 0.47;p = 0.007) compared
with traditional methods of assessments. Area under the Receiver
operating characteristic (ROC) curve values for NC-STAT
DPN-Check measured SA [0.88(0.71 0.97)] and velocity[0.91
(0.75 0.98)] were similar to the 10g monofilament test [0.83
(0.66 0.94)] for diagnosing DPN. However, sensitivity and
specificity of the device(SAsensitivity 88.9, specificity 86.4;
SSCVsensitivity 88.9, specificity 90.9) are more favourable compared
with the 10g monofilament (sensitivity 66.7, specificity 100).
Conclusions: The NC-STAT DPN-Check is a quick, simple,
objective and non-invasive hand-held device that provides an
accurate measure of peripheral nerve function. It could be the
preferred method of screening for DPN in the context of a busy
diabetic clinic as it is more sensitive than the 10g monofilament test
and has an excellent agreement with the gold standard CEE for
diagnosing DPN.
Clinical care and other categories posters: new therapies

P390
Reductions in glycaemic variability and
hypoglycaemia following islet
transplantation: studies using continuous
glucose monitoring systems within the first
month post-transplant in the Scottish islet
transplant programme
J Barclay1,2, S Chen3, G Brooks4, D Anderson1,2, K Duncan1,2,
T McGilvray2, J Davidson2, J Casey2,3 and S Forbes1,2,3
1
Department of Diabetes, Royal Infirmary of Edinburgh, Edinburgh, UK,

Transplant Unit, Royal Infirmary of Edinburgh, Edinburgh, UK,
3
Endocrinology, University of Edinburgh, Edinburgh, UK, 4Endocrinology,
Newcastle University, Newcastle-upon-Tyne, UK
2
Background: Subjects with Type 1 diabetes with impaired

awareness of hypoglycaemia may be eligible for islet
144
transplantation. These subjects often have glycaemic lability

which is not captured by conventional measures of glycaemic
control such as the HbA1c. Continuous glucose monitoring
systems (CGMS) sample the interstitial glucose reflecting blood
glucose measurements at 3min intervals over 6 day periods. The
immediate post-transplant period has not been well studied.
Aims: To observe glycaemic lability and duration of
hypoglycaemia pre-transplant, peri-transplant and up to 1 month
post-transplant.
Methods: Data on 14 patients with normal renal function and
functioning grafts are reported. All subjects wore the Ipro-2 or
MiniLink for a period of 6 days: pre-transplant, peri-transplant
(days 1 to 7 post-transplant), 1, 2, 3 weeks and 1 month
post-transplant. The mean amplitude of glycaemic excursion
(MAGE) at each stage were calculated; insulin requirements were
recorded pre-transplant and at 1 month post-transplant.
2014 The Authors.

Clinical care and other categories posters: pregnancy
Results: Recipient age at assessment pre-transplant was, median

(interquartile range), 45 (35 55) years, weight 70.7 (64.8 83.4)
kg, body mass index 26.3 (23.7 29.8)kg/m2, insulin requirements
0.53 (0.47 0.57) units/kg. MAGE was 7(5 9)mmol/l and
duration of hypoglycaemia 2.7 (0.02 0.10)%. Following islet
transplantation MAGE improved: peri-transplant 4(3 5)mmol/l;
1 week 3(2 5)mmol/l; 2 weeks 3(3 4)mmol/l; 3 weeks 5(4 7)
mmol/l and 1 month 4(3 6)mmol/l (p < 0.05), with significant
reductions in hypoglycaemia by 2 weeks post-transplant [0.1
(0 0.4)%; p < 0.05] and continued improvements thereafter. At 1
month insulin requirements were significantly reduced, 0.24
(0.07 0.40) units/kg.
Conclusion: Frequency of hypoglycaemia and glycaemic lability
improve rapidly with reductions in insulin requirements within the
first month following islet transplantation.
P391
Achievement of weight loss with
canagliflozin, sitagliptin and glimepiride in
combination with metformin or
metformin+sulfonylurea in patients with
Type 2 diabetes
L Stein1, G Thompson2, J Diels3, P Thilakarathne3 and I Girod2
1
Medical Affairs, Janssen-Cilag Ltd, High Wycombe, UK, 2Outcomes
Research, Janssen-Cilag Ltd, High Wycombe, UK, 3HEMAR EMEA, Janssen,
Beerse, Belgium
Objectives: Treating
Type
2
diabetes
with
existing
anti-hyperglycaemic agents (AHAs) is frequently accompanied by
weight gain that can increase the risk of cardiovascular diseases.
This analysis compares weight change achieved with three AHAs
in combination with metformin (MET): the sodium glucose
co-transporter-2
inhibitor
canagliflozin,
the
dipeptidyl
peptidase-4 inhibitor sitagliptin and the sulfonylurea (SU)
glimepiride. Canagliflozin was also compared with sitagliptin in
combination with MET+SU.
Methods: A meta-analysis and an indirect comparison, using
patient level data from four randomised trials (modified
intent-to-treat, n = 3,957), were conducted using a logistic
regression analysis comparing achievement of weight loss 3%
after 52week treatment with canagliflozin 100mg or 300mg,
glimepiride titrated from 1mg to 6 8mg/day (plus MET), or
sitagliptin 100mg (plus MET or MET+SU). A generalised
estimating equation (GEE) model was used to capture correlation
of subjects within trials, with treatment, baseline bodyweight
and the interaction between both as covariates.
Results: In patients on MET (average baseline weight 86.9kg), a
weight loss 3% was achieved in 62.4% (95% CI 59.9%, 64.8%)
DIABETICMedicine
and 55.2% (95% CI 51.3%, 58.9%) of patients on canagliflozin

300mg and 100mg respectively, compared with 24.6% (95% CI
23.9%, 25.2%) for sitagliptin and 11.3% (95% CI 10.7%,
11.9%) for glimepiride. In patients on MET+SU (average baseline
weight 90.0kg), 46.1% (95% CI 43.2%, 48.9%) of patients treated
with canagliflozin 300mg achieved a weight loss 3%, compared
with 35.6% (95% CI 34.8%, 36.3%) with canagliflozin 100mg
and 15.6% (95% CI 15.2%, 16.1%) with sitagliptin.
Conclusions: Type 2 diabetes patients are more likely to achieve
weight loss 3% on a MET background with both doses of
canagliflozin compared with sitagliptin or glimepiride. Similar
results were found vs sitagliptin in triple therapy (in combination
with MET+SU).
P392
A novel scoring system for assessing beta
cell function after clinical islet
transplantation (CIT)
A Smith1, S Imes2, A Malcolm2, PA Senior2 and AMJ Shapiro2
1
2
School of Medicine and Dentistry, University of Aberdeen, Aberdeen, UK,

Clinical Islet Transplant Program, University of Alberta, Edmonton, Canada
Aims: There are no validated tools to monitor graft function after

CIT. Previously, the beta score was developed for this purpose but
assigns categorical scores to continuous variables. We sought to
refine the beta score using all parameters as continuous variables.
Methods: Several model formulae were tested based on the
expected relationship between graft function and components of
the beta score (fasting c-peptide, fasting blood glucose, HbA1c and
insulin dose). The final formula was tested in the cohort used to
derive the beta score (n = 53, mean age 42.1 1.3 years; diabetes
duration 26.1 1.3 years; evaluated at 3, 6 and 12 months and then
yearly up to a maximum of 5 years). The new score was compared
with the beta score, stimulated unit of islet transplant objects
(SUITO) index and stimulated glucose levels following mixed meal
tolerance test (MMTT). Subgroup analysis was performed on 10
patients with follow-up to the present time (up to 13 years).
Results: The range for the new score was from 0 to 42 (mean 18).
The new score was highly correlated with beta score (r = 0.80),
SUITO index (0.92) and stimulated glucose levels after MMTT
(r = 0.71) (p < 0.01 for all). In the subgroup, the new score was
higher in those patients remaining insulin independent than in
those resuming insulin (S = 24 vs 12, p < 0.01).
Summary: This novel scoring system, incorporating relevant
parameters in a continuous fashion, shows potential as a tool to
monitor graft function following CIT. Its utility to predict
outcomes will be tested in future studies.

P393
Diabetes recognised in early pregnancy is
phenotypically Type 2 diabetes not
gestational diabetes
R Agha-Jaffar1, M Reddy1, GE Zhang2, L Phelan2, TG Teoh2,
C Yu2, D Gable1, NS Oliver1 and S Robinson1
1
Department of Metabolic Medicine, Imperial College, London, UK,

2
Department of Obstetrics and Gynaecology, Imperial College, London, UK
Refer to Oral A33
2014 The Authors.

P394
Fasting plasma glucose (FPG) alone will miss
persisting postnatal glucose abnormalities
in gestational diabetes
H Venkataraman1, N Sukumar1, C Wood2, N Anderson2 and
P Saravanan1,3
1
Metabolic and Vascular Health, Warwick Medical School, Coventry, UK,

Pathology, George Eliot Hospital, Nuneaton, UK, 3Diabetes and
Endocrinology, George Eliot Hospital, Nuneaton, UK
Refer to Oral A34
145
DIABETICMedicine
P395
Do differences in treatment for gestational
diabetes reflect/produce differences in
perinatal outcome? An audit of antenatal
care involving dietary advice alone, insulin
and/or metformin
DL Ashmore, G Blakey, L Brooke and H Panayiotou
1
Leeds Institute of Genetics, Health and Therapeutics, University of Leeds,
Leeds, UK
Refer to Oral A35
P396
Screening for diabetic retinopathy in
pregnant women with Type 1 and Type 2
diabetes: are guidelines being met?
L McVicker1, L Carmody2, O Kgosidialwa2, A Egan1,2 and
F Dunne1,2
1
Department of Medicine, National University of Ireland Galway, Galway,
Ireland, 2Galway Diabetes Research Centre, National University of Ireland
Galway, Galway, Ireland
Refer to Oral A36
P397
Association of continuous subcutaneous
insulin infusion (CSII) during pregnancy
with pregnancy related outcome and its
relationship with microvascular
complications
MS Ahmed, A Tahrani, S Ateeq, S Jones, H Buckley, P Dyer,
A Field, J Hand and M Karamat
Diabetes and Endocrinology Department, The Heart of England NHS
Foundation Trust, Birmingham, UK
Aims: To assess the impact of CSII on pregnancy outcome,

glycaemic control and microvascular complications in women with
Type 1 diabetes.
Methods: We examined all pregnant women who received CSII
treatment. Patients were divided into two groups: group A included
pregnancies in which CSII was started during pregnancy (n = 9);
group B included pregnancies in which CSII was started before
pregnancy (n = 12). Diabetic retinopathy (DR) was assessed using
retinal images acquired from the retinal screening programme.
Diabetic nephropathy was assessed using urinary albumin
creatinine ratio (ACR),in pregnancies that went through the three
trimesters(n = 8 and 10) in groups A and B respectively. Successful
outcome was defined as delivering a live birth.
Results: In group A 8/9 pregnancies (89%)and in group B 11/
12 pregnancies (92%) had a successful outcome. HbA1c improved
in both groups throughout pregnancy (group A 7.0% 0.3 vs
6.3% 0.4 vs 6.1% 0.5; group B 6.7% 0.5 vs 6.4% 0.8
vs 6.4% 0.8, for the first, second and third trimester respectively).
In group A, 4/8 (50%) patients developed sight-threatening DR
compared with 3/10 (30%) in group B, most of which were due to
the development of maculopathy. Any DR progression occurred in 7/
8 (87%) in group A vs7/10 (70%) in group B. ACR values were
normal in all patients in both groups throughout pregnancy.
Conclusions: CSII in pregnancy was associated with the
development of sight-threatening DR. This was greater in
patients started on CSII during pregnancy compared with
those who started before pregnancy,which might be related to
146
the greater decline in HbA1c observed in the first group. The safety
and benefits of using CSII in pregnancy need to be examined in
randomised clinical trials.
P398
Twin pregnancies in mothers with
pre-gestational diabetes: regional
population based cohort study
J Darke, SV Glinianaia and R Bell
Institute of Health and Society, Newcastle University, Newcastle-upon-Tyne, UK
Aim: To compare maternal characteristics and pregnancy

outcomes between twin and singleton pregnancies of mothers
with pre-gestational diabetes.
Methods: We analysed data on 32 twin and 2,264 singleton
pregnancies from 1996 to 2008 from the Northern Diabetes
in Pregnancy survey (NorDIP). Maternal characteristics and
pregnancy outcomes were compared between twin and singleton
pregnancies. Adjusted odds ratios (OR) with 95% confidence
intervals (CI) for combined adverse pregnancy outcome (fetal
loss < 24 weeks gestation, stillbirth, infant death or major congenital
anomaly) and extended perinatal mortality rate (EPMR) (still births
and neonatal deaths) were calculated using logistic regression.
Results: Twin mothers were more likely to be older (p = 0.017),
have Type 2 diabetes (p = 0.022), a higher body mass index (BMI)
(p = 0.003), attend pre-conceptual care (p = 0.006) and have lower
peri-conception HbA1c (p = 0.026) than singleton mothers. Twins
were more likely to be delivered by C-section (p = 0.001), pre-term
(p < 0.0001), with low birthweight (p < 0.0001) and require special
care baby unit admission (p = 0.041) compared with singletons.
The EPMR was significantly higher in twins than singletons
(RR = 2.84, 95% CI 1.19 6.77). 266/1,000 twin offspring had
an adverse pregnancy outcome. Twinning was not independently
associated with increased odds of adverse pregnancy outcome or
EPMR when adjusted for confounders. Higher peri-conception
HbA1c (OR 1.02, 95% CI 1.01 1.03), pre-pregnancy nephropathy
(OR 2.18, 95% CI 1.07 4.43) and no pre-conception folate (OR
1.51, 95% CI 1.01 2.24) independently predicted adverse
pregnancy outcome for twins and singletons.
Conclusion: Twinning is a less important predictor of adverse
pregnancy outcome in women with diabetes than diabetes-related
factors, but specific guidance is lacking in management of these
high risk pregnancies.
P399
Tele-Mum: a feasibility study for a
randomised controlled trial to explore the
potential of using telemedicine for the
diabetes care of those with gestational
diabetes
JE Given1, M OKane2, BP Bunting3, F Dunne4,5, K Moles6,
L King6, E McIvor6, T Gallacher7 and V Coates1,8
1
Institute of Nursing and Health Research, University of Ulster, Coleraine, UK,

Clinical Chemistry, Western Health and Social Care Trust, Londonderry, UK,
Psychology Research Institute, University of Ulster, Londonderry, UK, 4School
of Medicine, National University of Ireland, Galway, Ireland, 5Endocrinology,
Galway University Hospital, Galway, Ireland, 6Acute Care, Altnagelvin Area
Hospital, Londonderry, UK, 7Maternity, Letterkenny General Hospital,
Letterkenny, Ireland, 8Nursing Directorate, Western Health and Social Care
Trust, Londonderry, UK
2
3
Aim: To determine the feasibility and acceptability of using

telemedicine in the diabetes care of women with gestational
2014 The Authors.

diabetes (GD) and the possibility of replacing alternate diabetes

review appointments with telemedicine.
Methods: Fifty women with GD were randomised to usual care
(n = 26) or usual care plus telemedicine (n = 24), across two sites.
Telemedicine entailed weekly blood pressure and weight
measurements and transmission of these data, along with blood
glucose readings, for review by the healthcare team. Patients were
contacted as necessary. Management decisions made, and the time
taken to make them, were recorded for the intervention group at
each face-to-face and telemedicine review. Patients completed
questionnaires to measure their satisfaction with telemedicine or
blood glucose monitoring. Intervention group patients and
healthcare providers took part in qualitative interviews. Analysis
entailed descriptive statistics (satisfaction questionnaires), Cohens
kappa (management decisions) and framework analysis
(qualitative interviews).
Results: Telemedicine review took less time than face-to-face
review (mean 4.6 vs 11.7min) and allowed comparable
management decisions to be made (kappa was 0.54 for
telemedicine and face-to-face agreement, 0.65 for inter-rater
agreement and 0.44 0.77 for intra-rater agreement).
Eighty-seven percent of patients were satisfied with telemedicine
and would use it again. It was convenient and saved patients time
and money. Staff and patients found the equipment easy to use and
were positive about using it to replace alternate diabetes review
appointments in the future.
Conclusion: Telemedicine may help meet the growing demand on
diabetes services due to increasing numbers of women being
diagnosed with GD.
P400
GDm-Health: a pilot study examining
acceptability of mobile phone assisted
remote blood glucose monitoring for
women with gestational diabetes
L Loerup1, JE Hirst2, LH MacKillop2,3, A Farmer4, DA Kevat3,
L Tarassenko1, OJ Gibson1, K Bartlett3, YM Kenworthy2 and
J Levy5
1
Institute of Biomedical Engineering, University of Oxford, Oxford, UK,

Nuffield Department of Obstetrics and Gynaecology, University of Oxford,
Oxford, UK, 3Silver Star High Risk Pregnancy Unit, Oxford University Hospitals
NHS Trust, Oxford, UK, 4Primary Care Health Sciences, University of Oxford,
Oxford, UK, 5Oxford Centre for Diabetes, Endocrinology and Metabolism,
University of Oxford and OUH NHS Trust, Oxford, UK
2
Background: The increase in gestational diabetes (GD) is

challenging maternity diabetes services. Remote monitoring of
glycaemic control could potentially reduce outpatient encounters
for
well-controlled
women.
We
have
developed
a
smartphone-facilitated, remote blood glucose (BG) monitoring
system:GDm-Health. BG readings transmit via Bluetooth to a
smartphone app. Results are then labelled and transmitted via the
mobile phone data network (3G) to a secure website for clinician
review and feedback.
Aim: To determine womens views on GDm-Health and their
relationship with the diabetes team.
Methods: In a pilot study of 52 women (September 2012 to June
2013), BG was monitored using GDm-Health from diagnosis
until delivery. Following birth, women completed a structured
questionnaire assessing (i) general satisfaction, (ii) equipment, (iii)
relationship with the diabetes care team. Responses were scored on
a seven-point Likert scale with comments.
Results: Forty-nine of 52 women completed the questionnaire.
Thirty-two had new diagnoses of GD [gestation at recruitment 29
2014 The Authors.

DIABETICMedicine
4 weeks (mean SD)] and 17 had previous GD (18 6

weeks).45/49 women agreed their care was satisfactory and the
best for them. 47/49 and 43/49 agreed the equipment was
convenient and reliable respectively. 42/49 agreed GDm-Health
monitoring fitted into their lifestyle. 46/49 agreed they had a good
relationship with their care team. Written comments supported
these findings, with very positive reactions from the majority of
women.
Conclusion: This pilot demonstrates that GDm-Health is
acceptable and convenient for women. Effects on clinical and
economic outcomes are currently under investigation in a
randomised trial (clinicaltrials.gov NCT01916694).
P401
Investigating hypoglycaemia awareness in
Type 1 diabetes during pregnancy
L Vermont1, SS Hussain1, SJ Lam2, S Jarvis1, F Cheng2,
C Jairam1, B Jones2 and A Dornhorst1
1
Department of Diabetes and Endocrinology, Imperial College Healthcare
NHS Trust, London, UK, 2Department of Obstetrics and Gynaecology,
Imperial College Healthcare NHS Trust, London, UK
Aims: This prospective observational study aimed to determine

the prevalence of impaired awareness of hypoglycaemia (IAH) and
frequency of hypoglycaemia in pregnant patients with Type 1
diabetes.
Methods: IAH was assessed using three validated questionnaires
(Clarke, Gold and Pederson). These questionnaires were
amalgamated into a single questionnaire which also incorporated
the Edinburgh Hypoglycaemic Score, questions on causes and
worry for hypoglycaemia scored on a seven-point Likert scale.
Questionnaires, self-monitored blood glucose diaries and
Diasend meter downloads were collected from the pregnant
patients on a fortnightly basis.
Results: Eight pregnant (baseline mean SEM: age 28 1.24
years, duration of diabetes 12.5 3 years, HbA1c
65 7.88mmol/mol) and 22 matched non-pregnant female (age
30 1.23 years, duration of diabetes 16.32 2 years, HbA1c
70.77 4.26mmol/mol) patients with Type 1 diabetes were
recruited. In keeping with previous observations, there was
increased prevalence of IAH in pregnant patients compared with
non-pregnant subjects. There was no significant difference in severe
hypoglycaemia or proportion of blood glucose readings below
3.5mmol/l during pregnancy between IAH and aware patients.
Pregnant subjects were significantly more worried about nocturnal
hypoglycaemia than daytime hypoglycaemia (p = 0.007). Patients
with IAH felt more worried about hypoglycaemia compared with
aware patients (p = 0.04). Comparison of blood glucose diaries
and meter downloads highlighted under-reporting of
hypoglycaemic events in diaries.
Conclusions: This report demonstrates increased prevalence of
IAH during pregnancy. Further research is required to fully assess
whether IAH assessed via questionnaires in clinic can predict
higher incidence of hypoglycaemia or severe hypoglycaemia in
patients. This study also highlights the potential for improved
detection of hypoglycaemic episodes using technology.
147
DIABETICMedicine
P402
Nutrient intake of pregnant women with
Type 1 diabetes in the DAPIT trial:
relationships with anthropometry and
glycaemic control
AJ Hill1, C Patterson2, VA Holmes2, IS Young2 and
DR McCance3
1
Northern Ireland Centre for Food and Health, University of Ulster, Coleraine,
UK, 2Centre for Public Health, Queens University, Belfast, UK, 3Regional
Centre for Endocrinology and Diabetes, Royal Victoria Hospital, Belfast, UK
Aims: To examine the nutritional composition of diets of

pregnant UK women with Type 1 diabetes in the second
trimester in relation to body mass index(BMI) and glycaemia.
Methods: Women were participants in a randomised controlled
trial of antioxidant supplementation to prevent pre-eclampsia
(DAPIT [1]). Diet was assessed using a validated semi-quantitative
food frequency questionnaire at 26 weeks gestation. Mean daily
nutrient intakes were analysed using Q Builder nutritional software
(Tinuviel Software, UK) and SPSS version 20.
Results: Data were available for 597 women (78% of cohort),
mean age 30 (SD 5.4) years, with 14.5 (SD 8.0) years diabetes
duration and mean HbA1c at booking 7.8% (SD 1.4) (IFCC
61.7mmol/mol). Mean BMI at booking was significantly higher
than that of the background pregnant population2 27 (4.5)kg/m vs
24.2 (4.5)kg/m (p < 0.01) with 63% of women overweight or
obese, which is significantly higher than the background
population [2 4].Overall women reported adhering to diets high
in carbohydrate and fibre, average in glycaemic index and
glycaemic load, but low in fat and energy (mean 7,009 (2,161)
kJ). No relationship was observed between carbohydrate intake
and measures of glycaemia. While women with higher BMI
(>30kg/m) reported consuming less energy they required more
insulin at 28 weeks compared with normal weight women
(p < 0.01).
Conclusion: Obese women reported consuming lower energy
intake but were more likely to under-report. Obese women gained
more weight than recommended by IOM [4] but required more
insulin. Weight management in pregnancy may therefore have
more impact on glycaemic control than adjusting dietary intake.
References: 1. McCance DR, Holmes VA, Maresh MJ et al.
Lancet 2010; 376: 259 266.
2. Heslehurst N, Ells LJ, Simpson H et al. Br J Obstet Gynecol
2007; 114: 187 194.
3. Scott-Pillai et al. Br J Obstet Gynecol 2013; 120: 932 939.
4. IOM (Institute of Medicine) and NRC (National Research
Council). 2009. Weight Gain During Pregnancy: Reexamining the
Guidelines. Washington, DC: The National Academies Press.
P403
Gestation of stillbirths in women with
pre-gestational diabetes
N Holman1, M Maresh2, R Bell3, R Holt4, D Todd5, J Hawdon6,
M Morgan7, B Young8, N Lewis-Barned9 and H Murphy10
1
National Cardio-Vascular Intelligence Network, Public Health England, York,
UK, 2Central Manchester University Hospitals NHS Foundation Trust,
Manchester, UK, 3Institute of Health and Society, Newcastle University,
Newcastle-upon-Tyne, UK, 4Human Development and Health Academic Unit,
5
University Hospitals Leicester, Leicester, UK, 6Barts Health NHS Trust,
London, UK, 7Singleton Hospital, Swansea, UK, 8Diabetes and Endocrinology,
Salford Royal Hospital, Salford, UK, 9Northumbria Healthcare Foundation NHS
Trust, North Shields, UK, 10Institute of Metabolic Science, University of
Cambridge, Cambridge, UK
Aim: Women with pre-gestational diabetes have a significantly

increased risk of stillbirth. Current guidelines recommend elective
delivery at 38 completed weeks. We aimed to examine the
gestational age of stillbirths in a large UK cohort.
Methods: Data were collected on all pregnancies in women with
pre-gestational diabetes delivered in 2007 or 2008 in three regional
audits (Northern, North West and East Anglia) and all pregnancies
delivering between June 2010 and May 2011 in 13 units and two
regional (Northern and East Anglia) audits. The gestation of
stillbirths were identified and compared with data for all stillbirths
in England from the Office for National Statistics.
Results: The cohort included 2,370 singleton pregnancies, of
which 2,066 continued beyond 24 weeks. Of 29 stillbirths, nine
(31.0%) occurred between 24 and 31 weeks gestation and nine
(31.0%) occurred between 32 and 35 weeks. This compares with
43.1% and 17.7% of all stillbirths in England respectively. There
were two (6.9%) at 36 weeks gestation and five (17.2%) stillbirths
at 37 weeks gestation. Only four (13.8%) stillbirths occurred after
38 weeks gestation. The proportion of stillbirths occurring after
36 weeks gestation is similar for women with diabetes and for
England (37.9% and 39.1% respectively).
Conclusion: A smaller proportion of stillbirths among women
with diabetes occur between 24 and 31 weeks gestation than in all
stillbirths in England but a greater proportion occur between 32
and 39 weeks. However, this is partly a reflection of early
induction of women with diabetes.
P404
GDm-Health: a pilot study demonstrating
the feasibility of mobile phone assisted
treatment advice and medication
adjustment for women with gestational
diabetes
L Loerup1, OJ Gibson1, JE Hirst2, AJ Farmer3, KJ Bartlett4,
YM Kenworthy4, JE Blincowe5, DA Kevat4, LH MacKillop2,4,
L Tarassenko1 and JC Levy6
1
Institute of Biomedical Engineering, University of Oxford, Oxford, UK,

Nuffield Department of Obstetrics and Gynaecology, University of Oxford,
Oxford, UK, 3Department of Primary Care Health Sciences, University of
Oxford, Oxford, UK, 4Silver Star High Risk Pregnancy Unit, Oxford University
Hospitals NHS Trust, Oxford, UK, 5Horton Maternity Unit, Oxford University
Hospitals NHS Trust, Banbury, UK, 6Oxford Centre for Diabetes,
Endocrinology and Metabolism, University of Oxford and Oxford University
Hospitals NHS Trust, Oxford, UK
2
Background: Increasing diagnoses of gestational diabetes (GD)

are challenging maternity diabetes services. We have developed a
smartphone assisted blood glucose (BG) monitoring system
148
2014 The Authors.

(GDm-Health) aiming to reduce clinic visits for well controlled

women. BG monitors transmit via Bluetooth to a smartphone app.
Results are transmitted (via 3G) to a secure website for clinician
review and feedback including treatment modifications via phone
calls and text messages.
Aim: In a pilot cohort observational study, to determine whether
GDm-Health prevents hyperglycaemia in women with de novo
diagnosed (nGD) and previous pregnancy GD (pGD).
Methods: Six-point BG profiles were monitored and 1 week blood
glucose means were analysed at recruitment and before delivery.
Results: We recruited 28 nGD [mean (SD) gestation at
recruitment 30 (3) weeks; at delivery 38 (1) weeks] and 17 pGD
[18 (6) weeks; 39 (1) weeks]. Four women with atypical
presentations of GD were excluded from this analysis. At
recruitment, none of the women was receiving any form of
glucose-lowering treatment. Prescribed treatments at delivery were
diet only, metformin, mealtime insulin and background insulin,
respectively, in nGD (21%, 43%, 57%, 29%) and in pGD (24%,
47%, 41%, 12%). BG means (SD) at recruitment and delivery
were 6.2 (0.7) and 5.8 (0.6)mmol/l (p < 0.001) in nGD, and 5.5
(0.7) and 5.4 (0.6)mmol/l (p = 0.67) in pGD.
Conclusion: Smartphone assisted monitoring with GDm-Health
improved control in the nGDgroup and prevented development of
hyperglycaemia in the pGD group. Effects on clinical and
economic outcomes are currently under investigation in a
randomised trial (http://clinicaltrials.gov/, NCT01916694).
P405
ATLANTIC DIP (diabetes in
pregnancy): postpartum screening in
women with gestational diabetes, who
attends and why?
L Carmody1, A Egan1, B Kirwan1, A ODea2 and F Dunne1,2
1
2
Galway Diabetes Research Centre, University Hospital, Galway, Ireland,

School of Medicine, National University of Ireland Galway, Galway, Ireland
Background: The rate of gestational diabetes (GD) in Irish women

is 12.4%. GD is a well recognised risk factor for pre-diabetes and
Type 2 diabetes in later life. A recent meta-analysis has shown a
sevenfold lifetime increased risk of diabetes following GD. A robust
postpartum recall system is essential for early detection and
appropriate and timely intervention.
Aims: We sought to examine the rate of postpartum screening
amongst women with GD attending ATLANTIC DIP centres in the
west of Ireland and establish the prevalence of abnormal glucose
tolerance amongst women tested.
Methods: All women diagnosed with GD attending five antenatal
centres in the ATLANTIC DIP consortium during the years
2007 2012 were invited to have a postpartum glucose tolerance
test (OGTT) at 12 weeks postpartum. Postpartum glucose status is
defined by American Diabetes Association (ADA) criteria. The
service is coordinated by a central administrator.
Results: Of 1,520 women with GD, 1,149 (76%) attended for
postpartum OGTT over the 5 year period. Of those who attended,
6% had impaired fasting glucose, 9% impaired glucose tolerance
and 4% had Type 2 diabetes. Women who were older and/or
treated with insulin during pregnancy were more likely to attend
for postpartum screening.
Conclusion: GD is common with rising prevalence rates. The risk
of future pre-diabetes/diabetes post GD is high at 19%. A
coordinated recall programme such as occurs in ATLANTIC DIP
is effective in achieving an excellent uptake rate at 76% similar to
that seen in other national screening programmes.
2014 The Authors.

DIABETICMedicine
P406
Higher serum glucose levels during oral
glucose tolerance test (OGTT) allow early
identification of women who will
ultimately fail dietary intervention alone
and progress to pharmacological therapy
for gestational diabetes
CM McHugh1 and D ODonoghue2
1
Medicine, Sligo Regional Hospital, Sligo, Ireland, 2Medicine, National
University of Ireland Galway,Galway, Ireland
Aims and objectives: Individuals with gestational diabetes who

will ultimately require pharmacotherapy with insulin/metformin
compared with diet alone have higher glucose levels during OGTT
at diagnosis.
Methods: This was a retrospective observational study of all
women diagnosed with gestational diabetes from 2008 to 2012
screened using a 75g OGTT between 24 and 28 weeks gestation
based on risk factor identification.
Results: There were 287 pregnancies with gestational diabetes:
157 managed on diet alone, 130 needed insulin and/or metformin
in addition to diet alone. Three miscarriages (twodiet, one
metformin). Those who received metformin had a higher
booking body mass index than those on diet alone. Those
requiring pharmacotherapy had a higher fasting and serum
glucose and at 2h post glucose. There was no difference between
the groups at 1h post glucose load. Mean glucose concentrations
during the OGTT in the diet group were fasting 4.67 0.45, 1h
9.7 0.14, 2h 8.32 0.13mmol/l and in the pharmacotherapy
groupfasting 5.23 0.69, 1h 10.13 0.21, 2h 8.52 0.2mmol/
l. Those who subsequently failed metformin monotherapy and
required metformin and insulin had a higher fasting and 2h serum
glucose at diagnosis compared with those managed on metformin
monotherapy. There was no significant difference between
pharmacotherapy and diet groups in terms of fetal outcome.
Conclusions: Fasting and 2h serum glucose during screening
OGTT were higher in those individuals who progressed to insulin
and/or metformin. Early stratification of the results of the fasting
glucose level could allow identification of those most at risk.
P407
Study on the safety and efficacy of the use
of insulin glargine in gestational diabetes
CM McHugh1 and D ODonoghue2
1
Department of Medicine, Sligo Regional Hospital, Sligo, Ireland,

Department of Medicine, National University of Ireland Galway, Galway,
Ireland
2
Aims and objectives: Insulin glargine has comparable safety and

efficacy as licensed insulin or diet alone in treating gestational
diabetes.
Methods: This study is a retrospective observational study of all
completed pregnancies in patients with gestational diabetes, as
determined by oral glucose tolerance test, treated with glargine
either alone or in combination with other insulins at Sligo Regional
Hospital between January 2008 and July 2012.
Results: A total of 187 patients were diagnosed with gestational
diabetes with no fetal losses. There was no difference between the
treatment groups at baseline in terms of gestation at diagnosis,
body mass index, systolic blood pressure (BP) or diastolic BP.
Those receiving glargine and aspart had a significantly higher
fasting serum glucose (5.11 0.35mmol/l, p < 0.001), 1h
(10.93 0.62mmol/l,p < 0.001) and 2h (10.07 0.61mmol/l,
149
DIABETICMedicine
p < 0.001) during OGTT (p < 0.001) compared with those on diet
alone. There was no significant difference between the gestation at
delivery, birth weight, Apgar at 0 and 5min, mode of delivery, or in
systolic BP, diastolic BP after 20 weeks gestation. There was a
higher mean HbA1c in those receiving glargine and aspart
compared with diet alone (5.73 0.16% vs 5.58 0.2%)
(p = 0.002). This was not evident with any other group
comparison. There is a positive correlation between the birth
weight and mode of delivery (vaginal delivery or Caesarean
section) (p = 0.018)(r2 = 0.031), but no correlation between
treatment type and birth weight.
Conclusion: This study demonstrated the safety and efficacy of
glargine therapy during pregnancy compared with diet alone.
P408
Managing gestational diabetes: an
evaluation
EM Kewell, J Chaman, J Knott, S Matthias and T Richardson
Bournemouth Diabetes and Endocrine Centre, Royal Bournemouth Hospital,
Bournemouth, UK
Aims: This study evaluated the effectiveness of a dietitian-led

antenatal service within the Joint Obstetric Clinic (JOC) in
managing gestational diabetes (GD).
Methods: An oral glucose tolerance test (OGTT) was offered to
women with risk factors for GD at 24 28 weeks or 16 18 weeks
gestation if they had a previous history of GD. Mothers with GD
were managed through a parallel-run dietitian clinic run solely by a
dietitian. Help was available within the clinic environment from
the multidisciplinary team and patients could be switched between
the dietitian-led clinic and the JOC itself seamlessly. Treatment
included dietary and lifestyle intervention. Medication may have
been prescribed. Women presenting with GD between January
2009 and January 2010 were evaluated including demographics,
risk factors and postnatal outcomes.
Results: Ninety-two women, mean(SD), age 32years (5.3), body
mass index 29kg/m2(5.9). 54% had a family history of diabetes
and 14% had previous GD. 29% of individuals required the
addition of hypoglycaemic agents. 23% of deliveries were by
elective Caesarean section. 1% of neonates were macrosomic
(>4,500g). 92% had a postnatal OGTT, of which 8% showed
impaired glucose tolerance and 4% demonstrated Type 2 diabetes.
There was no correlation between birth weight and assisted
deliveries, or between patients booking body mass index/weight
and method of treatment.
Conclusions: The dietitian-led service is an effective intervention

in managing GD. The majority of individuals were well managed
with diet and lifestyle changes. At a time when GD numbers are
increasing dramatically, alternative methods of managing the
antenatal service are needed. A parallel-run dietitian-led service is
effective and safe and can help with the increasing burden of GD
within the JOC.
P409
Pre-conception care in diabetes and impact
on fetal macrosomia
S Krishnasamy1, R Indusekhar2, S Taylor1, L Varadhan1 and
AU Nayak1
1
Diabetes and Endocrine Department, University Hospital of North
Staffordshire, Stoke-on-Trent, UK, 2Obstetrics and Gynaecology Department,
University Hospital of North Staffordshire, Stoke-on-Trent, UK
Background and aims: Pre-conception care (PCC) in diabetes

improves adverse pregnancy outcomes, especially reducing rates of
neonatal malformations and premature delivery. However, its
impact on fetal macrosomia is not clear.
Methods: All women with pre-existing diabetes from the catchment
area, delivered in the year 2011 and 2012, were identified (n = 44).
Data on PCC, diabetes treatment during early pregnancy, HbA1c
(pre-conception and third trimester), neonatal birth weight, congenital
malformations and neonatal intensive care unit (NICU) admission
were obtained retrospectively from electronic patient records.
Results: Of the 44 women identified, 25 (57%) had Type 2
diabetes. Eight (18%) women had PCC (four each with Type 1 and
Type 2 diabetes). The HbA1c (mean SD) was lower in those
who had PCC (pre-conception HbA1c 62 17 vs 74 21mmol/
mol; third trimester HbA1c 50 14 vs 56 1421mmol/mol).
Proportions with congenital birth defects (0% vs 22%), fetal
macrosomia (birth weight 4.5kg) (0% vs 5.6%), NICU
admissions (0% vs 30%) were different, being higher in those
without PCC. Amongst those without PCC, proportion on statins,
ACEinhibitors, pioglitazone, GLP-1 analogues, DPP-4 inhibitors
were 2.8%, 5.6%, 2.8%, 5.6% and 2.8% respectively, whilst none
who had PCC was on any of these therapies.
Conclusions: Pre-conception care in women with diabetes can
impact on glycaemia and improve outcomes including a potential
reduction in fetal macrosomia. With the increase in prevalence of Type
2 diabetes in women of child bearing age, pre-conception care could
avoid potentially teratogenic pharmacotherapy in early pregnancy.
Clinical care and other categories posters:

psychological health
Aims: To investigate the emotional wellbeing and diabetes control

of adults with Type 1 diabetes following admission for diabetic
ketoacidosis (DKA).
Methods: Inventories measuring anxiety, depression (Hospital

Anxiety and Depression Scale), post traumatic stress disorder
(PTSD) symptoms (Impact of Events Scale
Revised) and
diabetes-related distress (Problem Areas in Diabetes, PAID) were
completed during admission and 3 months later. Additional
demographic and clinical data were also collated and analysed.
Results: Many of the 24 participants reported clinical levels of
anxiety (37.5%), depression (25%) and PTSD symptoms (37.5%)
during admission. Three months later, the prevalence of clinical
anxiety remained high (37.5%), although clinical levels of
depression (8.3%) and PTSD symptoms (8.3%) attenuated
markedly. There was no significant change in PAID scores across
150
2014 The Authors.

P410
Emotional sequelae during and following
hospital admission for diabetic ketoacidosis
in Type 1 diabetes: a prospective cohort
study
KY MacLennan, AJA Keen and AE Gold
Diabetes Centre, NHS Grampian, Aberdeen, UK
Clinical care and other categories posters: psychological health
time(p = 0.055). Pre-existing diabetes control was poorer in the

study sample than the overall local Type 1 clinical population
(96mmol/mol vs 74mmol/mol; p < 0.0001) and this did not
improve following discharge (p = 0.487).
Conclusions: Admission for DKA is associated with lasting
clinical anxiety for many with Type 1 diabetes. It is probable
that these levels of psychological distress during and after
admission go undetected and therefore are not used to inform
management and intervention approaches (at clinical levels these
will substantially impair cognitive processes such as memory,
concentration and decision making). Further investigation is
warranted to identify the longer-term course of emotional
distress following DKA to inform any associated care pathway
and raise awareness of psychological vulnerability of this subgroup
of people with Type 1 diabetes.
P411
Psychosocial impact of closed loop therapy
for adults with Type 1 diabetes: overnight
closed loop at home study
H Thabit2, R Hovorka2, E Walkinshaw3, A Lubina-Solomon3,
M Stadler4, T Tinati5, SR Heller3, SA Amiel4, ML Evans2 and
KD Barnard1
1
Southampton, Southampton, UK, 2Wellcome Trust, Medical Research Council
Institute of Metabolic Science, University of Cambridge, Cambridge, UK,
3
Diabetes Centre, Clinical Sciences Centre, Northern General Hospital,
Sheffield, UK, 4Diabetes Research, Weston Education Centre, Kings College
London, UK, 5Medical Research Council, University of Southampton, UK
Background: The burden of diabetes management presents

sustained challenges for people living with the condition.
Nocturnal hypoglycaemia is common and a serious complication
of insulin treatment in people with Type 1 diabetes representing a
critical problem in disease management. The closed loop (CL),
combining real-time continuous glucose monitoring with a
computer-based algorithm modulating insulin delivery, may
provide a solution and represent an additional realistic treatment
option.
Method: Participants in the Angela03 study (cross-over
comparison of overnight open vs CL) were interviewed within 2
weeks of completion of the CL arm to explore perceptions of the
impact of the 4 week home overnight CL use on their lived
experience. Content and thematic analyses were conducted.
Results: The participants were eight adults with Type 1 diabetes
aged >18 years. Key positive themes:better control (n = 7);
reassurance (n = 5); improved sleep leading to improved daily
functioning and diabetes control (n = 5); The best control I have
had for several years, Youve got more reassurance with the closed
loop, Anything suffers in comparison to the closed loop and I felt a
real sense of loss when I came off that.Key negative themes: alarms
beeping (n = 7); technical issues/calibration (n = 5); device/
equipment size (n = 4); The pump had quite a lot of pump
alarms, Making it smaller would be the biggest thing, One
inadvertent press of the power button shut the thing down.
Conclusion: Participation in the trial was reported as positive by
all. Increased reassurance, security and confidence were key
benefits. Technical issues and human factor aspects should be
addressed before people with diabetes are entirely comfortable
with the device.
2014 The Authors.

DIABETICMedicine
P412
Closed loop therapy overnight at home
study: psychosocial impact on adolescents
J Allen1, R Hovorka1, D Elleri1, H Thabit1, L Leelarathna1,
A Gulati1, M Nodale1, DB Dunger1, T Tinati2 and KD Barnard3
1
Wellcome Trust, Medical Research Council Institute of Metabolic Science,
University of Cambridge, Cambridge, UK, 2Medical Research Council
Lifecourse Centre, University of Southampton, Southampton, UK, 3Human
Development and Health, University of Southampton, Southampton, UK
Diabetes burden is relentless and can be particularly challenging

during adolescence. Nocturnal hypoglycaemia is common and a
serious complication of insulin treatment in people with Type 1
diabetes (Type 1 diabetes) represents a critical problem in disease
management. The development of the artificial pancreas, a system
combining
glucose
monitoring
with
computer-based
algorithm-dictated insulin delivery, may provide a solution and
represent an additional realistic treatment option for people with
Type 1 diabetes. Semi-structured, qualitative interviews were
conducted to explore participants perceptions of the impact of
the 3 week home overnight closed loop use on their lived
experience. Content and thematic analyses were conducted to
explore positive and negative experiences of the closed loop
technology. The participants were 15 adolescents (11 19
years). Key positive themes:improved sleep (n = 8); stable blood
glucose levels (n = 14); safety (n = 14); The perfect sugar control
overnight (015), I felt I could completely trust it at night and I felt
completely relaxed about it being there (004), At night I didnt
have to worry about being diabetic (006).Key negative
themes: calibration difficulties (n = 4); alarms beeping (n = 5);
discomfort/size of device (n = 8); The implants for the pump, they
were annoying the needle was sore (008), Just having this little
brick on my arm really Just the size of it (002). Perfect blood
glucose control and time off from diabetes were key benefits. All
participants reported a positive experience taking part in the trial.
Technological and practical difficulties should be addressed before
people with diabetes are entirely comfortable with the device.
P413
Depression, anxiety and self-care among
young adults with Type 2 diabetes: results
from the international Diabetes MILES
study
JL Browne1,2, G Nefs3, F Pouwer3 and J Speight1,2,4
1
Australian Centre for Behavioural Research in Diabetes, Diabetes Australia,
Melbourne, Australia, 2Centre for Mental Health and Wellbeing Research,
School of Psychology, Deakin University, Burwood, Australia, 3Center of
Research on Psychology in Somatic Diseases CoRPS, Tilburg University,
Tilburg, The Netherlands, 4AHP Research, Hornchurch, Essex, UK
Aims: Young adults with Type 2 diabetes have higher physical

morbidity and mortality than other diabetes subgroups, but
differences in psychosocial outcomes have not yet been
investigated. We compared depression and anxiety symptoms
and self-care behaviours of young adults with Type 2 diabetes with
two matched control groups.
Methods: Using cross-sectional data from the Australian and
Dutch Diabetes MILES national surveys, we matched 93 young
adults (aged 18 39) with Type 2 diabetes (cases), with (i) 93 older
adults (40 years) with Type 2 diabetes (Type 2 diabetes controls;
matched on country, gender, education, diabetes duration, insulin
use) and (ii) 93 young adults with Type 1 diabetes (Type 1 diabetes
controls; matched on country, gender, age, education). Groups
151
DIABETICMedicine
were compared on depression (PHQ-9) and anxiety (GAD scale)

symptoms, as well as frequency of selected self-care behaviours
(measured with a single item for each behaviour).
Results: Cases had higher depression scores than Type 2 diabetes
controls (p = 0.01) but statistically equivalent scores to Type 1
diabetes controls. Groups did not differ on anxiety. Cases were less
likely than control groups to take insulin as recommended (both
p < 0.01) but showed no differences in oral medication-taking
compared with Type 2 diabetes controls. Cases were less likely
than Type 2 diabetes controls to eat healthily (p = 0.03), and less
likely than Type 1 diabetes controls to be physically active
(p = 0.01). Effect sizes (Cohens d) ranged from 0.31 to 0.71.
Conclusions: Our results suggest that Type 2 diabetes is as
challenging as Type 1 diabetes for young adults and more so than
for older adults with Type 2 diabetes. Young adults with Type 2
diabetes may require more intensive psychological and self-care
support than their older counterparts.
P414
Does one size fit all? Towards an
understanding of the maintaining factors of
fear of hypoglycaemia
C Shaban, D Kerr and J Knott
Bournemouth Diabetes and Endocrine Centre, Royal Bournemouth Hospital,
Bournemouth, UK
Aims: The purpose of this study was to evaluate the relationship

between aspects of fear of hypoglycaemia (FoH), HbA1c and
diabetes distress in three groups of patients with Type 1 diabetes.
Method: Complete questionnaires to assess FoH (Hypoglycaemia
Fear Survey, HFS-II) and diabetes distress (Problem Areas in
Diabetes Scale, PAIDS) were obtained from 210 people (n = 96
male) with Type 1 diabetes attending either intensive insulin
training (BERTIE), initiation of insulin pump therapy (pump start
PS) or pump follow-up clinics (PC). The HFS-II measured three
aspects of FoH: worry (HFS-W), actions to maintain high blood
glucose (MHBG) and actions to avoid hypoglycaemia (AH).
Analysis of variance and Pearson correlations were used.
Results: Between group comparisons showed those established on
insulin pump therapy (PC) reported the lowest HbA1c and PAIDS
scores overall. PC reported lower HFS-W than the pump start (PS)
group [mean(SD) 14.56(12.75), 23.97(15.17), p < 0.001]. BERTIE
and PS were differentiated by the report of greater hypoglycaemia
avoidance by the PS group [9.00(6.97), 6.08(6.21), p < 0.02].
MHBG was related to poorer glycaemic control in both pump
groups (r = 0.28, p < 0.02; r = 0.27, p < 0.02). The BERTIE
group showed a positive association between HFS-W and HbA1c
(r = 0.25, p < 0.04).
Conclusions: The findings demonstrate that different aspects of
FoH are prevalent in the three diabetes treatment groups and are
differentially associated with glucose control and diabetes distress.
This contributes to an understanding of the coping and
maintaining factors of FoH and highlights the need for
interventions that place different emphasis on either behaviour or
thinking at different points in the care pathway.
152
P415
Beliefs about insulin in people with newly
diagnosed Type 2 diabetes: findings from
the South London Diabetes Cohort (SOUL-D)
study
M Boughdady1, K Winkley2, K Ismail2 and SA Amiel3
1
School of Medicine, Kings College London, London, UK, 2Department of
Psychological Medicine, Institute of Psychiatry, Kings College London,
London, UK, 3Diabetes and Nutritional Sciences Division, Kings College
London, London, UK
Aims: Psychological insulin resistance (PIR) is the unwillingness to

accept insulin therapy due to psychological barriers. The aim of
this cross-sectional study was to determine whether ethnicity is
associated with more PIR.
Methods: This was a cross-sectional study of a population-based
cohort of individuals with newlydiagnosed Type 2 diabetes
recruited through 6-monthly searches of 96 GP databases in
southeast London. Patients completed the Barriers to Insulin
Treatment (BIT) questionnaire to determine PIR, with scores > 5
associated with PIR. Data were also collected on sociodemographics, biomedical factors and depression status (Patient
Health Questionnaire-9).
Results: In all, 1,790 participants were recruited to the SOUL-D
cohort. Mean age was 56 years (SD 11.0), 55% were male, 49%
were of white ethnicity, 40% were black and 11% were of Asian/
other ethnicity. Participants of black, Asian/other ethnicity were
three times more likely to score in the top quartile range on the BIT
(mean score 5.58) than white participants (odds ratio 3.67, 95%
CI 2.86 4.70, p < 0.001). This relationship persisted when
adjusted for age, gender, HbA1c, duration of diabetes (months),
diabetes treatment (insulin, oral medication, diet only), and
depressive symptoms (odds ratio 3.34, 95% CI 2.45 4.57,
p < 0.001).
Conclusions: Preliminary results demonstrate that PIR is
associated with ethnicity in people with a recent diagnosis of
Type 2 diabetes. This suggests that clinicians need to recognise that
people of non-white ethnicity are likely to have more barriers to
commencing insulin therapy. This is of clinical importance as these
patients are more likely to have an earlier onset of diabetes.
P415A
Outcome of individual psychological
interventions following a cognitive
behavioural therapy and mindfulness
model applied to adults with Type 1 and
Type 2 diabetes
KS Woldt
Department of Diabetes, Crichton Hospital, Dumfries, UK
Aims: The main aims were to determine whether psychological

interventions following a cognitive behavioural therapy approach
were improving glycaemic control and psychological outcomes as
measured by the Hospital Anxiety and Depression Scale (HADS).
Method: All clients with Types 1 and 2 diabetes who received
psychological treatment in a year locally regarding their mood
issues were entered in the study. Pre-and post-treatment results
were compared using repeated t-tests and at a 6 month
follow-up. The study compared HbA1c levels (in mmol/mol) and
HADS data. A random selection of clients on the SCI-DC data set
who had a heightened HADS score (of 9 or above) were entered in
the treatment as usual group (TAU), in addition to having
2014 The Authors.

heightened HbA1c levels of above 65mmol/mol (at baseline) and

not having psychology input.
Results: A total of 65 patients entered the study. The HbA1c
levels in the treatment group significantly reduced from baseline
91.96 (p = 0.000) to 80.23 after treatment (p = 0.000) and again
to 75.8mmol/mol (p = 0.002) at 6 months follow-up. The initial
mean reduction was larger in the treatment group with
11.73mmol/mol in comparison to only 2.03mmol/mol in the
TAU group. In the treatment group the HADS depression scores
dropped from pre-treatment 10.94 to post-treatment 3.28
(p = 0.000). Also their mean anxiety scores reduced significantly
(13.53 pre-treatment) to 4.88 (p = 0.000).
Conclusions: This study has shown that psychological input can
reduce both HbA1c levels over treatment time and 6 month
follow-up and lead to larger reductions of anxiety and depression
levels than in a TAU group.
P417
Psychosocial impact on parents of
adolescents with Type 1 diabetes: overnight
closed loop at home study
D Alleri1, R Hovorka1, J Allen1, H Thabit1, L Leelarathna1,
A Gulati1, M Nodale1, DB Dunger1, T Tinati3 and KD Barnard2
1
University of Cambridge, Cambridge, UK, 2Human Development and Health,
University of Southampton, Southampton, UK, 3Medical Research Council,
University of Southampton, Southampton, UK
Parents of a child with Type 1 diabetes have increased rates of

stress and depression. Fear of nocturnal hypoglycaemia is common
causing heightened anxiety and leading to maladaptive coping
strategies. The development of the artificial pancreas, a system
combining
glucose
monitoring
with
computer-based
algorithm-dictated insulin delivery, may represent an additional
realistic treatment option for people with Type 1
diabetes. Semi-structured, qualitative interviews were conducted
to explore perceptions of the impact of the 3 week home overnight
closed loop use. Content and thematic analyses were conducted
exploring positive/negative experiences of closed loop
technology. The participants were 13 parents of participating
adolescents. Key positive themes: felt safe/reduced anxiety
(n = 13); stable blood glucose levels (n = 8); improved sleep
(n = 6); Knowing that theyre going to be OK in the night
thats a safety thing (004), Not having the worry not having to
drag yourself out of bed testing at 2 or 3 oclock in the morning
It was reliable (002), Improved HbA1c was probably best (001).
Key negative themes: calibration difficulties (n = 5); size of device
(n = 4); alarms beeping (n = 2); It sometimes wouldnt connect
which took quite a long time to sort out(013), Probably the alarm
because they were just really loud (010), It was quite big and Id
quite like to have a slightly littler one (012). All participants
reported a positive experience taking part in the trial. Reduced
anxiety and feeling their child was safe were key benefits. Alarms
and calibration difficulties were noted as barriers to the
technology.
2014 The Authors.

DIABETICMedicine
P418
Implementing a physical activity
consultation service for adults with
diabetes: promising results from a 12 month
evaluation
M McCallum1, L Matthews2, A Kirk2, N Mutrie3 and AE Gold1
1
Diabetes Centre, NHS Grampian, Aberdeen, UK, 2Physical Activity for Health
Research Group, School of Psychological Sciences and Health, University of
Strathclyde, Glasgow, UK, 3Institute for Sport, Physical Education and Health
Sciences, University of Edinburgh, Edinburgh, UK
Objective: To provide follow-up data on the implementation of

an evidence-based physical activity consultation service for adults
with diabetes (published Diabetes UK 2013).
Methods: A 12month physical activity intervention is being
offered to adults with Type 1 or Type 2 diabetes within NHS
Grampian. Promotion of the service is provided within primary
and secondary care. Patients self-refer. Patients receive an initial
30 45min face-to-face consultation, monthly follow-up
consultations for 6 months (face-to-face, email or telephone)and
further follow-up consultations at 6 and 12 months. Consultations
are guided by evidence-based behaviour change strategies.
Results: During the initial 18 months, 47 patients enrolled in the
service (60.9 10.2 years; body mass index 33.1 6.9; 51.2%
female; 83.7% with Type 2 diabetes). A significant increase was
observed from baseline to 6months in the number of participants
achieving the physical activity recommendations (IPAQ) (21.3%
vs 66.7%; p = 0.002). This was maintained from baseline at
12 months (21.3% vs 54.5%;p= 0.012). A significant increase was
observed in positive affect (PANAS) from baseline to 12 months
(mean difference 6.5, SD 6.1; 95% CI 1.4, 11.6; p = 0.019).
Positive but non-significant changes were also observed in
perceived levels of anxiety, depression and negative affect.
Process evaluation demonstrated high adoption by health
professionals. Reasons given for adoption included protocol
integrated with current diabetes care, minimal time requirements
from health professionals, and delivery by an experienced
exercise/health psychologist.
Conclusions: Physical activity consultation should be considered
as an effective method of supporting people with diabetes and
comorbidities to increase physical activity, promote weight loss
and improve psychological wellbeing.
P419
Enhancement of structured diabetes care
for adults with diabetes and a severe
mental health condition
C Simpson
Community Diabetes Team, Bradford Teaching Hospitals NHS Foundation
Trust, Bradford, UK
Aim: To improve care and engagement with patients with both a

severe mental health condition and diabetes resulting in improved
glycaemic control and reduction in short- and long-term
complications.
Methods: Increase the amount of input from a diabetes specialist
nurse (DSN) and a diabetes specialist dietitian in the local mental
health hospital and more joint working in the community
psychiatric services.This time will be spent implementing the
following: establish multidisciplinary team clinics at the mental
health hospital for inpatients; ongoing inpatient on ward reviews
by a DSN; crisis management on discharge; review patients in the
community where they choose to be seen; on ward education
153
DIABETICMedicine
sessions and off ward link nurse education sessions; new blood
glucose monitors and care plans on the wards.
Results: Thirteen patients took part in the study. Baseline average
HbA1c 76mmol/mol, average HbA1c after 3 months 62mmol/mol
and at 6 months 54.7mmol/mol. At baseline the percentage of
patients having all annual diabetes checks was 48%; at 6 months
this increased to 83.3%. Staff who engaged with the education
reported to be more confident at managing patients with diabetes
on the wards.
Conclusion: A more systematic and flexible approach to looking
after patients with diabetes and mental health problems creates a
reduction in HbA1c and improved patient satisfaction.
P420
Levels of anxiety and depression in
secondary care diabetes outpatient clinics:
the need for service change
SG Wijetilleka, S Pais, W Liberty and TM Galliford
Department of Diabetes and Endocrinology, West Hertfordshire Hospitals
NHS Trust, Watford, UK
Aims: Rates of depression in patients with diabetes are twice that

of the general population which has detrimental effects on
glycaemic control and complications. Provision of psychological
support therapies is patchy despite proven benefits in terms of
improved quality of life and recommendations in the National
Service Framework.
Methods: We surveyed unselected consecutive patients attending
for review in a specialist diabetes clinic within a district general
hospital over 7 weeks. We obtained demographics and patient
opinion about psychological support. Responders completed a
modified established questionnaire assessing levels of anxiety and
depression (Hospital Anxiety and Depression Score).
Results: Analysis of the 67 responders was as follows: mean age
48 years, 84% of patients had a diabetes duration of 6 years plus,
60% of patients had Type 1 diabetes, 46% had complications.76%
of responders reported to feeling happy and relaxed regarding their
condition; however, formally assessed levels of anxiety were
38%. Levels of depression were low although 39% had lost
interest in their appearance compared with pre-diagnosis. 46% of
respondents wanted to see a psychologist.
Conclusions: This study identifies a need and desire for
psychological support in patients with complex and
long-standing diabetes in West Hertfordshire. Best practice tariff
for children with diabetes mandates access to psychological
services; this is essential for adults and needs to be recognised by
commissioners.
P421
Diabetes care in the acute mental health
setting
AF Conlon, H Atkins and KS Higgins
Diabetes and Endocrinology, University Hospitals of Leicester Acute Trust,
Leicester, UK
Introduction: Schizophrenia is an independent risk factor for

diabetes. Atypical antipsychotics have fewer extrapyramidal side
effects but have a negative impact on glycaemic control. Any
admission for this group of patients should be seen as an
opportunity to screen for or monitor pre-existing diabetes.
Aim: To assess if people with diabetes admitted to a mental health
unit are receiving standards of care recommended by the National
Institute of Clinical Excellence (NICE) and DiabetesUK.
154
Method: A snapshot audit reviewed the clinical notes for each

patient over 1 month. Areas audited included blood glucose
monitoring, screening blood tests including cholesterol and
HbA1c, foot assessment and documentation of hypoglycaemic/
hyperglycaemic episodes.
Results: Thirty-six of 245 patients were recorded as having
diabetes, 21 male, 15 female, median age 62years. 100% (36)
had a blood glucose monitoring chart. 77.8% (28) had blood tests;
however, only 38.9% (14) had their HbA1c checked. 47.2% (17)
had their feet checked during admission.Compliance with
medication and glucose monitoring was as an issue for 50% (18)
of patients. Seventeen hypoglycaemic episodes were recorded but
there was little or no documentation of treatment that followed.
Conclusion: Further questioning revealed many care staff and
psychiatric nurses were not confident in treating hypoglycaemic
episodes. Their knowledge of insulin was poor. This prompted the
formulation of a staff knowledge questionnaire with the intention
of subsequent education sessions. We also need to educate
regarding foot checks to reduce the number of limb amputations
for this group of patients.
P422
Impact of diabetes on the quality of life of
family members: UK findings from the
second Diabetes Attitudes, Wishes and
Needs (DAWN2TM) study
KD Barnard1, MH Cummings2, MJ Davies3, N Kanumilli4 and
N Munro5
1

Diabetes and Endocrine Centre, Portsmouth NHS Trust, Portsmouth, UK,
3
Leicester Diabetes Centre, University of Leicester, Leicester, UK, 4South
Manchester Clinical Commissioning Group, Northenden Group Practice,
Northenden, UK, 5Postgraduate Medical School, University of Surrey,
Guildford, UK
2
Aims: To explore the impact on family members (FMs) of living

with a person with diabetes using a descriptive sub-analysis of the
UK data from the DAWN2TM study.
Methods: Between 6 March 2012 and 22 May 2012, FMs
completed the DAWN2 survey covering quality of life (QoL),
wellbeing and problem areas (WHO-QOL-BREF; WHO-5;
PAID-5). FMs were living in the same household as people with
diabetes who received insulin (n = 46) or were not on insulin
medication (n = 76).
Results: Overall, FMs of those not on insulin medication reported
slightly higher QoL (68% WHO-QOL-BREF good/very good
scores) than FMs of people taking insulin (54%). Wellbeing
(WHO-5) was the same for both groups (67% good wellbeing),
but 9% of FMs across both groups scored as potentially depressed.
For PAID-5, 24%26% of all FMs reported worrying about the
future and possible complications as a serious or somewhat
serious problem; 4% in both groups felt depressed when thinking
about the fact that the person they live with has diabetes. Diabetes
had a positive impact on 24%28% of FMs relationships with the
person they live with; however, managing the diabetes of that
person was much less of a burden for FMs of people not taking
insulin than for FMs of people insulin treated (61% vs 35% no
burden).
Conclusions: The burden of diabetes extended beyond the person
who has the condition. Whilst many family members reported
good QoL, diabetes was associated with concerns for the future
and risk of long-term diabetic complications for their loved ones.
2014 The Authors.

P423
What are the factors that influence
glycaemic control in the emerging adult
group of people living with Type 1 diabetes?
CM Reidy1, D Wellsted1, TM Galliford2 and S George3
1
Centre for Lifespan and Chronic Illness Research, University of Hertfordshire,
Hatfield, UK, 2Department of Diabetes and Endocrinology, West
Hertfordshire Hospitals NHS Trust, Watford, UK, 3Department of Diabetes
and Endocrinology, East and North Hertfordshire NHS Trust, Hertford, UK
Background: Extensive research has been conducted focusing on

children or adults living with Type 1 diabetes, but those between
16 and 25 are under-researched. However, there is evidence that
patients cognitions (thoughts about illness) and emotional
reactions combine to form what can be described as illness
perceptions that drive behaviour.
Aims: To explore how illness perceptions, clinical and significant
life events influence glycaemic control in the emerging adults living
Methods: This was an observational, cross-sectional approach
(N = 70) using a self-report questionnaire pack: Brief Illness
Perceptions Questionnaire, Diabetes Psychological Adjustment
Scale, Body Shape Questionnaire 8B, Patient Health
Questionnaire 8, Generalised Anxiety Disorder Assessment 7,
and an additional questionnaire looking at clinical and significant
life events.
Results: A hierarchical logistic regression model found that illness
perceptions of personal control and understanding were the best fit
for predicting glycaemic control (multiple R = 75.3) and account
for 49.4% of the difference in the groups (p = 0.002). BSQ-8B
scores demonstrated that a more negative perception of ones own
body shape is related to higher HbA1c levels in women. The
PHQ-8 also demonstrated that higher levels of depression are
related to higher HbA1c levels.
Discussion: An interplayof understanding and personal control
could have a significant effect on glycaemic control along with
negative body image (in women) and higher levels of depression.
Where conventional methods of glycaemic control are less effective
in this group, the findings suggest that targeted and specific
psychological support could improve glycaemic control.
P424
Depression and glycaemic control in Emirati
patients with diabetes: how strong is the
association?
A Khalili, N Lessan, Z Hannoun and MT Barakat
Research, Imperial College London Diabetes Centre, Abu Dhabi, United Arab
Emirates
Aims: An association between depression and diabetes has been

described in several studies. The presence and the strength of this
association in the Emirati population is unknown. We have
therefore investigated the value of HbA1c as a surrogate marker
of depression. We have also aimed to identify patients at higher
risk in the population examined.
Methods: In a prospective cohort study, adult Emirati patients
with pre-diabetes and diabetes (n = 609) and normal control
subjects (n = 85) attending the Imperial College London Diabetes
Centre between August and November 2012 had depression
screening using an Arabic translation of the PHQ9 depression
screening questionnaire. Demographic and metabolic parameters
of the same visit were retrieved from patients medical records.
Basic and comparative statistics were performed (SPSS20).
2014 The Authors.

DIABETICMedicine
Results: A weakbut highly significant positive correlation was

seen between PHQ9 and HbA1c (p = 0.008, r = 0.107). Women
were found to have a significantly higher PHQ9 compared with
men (mean SE: 7.26 0.25 vs 4.5 0.34, p < 0.0001). Also,
patients with age 40 years, body mass index 30kg/m2, HbA1c
10% and those with Type 1 diabetes had a significantly higher
depression score. There were no significant correlations between
PHQ9 scores and vitamin D level, smoking habit, LDL cholesterol
or blood pressure.
Conclusion: In the Emirati patients, younger age, obesity, female
gender, Type 1 diabetes and poor glycaemic control were found to
be independent risk factors for depression. HbA1c used alone was
not a good surrogate indicator of depression in the population
studied.
P425
The influence of psychological wellbeing on
glycaemic control in adults treated with
continuous subcutaneous insulin infusion
(CSII)
MA Jones1, A Al-Janabi1, S Ng1, K Davenport1, C Ward1 and
ML Evans1,2
1
Wolfson Diabetes and Endocrinology Clinic, Cambridge, UK, 2Institute of
Metabolic Science, University of Cambridge, Cambridge, UK
Aims: We tested the hypothesis that suboptimal glycaemic control

in patients with Type 1 diabetes starting on CSII is related to a fear
of hypoglycaemia.
Methods: As part of an ongoing clinical audit, non-pregnant
adult patients completed a hypoglycaemic fear survey (HFS) on
commencing pump therapy and annually. Of the 85 patients
recorded on CSII 39 were eligible with completed paired data.
Patients were divided into two groups based on their annual
HbA1c (8.5% and 8.5%) comparing raw HFS scores at 1 year
and change in HFS score over a year on CSII.
Results: As expected there was a significant improvement in fear
of hypoglycaemia (HFS 68.9 2.3 to 58.8 2.1, p < 0.025,
n = 39) within a year of commencing pump therapy. Fear of
hypoglycaemia (score > 50%) was found in 28% of patients;
however, it was not significantly correlated with glycaemic control.
Conclusion: Fear of hypoglycaemia is not correlated with HbA1c
in patients using CSII so it is not a major determinant of glycaemic
control at a clinic level. However, fear of hypoglycaemia is present
in many CSII patients and is still important in determining
psychological wellbeing.
P426
Cross-sectional survey of depression and
anxiety in a diabetes clinic population:
levels and role in blood sugar control
HA Lakhani1, I Cane1, M Taylor2, P Chadwick2 and
M Rosenthal3
1
Medical Sciences, University College London, London, UK, 2Psychology,

Royal Free Hospital, London, UK, 3Endocrinology, Royal Free Hospital,
London, UK
Aims: To evaluate the prevalence of depression and anxiety in

adult outpatients with diabetes and identify the correlation with
diabetes management behaviours and clinical outcomes.
Methods: Cross-sectional analyses were performed on 142
diabetic patients (Type 1 diabetes n = 67, Type 2 diabetes
n = 60) attending outpatient clinic. Depression and anxiety were
155
DIABETICMedicine
evaluated using the Patient Health Questionnaire (PHQ-2) and the

General Anxiety Disorder Questionnaire (GAD-2). Validated
survey items were used to measure patient characteristics and
clinical outcomes.
Results: Prevalence of psychological symptoms was high, with 39
patients (27.3%) having a score indicative of depression compared
with <10% in the general population, and 29 (20.3%) with
general anxiety disorder. Regression analysis showed that anxiety
and quality of life correlated strongly with depression (R = 0.59
and 0.43 respectively). At multivariate analysis, depression and
anxiety were significantly associated with higher self-reported
blood glucose levels, greater number of diabetes related
complications (p < 0.01) and less physical activity (p < 0.05).

Poorer satisfaction with treatment and lacking self-monitoring of
glucose also correlated with depression (p < 0.05). More than half
of the patient population felt that it was very important for
psychological support to be diabetes specific.
Conclusion: In our sample, patients with depression and
anxiety had poorer clinical outcomes with depressed patients
also exhibiting worse diabetes management behaviours.
This highlights the importance of screening for depression and
anxiety in outpatient visits, which would enableidentification
and management of psychological symptoms, thus improving
glycaemic control and overall wellbeing.
Clinical care and other categories posters: retinopathy
Refer to Oral A63
under eye hospital care and three patients were kept in digital
surveillance at 6 monthly intervals.
Conclusion: The
diabetic
prisoner
population
in our
county showed a higher level of any diabetic retinopathy (40%)
than the normal diabetic population (32%). In our county during
the year 2011 2012 there were 27,219 digital screenings: 1,488
(5.4%) had sight-threatening retinopathy and 8,654 (32%) had
any retinopathy. However, 10% of diabetic prisoners had
sight-threatening diabetic retinopathy which is double the normal
diabetic population level. The results emphasise the need to screen
this group of the population although it requires extra time,
expense and effort on the part of the screening programme.
P428
P430
Retinopathy post-pancreas transplant: are

we blind to the risk?
Do we need to consider ethnicity when

estimating risk of progression to
sight-threatening retinopathy? A
retrospective cohort analysis of patients
from an ethnically diverse diabetic eye
screening programme with at least three
screening episodes
P427
Personalised risk estimation for progression
to sight-threatening diabetic retinopathy:
how much does clinical information add to
screening data?
IM Stratton1, SJ Aldington1, AJ Farmer2 and PH Scanlon1
1
Gloucester Diabetic Retinopathy Research Group, Gloucester Hospitals NHS
Foundation Trust, Cheltenham, UK, 2Primary Health Care Sciences, University
of Oxford, Oxford, UK
A Nicholls, F Akiboye, D Cavan and D Kerr

Bournemouth Diabetes and Endocrine Centre, Royal Bournemouth and
Christchurch Hospitals NHS Foundation Trust, Bournemouth, UK
Refer to Oral A9
P429
Screening imperative: 10% of prisoners
with diabetes have sight-threatening
retinopathy
C Wallis, B Jennings and W Gatling
Dorset Diabetic Eye Screening Programme, Dorset Healthcare University NHS
Foundation Trust, Ferndown, UK
IM Stratton1, S Sivaprasad2, C Connor3, SJ Aldington1 and

PH Scanlon1
1
Gloucestershire Diabetic Retinopathy Research Group, Cheltenham, UK,

Ophthalmology, Kings College Hospital NHS Foundation Trust, London, UK,
Diabetic Eye Complications Screening Service, Guys and St Thomas NHS
Foundation Trust, London, UK
2
3
Aim: To assess levels of diabetic retinopathy within the prison

population of our county in 2012.
Method: All prisoners with diabetes were identified, prior to the
screening visit, by the medical staff at the prisons and details were
given to the screening programme. This has to happen close to the
visit date because the prison population is constantly changing and
prisoners are being moved. The camera and other equipment
needed were taken into the prison medical units. Patients were
screened and graded according to local Diabetic Eye Screening
Programme protocols.
Result: A total of 59 prisoners were screened: 35 had no diabetic
retinopathy (R0 M0); 19 had background retinopathy (R1);
threehad preproliferative retinopathy (R2); twohad proliferative
retinopathy (R3). Six patients had diabetic maculopathy (M1).
Three were referred to hospital eye unit care, one was already
Aims: We have developed a model stratifying risk of progression

to sight-threatening diabetic retinopathy (STDR) in those reported
with no STDR at either of two previous consecutive screening
episodes. We examine the effect of ethnicity on progression rates
by ethnic group after fitting the current risk model.
Methods: Outcomes from digital retinal screening and ethnic
group were extracted from the screening programme database.
Proportional hazards models were used to examine the effect of
ethnicity in ethnic groups where at least 100 progressed to STDR.
Results: Data were available for 18,504 people with no STDR at
two consecutive screening episodes, age 64 (15) years (mean,SD),
duration of follow-up 31 (13) months. Of these 9,478 (51.2%)
were White British Caucasians, 1,728 (9.3%) Other White, 3,993
(21.6%) Black Caribbean and 3,305 (17.9%) Black African.
Overall, 1,410 (7.6%) progressed to STDR. After fitting a model
including the presence of background retinopathy in neither, one
or both eyes at two consecutive screening episodes, ethnicity was
156
2014 The Authors.

Clinical care and other categories posters: retinopathy
highly significant: Black African hazard ratio 1.55 (95%CI

1.33 1.80), Black Caribbean 1.58 (1.38 1.81), Other White
1.24 (1.03 1.48). There were fewer than 100 STDR events for
mixed race groups but for White and Black African and White and
Black Caribbean mixed ethnic groups hazard ratios were 2.1
(1.35 3.35) and 1.5 (1.08 2.11) respectively.
Conclusions: Those with diabetes from these ethnic groups are at
higher risk of progression to STDR than White British Caucasians.
Ethnicity should be recorded by screening programmes and
included when estimating risk.
P431
Diabetic retinopathy in the Kingdom of
Swaziland: is there a need for screening?
HA Burn1 and J Pons2
1
2
Norwich Medical School, University of East Anglia, Norwich, UK,

Ophthalmologist, Good Shepherd Hospital, Siteki, Swaziland
Introduction: The prevalence of Type 2 diabetes is rising in

southern Africa and becoming an increasingly important
contributor to low vision. In 2012 3.8% of the population of
Swaziland had Type 2 diabetes, and this is set to increase
dramatically in the next decade. While the prevalence is
increasing, diabetes services in Swaziland remain extremely
limited leading to high complication rates.
Methods: This study is a retrospective survey of the medical
records of patients presenting to two eye clinics in Swaziland in the
18 month period from January 2012 to June 2013. Patients who
were placed in the diagnostic category of retinopathy were
selected. Of these 136 patients, 82 had diabetic retinopathy and
their age, sex, visual acuity, random blood sugar, diagnosis,
management and comorbidities were recorded.
Results: The four main findings are as follows. 38% of patients
presenting with diabetic retinopathy had a corrected visual acuity
of <3/60 at the time of presentation. 76% of patients had
proliferative diabetic retinopathy at the time of presentation.
84% had a random blood sugar of >8.0mmol/l at presentation to
the eye clinic. 68% had a coexisting hypertension.
Discussion: The high prevalence of advanced proliferative
disease, low visual acuity and poor diabetic control of patients
presenting to the eye clinic provides strong justifications for the
implementation of a diabetic retinopathy screening programme in
Swaziland. We propose a pilot screening programme involving one
retinal camera in Swaziland, an ophthalmologist in Canada to
interpret the results, and one ophthalmologist in Swaziland to
carry out laser treatment.
P432
Uptake of annual retinal screening in
adolescents with diabetes in Grampian
DA Cole1, S Philip2, JD Olson3 and A Mayo4
1
Foundation Programme, NHS Grampian, Aberdeen, UK, 2Diabetes and
Endocrinology, NHS Grampian, Aberdeen, UK, 3Ophthalmology, NHS
Grampian, Aberdeen, UK, 4Adolescent Diabetes, NHS Grampian, Aberdeen,
UK
Introduction: Current guidelines state that retinal screening

should be offered yearly to people with diabetes over the age of
12. We performed an audit of uptake of retinal screening,
prevalence of retinopathy and risk factors such as poor glycaemic
control among the adolescent population in Grampian with
diabetes.
2014 The Authors.

DIABETICMedicine
Method: Data regarding retinal screening attendance, age at

diabetes diagnosis, weight, blood pressure, HbA1c, and grade of
retinopathy were collected from the regional diabetes register. 672
patients between the ages of 12 and 20 attended the Grampian
screening programme during the calendar years 2006 2010.
Results: The uptake in screening increased from 62% in 2007 to
71% in 2010. The mean age at time of screening in 2010 was 18
years(SD 2.5) and duration of diabetes 7 years (4.6years). In 2010,
19% of the patients had some retinopathy. Patients with
retinopathy had diabetes for longer (11 years vs 8years,p
< 0.001). Fourteen people were diagnosed with referable eye
disease (maculopathy 13, severe background retinopathy 1) over
the whole study period. Compared withthe rest of the cohort
screened they had a higher mean HbA1c (95.6mmol/l vs
82.4mmol/l; p < 0.001) and higher mean arterial pressure
(91mmHg vs 86mmHg;p < 0.05).
Conclusion: There was increase in the uptake in retinal screening
from 2007 to 2010; however, there remains scope for
improvement. The prevalence of any retinopathy and that of
referable retinopathy was low.
P433
Retinopathy in cystic fibrosis related
diabetes: more common than we think
M Choudhury1, SF Wong1, J Bell2, L Evans2, L Speight2,
J Duckers2 and RI Ketchell2
1
Institute of Molecular and Experimental Medicine, Cardiff University School
of Medicine, Cardiff, UK, 2All Wales Adult Cystic Fibrosis Centre, University
Hospital Llandough, Penarth, UK
Background: Cystic fibrosis related diabetes (CFRD) occurs in

people with cystic fibrosis (CF) where viscous secretions due to
defective chloride channels damage the pancreas. CFRD is
becoming increasingly prevalent in patients with CF, along with
microvascular complications including diabetic retinopathy. We
examined the prevalence of diabetic retinopathy (DR) in adult
CFRD patients attending the All Wales Adult CF Centre.
Methods: Retinal screening data were analysed in patients with
CFRD between the years 2010 and 2012. Baseline clinical
characteristics and mean HbA1c were examined.
Results: Sixty-four patients had CFRD in 2012 (age range 18 49
years) of whom 54 (84%) were referred for retinal screening
between 2010 and 2012. Forty-three (80%) patients had attended
their most recent retinal screening appointment. Nineteen patients
(44%) had evidence of DR based on their most recent retinal
screen (mean HbA1c 69mmol/mol; CI 61 75mmol/mol). Of this
group, four had CFRD for a duration of 5 years including two
patients who had CFRD for 2 years. Of the 24 patients with no
DR, the mean HbA1c was 54.1mmol/mol (CI 50.8 59.6mmol/
mol) with nine patients having a duration of CFRD 5years. The
difference in the mean HbA1c between the two groups was
significant (p = 0.015).
Conclusions: Our study shows that DR is more common than
anticipated in CFRD and evident in patients with a duration 2
years which may highlight the fact that these patients may have
had dysglycaemia not revealed by the annual standard OGTT used
to diagnose CFRD.
157
DIABETICMedicine
P434
Prevalence of diabetic retinopathy at first
screen in children with diabetes diagnosed
before age 12 years in Scotland
HC Looker1 and DT Cromie2
1
Diabetes Epidemiology Unit, University of Dundee, Dundee, UK, 2Public
Health, NHS Lanarkshire, Glasgow, UK
Aims: UK screening programmes for diabetic retinopathy offer

screening from the age of 12 years but there is little published on
the prevalence of retinopathy at first screening. The aim was to
measure the prevalence of sight-threatening diabetic retinopathy at
first screening for children diagnosed with diabetes prior to age 12
years.
Methods: We used national retinopathy screening data for those
diagnosed under the age of 12 years and before May 2008 and
screened for the first time at 1213 years of age in the period 2006
2011 to calculate prevalence and risks associated with diabetes

duration.
Results: In all, 1,010 children (50.1% male) were diagnosed
before 12 years and screened for diabetic retinopathy for the first
time aged <14 years during the study period. Median diabetes
duration at screening was 5.3 years (interquartile range 3.18.0)
with 100 (9.9%) having a duration 10 years. There were no cases
of pre-proliferative or proliferative retinopathy at screening. 911
(90.2%) had no retinopathy and 9.8% [standard error (SE) 0.9]
had some retinopathy. Retinopathy prevalence was associated with
diabetes duration: 7.5% (SE 1.2) of children screened at <5 years
duration had retinopathy compared with 10.9% (SE 1.5) and
16.0% (SE 3.7) at 59 years and 10 years duration respectively.
The odds ratio for retinopathy adjusted for age and sex was 1.10
per year of diabetes duration (95% confidence interval 1.031.18).
Conclusions: No children had sight-threatening retinopathy at
their first screen which supports the current lower age limit for
retinopathy screening.
Clinical care and other categories posters: screening and

pre-diabetes
P435
Staff attendance for cardiovascular and
Type 2 diabetes risk screening at the
workplace: a qualitative study of drivers
and barriers
EM Di Battista1,2, RM Bracken3, JW Stephens3, K Morgan4,
M Thomas5, SP Williams6, M Williams4, S Rice4, SD Mellalieu2
^ r Group7
and Prosiect Sir Ga
1
Nutrition and Dietetics, Hywel Dda Health Board, Llanelli, UK, 2Applied
Sports Technology Exercise and Medicine A-STEM Research Centre, Swansea
University, Swansea, UK, 3Diabetes Research Group, College of Medicine,
Swansea University, Swansea, UK, 4Diabetes Centre, Hywel Dda Health
Board, Llanelli, UK, 6Tata Steel Packaging and Recycling, Llanelli, UK
Aim: Recruiting individuals for cardiovascular disease and Type 2

diabetes risk assessment (RA) is fraught with challenges. This study
aims to provide a deeper understanding of the drivers and barriers
of RA recruitment in the workplace,specifically by investigating
staff views and perceptions of a multi-agency project delivered via
occupational health departments in Carmarthenshire, Wales, UK.
Methods: Semi-structured face-to-face interviews were conducted
at the workplace with steel industry staff, nursing staff and other
hospital staff (n = 40). Purposive sampling ensured maximum
demographic variation within the study and non-responders
(n = 13) were sought to investigate access barriers. All interviews
were audio recorded, transcribed verbatim and analysed using
thematic content analysis via the constant comparison method.
Results: A total of nine themes were identified which generally
had an inverse relationship between drivers and barriers of
attendance: convenience of appointment booking, flexibility
of working demands, line management support, family history of
heart disease or diabetes, perceptions of personal health, history of
health promoting behaviours, perceptions of primary care services,
perceptions of occupational health department, access to
promotional material. Interestingly, despite other barriers, all
non-responders said they would attend if a line manager sensitively
encouraged them in a private one-to-one conversation.
Conclusions: Little qualitative literature exists regarding the
complexities of targeting staff for RA at workplaces in the UK.
158
This novel qualitative study reveals workplace dynamics that

facilitate RA attendance at occupational health. To maximise
recruitment, future workplace interventions should empower line
managers to fittingly approach staff.
P435A
Trends and outcomes from the use of the
glycated haemoglobin test to screen for
Type 2 diabetes in a diverse West Midlands
population: opportunities for diabetes
prevention
P Singh1, A Hayling2 and PH Davies1
1
Diabetes Centre, Sandwell Hospital, Sandwell and West Birmingham
Hospitals NHS Trust, West Bromwich, UK, 2Department of Pathology,
Sandwell and West Birmingham Hospitals NHS Trust, Birmingham, UK
Background: Whilst there is no gold standard to diagnose Type

2 diabetes, adoption of glycated haemoglobin (HbA1c) by WHO,
as an acceptable screening test, offers certain advantages. Since
2012 our electronic ordering system (Order-coms) has
distinguished between HbA1c requests for surveillance of
diabetes and those used for screening.
Aims: To examine trends in HbA1c requests for screening vs
overall use; trends in HbA1c requests by practice size/type;
outcomes of HbA1c used for screening.
Methods: Laboratory telepath data July 2012 to June 2013 were
analysed in Microsoft Excel.
Results: Over this period 35% of HbA1c requests were for
screening. This increased from 32% to 40% over 12 months
(45% to 51% after excluding test requests not in Order-coms).
Overall, HbA1c requests went up by 35% from ~8,000 to ~11,000
per month. Order-coms use increased with 84% of general practices
using this system at the beginning and 94% at the end of this period.
Practice characteristics predicted use of electronic ordering, with less
use from single handed practices and those requesting fewer HbA1c
tests. Using an HbA1c cutoff of 48mmol/l, 5,784 cases (15%) of
diabetes were diagnosed; 23,046 (58%) had HbA1c < 42mmol/l
and 10,965 (27%) had results of 42 47mmol/l.
2014 The Authors.

Clinical care and other categories posters: screening and pre-diabetes
Conclusion: HbA1c testing is increasing rapidly, especially when

used to screen for diabetes, and screening now exceeds surveillance
as the chief reason for HbA1c requests. Making the purpose of
HbA1c requests explicit in this way offers potential to define the
large number of people at risk of diabetes, offering new
opportunities for diabetes prevention.
P437
The role of non-invasive screening methods
in estimating prevalence of diabetes in
patients admitted to hospital with acute
coronary syndrome
MA Karamat1,2, W Hanif3, SE Manley4, AA Tahrani1,2 and
MJ Stevens1,2
1
Diabetes, Heartlands Hospital, Birmingham, UK, 2Institute of Biomedical
Research, University of Birmingham, Birmingham, UK, 3Diabetes, Queen
Elizabeth Hospital, Birmingham, UK, 4Biochemistry, Queen Elizabeth Hospital,
Birmingham, UK
Aims: To determine the prevalence of undiagnosed diabetes and

impaired glycaemic state (IGS) and compare WHO 1998 and IEC
criteria in patients admitted to hospital with acute coronary
syndrome. In addition we also looked at using novel screening
algorithms to determine glycaemic status.
Methods: This was a prospective 3 year study carried out in two
large inner city hospitals in the UK.
Results: In all 118 patients were included in the analysis. The
prevalence of diabetes was 20% and 16% respectively according to
the WHO and IEC criteria at baseline. The prevalence of diabetes
remained similar at 3 months at 21%. However, two-thirds of
participants with IGS and a third of those with diabetes changed
their glycaemic status at 3 months. This could possibly be due to
stress hyperglycaemia as urinary cortisol creatinine ratio was
elevated in patients who had Type 2 diabetes at baseline compared
with NGT and IGS. The two diagnostic criteria appeared to
identify different cohorts of patients. Our screening algorithm had
a sensitivity of over 85% at baseline in comparison with WHO
1998 criteria. We also designed a diabetes predictor score based on
age, fasting plasma glucose and HbA1c and it had an excellent
sensitivity of over 80% and negative predictive value of over 90%.
By contrast the sensitivity of IEC criteria was only 57%.
Conclusion: The WHO and IEC diagnostic criteria identify
different populations with diabetes at baseline as well as 3
months. This is clinically relevant as we are basing screening in a
high risk population on these criteria. The IEC criteria do not
identify patients with IGS which is known to be associated with
increased cardiovascular morbidity and mortality.
P438
Screening for diabetes in the SW peninsula:
assessing the use of different risk scores
and HbA1c
F Brown1, C OMahoney1, S Creanor2, N Shaw1 and
BA Millward1
1
Diabetes Clinical Research Centre, Plymouth University Peninsula Schools of
Medicine and Dentistry, Plymouth, UK, 2Centre for Medical Statistics and
Bioinformatics, Plymouth University Peninsula Schools of Medicine and
Dentistry, Plymouth, UK
Aim: To compare the utility of the Diabetes UK and Cambridge

riskscores together with HbA1c in predicting diabetes (Type 2
diabetes) and pre-diabetes in a Caucasian population in the
southwest peninsula of England.
2014 The Authors.

DIABETICMedicine
Background: Diabetes is a serious condition with macrovascular

and microvascular complications damaging blood vessels, nerves,
kidneys and eyes; modifiable risk factors include weight and blood
pressure. Tools available to determine a persons risk of developing
Type 2 diabetes include the Diabetes UK riskscore (DRS), the
Cambridge riskscore (CRS) and HbA1c testing; these tools perform
differently.
Methods: DRS and CRS were calculated for 1,044 volunteers
who had consented to participate in diabetes research; none had
known diabetes. Traffic light colours were assigned to the DRS
(0 6 green; 7 15 flashing amber; 16 24 amber; 25 red). The
CRS calculates the Cambridge Probability Score. HbA1c levels
(mmol/mol), available in 823 subjects, were assigned a traffic
light colour (<37 green; 37 41 yellow; 42 47 amber (high
risk); 48 red (likely diabetes). The two riskscores were compared
and HbA1c was added to those with a positive DRS (16) or CRS
(0.128) score.
Results: DRS identified 591/818 (72.3%) and CRS 602/818
(73.6%) subjects at high risk of developing diabetes. HbA1c was
applied to those identified as high risk on either score (653/818).
This identified 26 (3.2%) with likely diabetes and 87 (10.6%) at
high risk.
Conclusions: Simple tests of diabetes riskassessment can ascertain
those with/at risk of diabetes in primary care. Utilizing stratified
risk assessments with more than one assessment tool may be a
cost-and time-efficient method of screening for diabetes and
pre-diabetes.
P439
Targeted screening of a high risk population
using HbA1c identifies a high yield of
patients with impaired glucose metabolism
(IGM)
HJ Wallace1, S Ferguson2, D Bernatowicz2, J Dougan2 and
SJ Hunter1
1
Regional Centre for Endocrinology and Diabetes, Royal Victoria Hospital,
Belfast, UK, 2Rapid Access Chest Pain Clinic, The Heart Centre Northern
Ireland Royal Victoria Hospital, Belfast, UK
Aims: The prevalence of IGM (impaired fasting glucose, impaired

glucose tolerance and Type 2 diabetes) is high. An improved
strategy of identification is essential to offer earlier care and
improve outcomes. We sought to assess the effectiveness of
identifying patients with IGM using a targeted screening
approach in the chest pain clinic, a group of patients who are at
high risk of developing Type 2 diabetes.
Methods: There were 331 patients attending the chest pain clinic
over a 14 week period; 303 patients were not previously known to
have diabetes. These patients were risk stratified after completion
of the American Diabetes Association (ADA) risk assessment
score. Those who were deemed to be high risk underwent HbA1c
testing. Patients were considered to have diabetes or pre-diabetes
when HbA1c measured 48mmol/mol or 38 47mmol/mol
respectively, as per ADA guidelines.
Results: One hundred and sixty-two (53.5%) patients were
classed as high risk on the basis of the risk assessment score. Of
the 148 patients who had HbA1c measured, 54(36.5%) were
defined as pre-diabetic. Five patients (3.4%) were diagnosed with
Type 2 diabetes. Only 13 patients (4.5%) were found to have an
elevated random glucose and considered to have IGM when using a
random plasma venous glucose, the standard method of testing
used in the chest pain clinic.
159
DIABETICMedicine
Conclusions: The risk assessment questionnaire is a simple, quick

and inexpensive tool which can be used to identify patients at high
risk of IGM who should undergo further investigation. Targeted
screening using HbA1c identifies a significantly higher yield of
patients with previously unknown IGM compared with random
plasma venous glucose alone.
P440
Corneal confocal microscopy detects early
nerve damage in subjects with impaired
glucose tolerance and predicts progression
to diabetes
S Azmi1, I Petropoulos1, MF Ferdousi1, G Ponirakis1, U Alam2,
HF Fadavi1, M Tavakoli1, A Kheyami1, M Jeziorska1 and
RA Malik1,2
1
Centre for Endocrinology and Diabetes, Institute of Human Development,

University of Manchester, Manchester, UK, 2Diabetes Centre, Central
Manchester NHS Foundation Trust, Manchester, UK
Aim: To identify neuropathy in subjects with impaired glucose

tolerance (IGT) using corneal confocal microscopy (CCM) and
assess whether baseline pathology predicts outcomes in relation to
glucose tolerance.
Methods: Thirty subjects with IGT and 20 control subjects
underwent functional [warm threshold (WT) and warm induced
pain (WIP)] and structural [corneal nerve fibre density(CNFD),
branch density(CNBD), fibre length(CNFL), intra-epidermal nerve
fibre density (IENFD)] assessment of small fibre damage at baseline
and 12 months.
Results: At baseline IGT and controls were matched for age
(57.8 2.1 vs 58.9 1.76, p = 0.342), blood pressure (mmHg)
(136.92 3.71/74.6+1.93
vs
136 3.69/75.53 2.31,
p = 0.79), cholesterol (5.05 0.23 vs 5.5 0.22,p = 0.134)
and triglycerides (2.18 0.3 vs 1.89 0.2,p = 0.84) but not
HbA1c (6.15 0.9 vs 5.69 0.62%, p < 0.000) or body mass
index (31.5 0.89 vs 27.68 0.9kg/m2, p = 0.007). There was a
significant elevation in WT (40.01 0.89 vs 38.33 0.59,
p = 0.03) and WIP (47.26 1.12 vs 45.95 0.6, p = 0.003)
and reduction in CNFD (23.28 1.25 vs 30.77 1.53,
p = 0.001), CNBD (28.46 2.66 vs 37.03 2.71, p = 0.025)
and CNFL (14.39 0.63 vs 20.43 3.14, p = 0.005) but not
IENFD (7.75 0.84 vs 8.93 0.65, p = 0.77) in subjects with
IGT. Six subjects progressed to Type 2 diabetes and five regressed
to normal glucose tolerance (NGT) at 1 year. Whilst those who

progressed to Type 2 diabetes had a lower baseline CNFL
(13.03 1.15) and CNFD (19.96 2.27), CNBD increased
significantly in subjects who regressed to NGT (27.08 5.6 vs
37.97 6.82, p = 0.046) and reduced in those who developed
Type 2 diabetes (26.82 4.69 vs 18.7 4.11, p = 0.046).
Conclusion: CCM demonstrates significant small fibre damage in
subjects with IGT. Subjects who progressed to Type 2 diabetes had
greater corneal nerve pathology at baseline and showed reduced
regeneration compared with those who became glucose tolerant.
P441
Screening for diabetes following pancreas
transplantation
A Nicholls, F Akiboye, D Cavan and D Kerr
Bournemouth Diabetes and Endocrine Centre, Royal Bournemouth and
Christchurch Hospitals NHS Foundation Trust, Bournemouth, UK
Background: There is little evidence to guide the screening of

graft function in patients post simultaneous kidney pancreas (SPK)
or pancreas transplant alone. There is some evidence that an oral
glucose tolerance test is a better predictor of future graft failure.
Aim: To determine current practice in the screening for graft
failure following SPK or pancreas transplant.
Method: We looked at all patients managed in a district general
hospital who had undergone pancreas transplantation in the 5
years 2006 2011. Our primary outcome was to check whether
patients had a screen for diabetes on an annual basis. We
determined those who were tested with HbA1c, oral glucose
tolerance test (OGTT) or both.
Results: Twenty-two patients (18 SPK) were transplanted in the
years 2006 2011. Eight (36.4%) of these patients remained under
the follow-up of a diabetologist because of the presence of
significant pre-existing complications of diabetes, usually
neuropathy. Nineteen patients (86.4%) had an annual HbA1c.
Only two patients (9.1%) had an oral glucose tolerance test within
the last 12 months. Four patients (18.2%) were restarted on
treatment for diabetes within 5 years of transplantation.
Conclusion: Screening for diabetes post-transplantation is
variable, and not usually carried out by a diabetologist. Due to
the lack of clear guidelines there may be some patients who are
either missed from screening or who may benefit from additional
testsor review by a diabetologist.
Clinical care and other categories posters: secondary care

P442
Glycaemic control in relation to insulin
doses and treatment intensity in a large
secondary care diabetic clinic
Aims: We aimed to assess the achievement of glycaemic targets in

relation to the insulin regimen, total insulin dose and the type of
insulin used, in a large secondary care diabetic clinic.
Methods: Data from 1,048 patients, on insulin for at least 3
months, representing 94% of all insulin treated patients with Type
2 diabetes under our follow-up were available for analysis. Data

were collected from our clinical database.
Results: There were 600 male and 448 female patients with a
mean age (SD) of 66.4 11.8 years and a body mass index
(BMI) of 33.2 1.2kg/m2 on a total daily insulin dose of
67 2U, equivalent to 0.72 0.52U/kg. 137 (13%) patients
were on insulin once/day, 576 (55%) on twice/day, 102 (10%) on
thrice/day and 233 (22%) on four or more injections/day. Patients
on analogue insulins were as follows: 234 (94%) out of 248 on
rapid acting insulin, 538 (81%) out of 668 on pre-mixed insulin
and 314 (84%) out of 376 on long acting insulin. HbA1c
was 7.5% (group1) in 342 (33%) patients, 7.5% 8.0%
(group2) in 152 (14%), 8.0% 9.0% (group3) in 245 (24%),
9% 10% (group4) in 157 (15%) and >10% (group5) in 150
160
2014 The Authors.

LS Cozma, S Benjamin and V Eligar

Diabetes Centre, Princess of Wales Hospital, Abertawe Bro Morgannwg
University Health Board, Bridgend, UK
Clinical care and other categories posters: structure/systems of care and healthcare delivery
(14%) patients. Patients in group 5 were on significantly more

complex insulin regimens compared with groups 1 and 2. Insulin
dose per kilogram was significantly higher in groups 3, 4 and 5 vs
group1 (ANOVA across groups, p < 0.0001). BMIs were
significantly higher in groups 4 and 5 vs group1 (ANOVA across
groups, p < 0.0001).
Conclusion: The majority of insulin treated patients with Type 2
diabetes are obese and require large doses of insulin to control their
diabetes. Good (HbA1c 7.5%) or acceptable (HbA1c 8.0%)
glycaemic control is achieved in less than half of our patients.
P443
Clinical profile of patients with Type 1
diabetes attending an Irish teaching
hospital
TG Cotter, T OBrien, SF Dinneen, F Dunne, M Bell,
FM Finucane and M Gately
Endocrinology, Galway University Hospital, Galway, Ireland
Objectives: (1) To characterise a cohort of patients with Type 1

diabetes (Type 1 diabetes) attending a diabetes clinic in a teaching
hospital in Ireland and (2) to compare outcomes with the 2011
Scottish Diabetes Survey and the American Diabetes Association
(ADA) recommendations.
DIABETICMedicine
Methods: This was a retrospective review of the most recent

clinical and laboratory records of Type 1 diabetes patients who
were managed at Galway University Hospitals Diabetes Centre
between June 2011 and June 2013.
Results: In total 905 patients, including children, were eligible for
inclusion. There were 474 (52%) males and 431 (48%) females. The
mean age was 37 years (SD 16.78). The average body mass index
(BMI) was 25.7kg/m2 (SD 5.14) and 159 (18%) had BMI 30. The
average glycosylated haemoglobin (HbA1c) was 70.3mmol/mol
(SD 17.7) with 117/797 (15%) of adults having an
HbA1c < 53mmol/mol. 613/872 (70%) had blood pressure (BP) <
140/80mmHg. 453/736 (62%) had LDL cholesterol (LDL-C)
< 2.6mmol/l. 183/769 (23.8%) had increased albuminuria. 102/905
(11%) had documentation of loss of hypoglycaemic awareness.
Conclusion: ADA targets for BP and LDL-C were achieved for
most patients. 15% of adults were within the ADA target for
HbA1c demonstrating the challenge in obtaining optimal
glycaemic control in this patient population. BP, HbA1c and
BMI results compare favourably with the Scottish Diabetes Survey.
These data have prompted the restructuring of diabetes clinics to
cohort patients with Type 1 diabetes to a dedicated clinic and place
particular emphasis on glycaemic control through use of structured
education and technology. Improved documentation of
hypoglycaemia is a future goal. Improved BP control and LDL-C
levels also represent ongoing challenges.
Clinical care and other categories posters: structure/

systems of care and healthcare delivery
P444
Integrated diabetes care: the benefit of
e-consultations
P447
M Mohammad and M Freeman
J Burke and D Mellor
Diabetes and Endocrinology, Dewsbury District Hospital NHS Trust,

Dewsbury, UK
Department of Clinical Sciences, University of Chester, Chester, UK
Refer to Oral A10
P445
Adherence to local protocol for the initial
management of adult patients presenting
with diabetic ketoacidosis to a district
general hospital
DH Foster, C McCarthy and A Helmy
General Medicine, South West Acute Hospital, Enniskillen, UK
Refer to Oral A45
P446
The InterCare Diabetes Service: a whole
system integrated approach to diabetes
care
P King1, K Gale1, S Ashton-Cleary1, P Dindsa1, K Farrell2 and
M Munir3
1
Diabetes and Endocrinology, Derby Hospitals NHS Foundation Trust, Derby,
UK, 2Willington Partnership, Willington, UK, 3Wilson Street Surgery, Derby, UK
Refer to Oral A46
2014 The Authors.

Knowledge of nutrition for diabetes

management among student nurses
Aims: Research suggests that access to specialised dietetic services

for patients with diabetesis not universal. As the front line of patient
care, nurses are providing a significant amount of nutritional advice
although their knowledge of nutrition and diabetes has been found
to be limited. This study investigated the level of knowledge of
nutrition for diabetes management among student nurses.
Methods: One hundred and ten pre-registration student nurses
without prior notice completed a self-administered, 43-item
questionnaire assessing the levels of clinical knowledge of
nutrition specific to diabetes, under near examination
conditions. The knowledge question scoring procedure attempted
to ensure that answers to all questions were treated consistently as
being either correct or incorrect including those within multiple choice
questions. The adequacy of knowledge was tested using a one-sample
Wilcoxon signed rank test (p < 0.05) to compare the median score of
the sample group against the university grade for competency.
Results: Seventy percent of the nurses were found to be competent
or above. Whilst not statistically significant, nurses who had
professional dietetic input into their education programme had a
higher median score than those who had not. Of clinical
importance, only 28% of participants correctly identified the best
method in treating a patient with symptomatic hypoglycaemia.
Conclusion: Our data indicated that the student nurses did
generally have an adequate knowledge of nutrition for diabetes
management and suggest that interdisciplinary education could be
161
DIABETICMedicine
relevant to the Quality and Outcomes Framework indicator

DM013, which records annual dietary reviews undertaken by a
suitably competent professional.
P448
An integrated diabetes service facilitates
the transition from inpatient to outpatient
and is associated with reduced readmissions
MB Whyte, A Cox, GM Noble-Bell, M Boerkamp and R Malik
Diabetes, Kings College Hospital NHS Trust, London, UK
Aims: Patients with diabetes have high risk of readmission following

discharge. International standards of care include follow-up within
1 month for patients with in-hospital dysglycaemia. We reviewed our
urban tertiary care hospital service that provides diabetes care from
admission through to clinic review 1 month post discharge.
Methods: All emergency admissions are reviewed on day 1 on the
medical assessment unit for the presence of diabetes and/or
hyperglycaemia. An electronic referral system was established for
other wards. Management and education was delivered on the
ward, reinforced where necessary via telephone and reviewed at a
dedicated discharge clinic (DC) within 1 month of discharge.
Results: Over 12 months 199 new referrals were seen at DC; 52
did not attend (21%); 16% Type 1 diabetes, 60% Type 2 diabetes,
14% ketosis-prone diabetes, 10% other. Median (IQR) time from
discharge to DC was 28 (21 45) days. 97 (49%) received
telephone advice between discharge and DC. From clinic, 30%
discharged to primary care, 42% secondary care follow-up, 28%
further review at DC. 147 (74%) had inpatient HbA1c measured;
HbA1c dropped from 11.7% to 9.2% at DC (p < 0.001) in those
with inpatient HbA1c 7.5%. Three of 199 (1.5%) patients
attending DC were readmitted within 6 months with diabetes
destabilisation vs four of 51 (7.8%) who did not attend the DC
(p = 0.03). The magnitude of fall in HbA1c between admission and
DC was associated with reduced risk of all-cause readmission
(p = 0.02) but not the absolute HbA1c (either as inpatient or at DC).
Conclusions: For patients who demonstrate poor inpatient
control, a package of care that continues after discharge may
help reduce all-cause and diabetes-specific readmission.
P449
An evaluation of the impact of virtual clinics
for people with Type 2 diabetes in primary
care
N Basudev1, A Forbes2 and S Thomas3
1
Diabetes Intermediate Care Team, Lambeth Clinical Commissioning Group,
London, UK, 2Department of Primary and Intermediate Care, Kings College
London, London, UK, 3Department of Diabetes and Endocrinology, St
Thomas Hospital, London, UK
Objective: Many patients with Type 2 diabetes (Type 2 diabetes)

experience suboptimal management in primary care. This
study assessed the impact of a collaborative virtual clinic model
of care (DViC) on patients with Type 2 diabetes, in terms of
metabolic outcomes, care assignment and organisation of care. The
setting was primary care (n = 6 general practices in inner London)
and the participants werepatients with poorly managed HbA1c
69mmol/mol and diabetes 12 months diagnosis.
Methods: A parallel arm randomised, controlled clinical trial
compared clinical outcomes (primary outcome HbA1c) in
patients randomised to intervention (DViC, n = 93) or control
(usual care, n = 116) groups. Patients were followed up for
6 months. Analysis was by intention to treat.
162
Results: A third of patients in both groups were on suboptimal

glucose lowering medications at baseline (n = 72, 34%). There
was a significant improvement in HbA1c of 7mmol/mol in the
intervention group (95% CI 2 10, p = 0.005), but no difference
compared with the control group (p = 0.785). More medication
changes were made in the control group (51% vs 31%, p = 0.001).
Conclusions: Many people with Type 2 diabetes continue to
experience suboptimal clinical care. This study shows that care
collaboration via the DViC is a viable option for addressing poor
diabetes control with no adverse effects. While the data did not
show superiority for the DViC compared with usual care, net
benefits were identified in the overall care patients received. The
lack of a differential effect between the study groups may be
attributable to the fact that the DViC is not isolated and that
changes in clinician behaviour following the DViC may have
extended to both study groups.
P450
A paradigm shift in the management of
diabetic ketoacidosis: the introduction of
the Joint British Diabetes Societies 2010
guideline into the Royal Surrey County
Hospital
C van Zeller, T Milner, R Herring and D Russell-Jones
Cedar Centre, Royal Surrey County Hospital NHS Foundation Trust, Guildford, UK
Aims: In light of the recent development of the Joint British

Diabetes Societies (JBDS) 2010 diabetic ketoacidosis (DKA)
guideline, this audit aimed to assess DKA management at the
Royal Surrey County Hospital.
Methods: DKA patients were identified via medical coding and a
retrospective audit of clinical records was performed between January
2011 and March 2013 assessing practice against current guidelines.
Results: The audit identified 20 patients, with an average age of
31.7 years and a male to female ratio of 11:9. Patient care was
delivered in 11 separate wards. Significantly fewer deviations from
the guideline occurred when patients were on high input wards
[intensive care unit (ICU) and acute dependency unit]
(OR = 0.0004, 95%CI < 0.005 0.51, p = 0.0457). Initial
management was well performed, with a short average time to
diagnosis (69min) and treatment (88min). However, ongoing ward
care could be improved: only 54% of guideline recommended
blood tests were performed in the first 12h of admission, 70% of
patients did not receive adequate potassium replacement, and
average time on sliding scale (51.6h) was long.
Conclusions: Following assessment of audit outcomes, a focused
introduction of the JBDS guideline occurred. The first 24h of
management would occur in three clinical areas: accident and
emergency, ICU and the emergency assessment unit. A business
case was developed for capillary blood ketone monitoring and staff
education. The JBDS guideline introduction adopted a multimodal
approach, using audit-based learning, staff education through
trust-wide presentations, ward-based teaching, and introduction of
a proforma to guide adequate monitoring and treatment.
2014 The Authors.

P451
Prevention of recurrent glycaemia related
admissions through case management and
mental health intervention
D Simmons1, S Hartnell1, J Watts1, E Gunn2, A Jenaway2,
C Ward1 and K Davenport1
1
Wolfson Diabetes and Endocrinology Clinic, Cambridge University Hospitals,
Cambridge, UK, 2Psychiatry Liaison, Cambridge and Peterborough
Foundation Trust, Cambridge, UK
Background and aims: Recurrent admissions for hyperglycaemia

or hypoglycaemia often reflects a range of psychological,
educational, social and/or health service related barriers to
care. We have tested whether readmissions from acute glycaemic
events can be prevented.
Methods: A novel joint diabetes/mental health service was funded
for 12 months (173,228). The service included 2.0 diabetes
specialist educators, 0.1 diabetologist, 0.05 psychiatrist and 0.4
psychological wellbeing practitioner. An inpatient identification/
adoption process with post-discharge ambulatory diabetes, social
and mental health case management was implemented. A Barriers
to Diabetes Care framework was used to identify and
systematically address the problems that impacted on this group
of patients including links with a variety of outside agencies (social
services, welfare and benefits offices, citizens advice, county
council housing departments, community mental health services).
Results: Over the first 7 months, 44 patients (mostly Type 1
diabetes) were accepted: 23 (52%) warranted a mental health
review. Expedited clinical interventions occurred including DAFNE
(Dose Adjustment for Normal Eating, n = 5), insulin pump therapy
(n = 7), cognitive behavioural therapy (n = 6) and cognitive
analytical therapy (n = 2). Acute glycaemia readmission was
reduced 60% from 15 to 4 6/month. Glucose control improved
[HbA1c fell from 97 20 (11.0 4%) to 90 21 mmol/mol
(10.4 4.1%) (paired t-test p = 0.0.35)] and length of stay
dropped (4.3 to 2.5 days/admission). Estimated financial savings
from admission avoidance was 265,032: predicted annual net
saving was 91,804 return on investment approximately 1.5:1.
Conclusions: A barriers based approach by an integrated
diabetes/mental health service for patients with recurrent
glycaemia related readmissions can prevent readmission, reduce
length of stay, improve glucose control and is cost effective.
P452
Mapping the diabetes landscape: matching
specialist community resource to areas of
need
SL Woodman, KE Fayers, H Alsafadi and C Atkinson
Foundation Trust, Lyndhurst, UK
Aims: West Hampshire Community Diabetes Service was

launched in 2010 alongside a locally enhanced service agreement
(LES) to support insulin initiation. This survey identified the level
of engagement of practices with diabetes professional education
programmes in West Hampshire Clinical Commissioning Group
(CCG), combined with diabetes outcomes for each practice. A
scoring system was developed to identify practices most in need of
specialist support.
Methods: Inter-professional learning (IPL) students undertook a
2 week survey of local practices and recorded levels of practice
nurse and GP training against diabetes outcomes/admissions
(2011 2012 data).
2014 The Authors.

DIABETICMedicine
Results: 72% of the 54 practices signed up to the LES for insulin

initiation. 74% of lead GPs have attended diabetes training (83%
were up to date within 2 years). 92% of lead practice nurses
attended diabetes training (89% were up to date). Marked
variation was observed for all parameters across all practices,
expressed as a percentage of patients on the practice diabetes
register (key care processes 85% 97%, diabetic ketoacidosis
admissions 0% 3.8%, hypoglycaemia admissions 0% 1.6%,
HbA1c below 53mmol/mol 55% 82%, exemption rate
reporting 3.3% 26.5%). Practices were given an arbitrary score
for each parameter. Higher scores indicated poorer performance.
We identified the top and bottom performing practices.
Conclusion: These scores indicate the complexity of assessing the
quality of diabetes care across a large CCG. As a community
service we aim to target appropriate resource to best support
patients and reduce postcode variation in primary care. We have
used this information to target diabetes nurse support for practices
in 2013 2014.
Acknowledgement: We wish to acknowledge the interprofessional learning students from the Universities of
Southampton and Portsmouth who collected the data.
P453
Factors influencing safe and appropriate
discharge from a secondary care rapid
access clinic (RAC) to primary care
PJ Evans1, M Walsh1 and AL Lewis2
1
Diabetes Unit, Royal Gwent Hospital Newport, UK, 2Diabetes Unit, Royal
Gwent Hospital, Newport, UK
Background: The RAC

is
a
secondary
care
based
multidisciplinary clinic for the initiation of insulin therapy set up
in 2006. It has proved to be an invaluable service for primary care
patients requiring insulin initiation. Whilst primarily for insulin
initiation, GLP-1 therapy now predominates. Follow-up and
discharge of patients back to primary care has proved difficult.
Aims: To develop a seamless pathway between primary and
secondary care for patients with poorly controlled Type 2 diabetes
requiring injectable therapy initiation.
Methods: A clinical review of the RAC activity was undertaken
on 67patients (44 male,23 female) with an average baseline
HbA1c of 9.8% over a 12month period. Twenty-two surgeries
were involved.This review was carried out by a diabetes education
nurse facilitator from Novo Nordisk Ltd. An analysis of diabetes
education and skills was collected, by completion of a
questionnaire by all local referring surgeries with a 50%
response rate.
Results: An average reduction of 1.4% in HbA1c was
demonstrated. In contrast with 2006 50% of patients were
commenced on GLP-1 therapy. Whilst aiming for one follow-up
visit 32% of patients were reviewed on more thantwo occasions.
The questionnaire highlighted education deficiencies at a number
of surgeries inhibiting discharge back to the general practitioner.
Outcomes: We aim to improve communication between primary
and secondary care and implement appropriate education
opportunities and mentorship programmes for practice nurses
and general practitioners. This process will be audited annually.
163
DIABETICMedicine
P454
Clinical audit of treatment of diabetic
ketoacidosis (DKA): the Joint British
Diabetes Societies (JBDS) guideline (2010)
in practice
B Freudenthal1, J Ironside1, T Khilji2, F Rizvi1 and S Russell1
1
Department of Diabetes and Endocrinology, Chase Farm Hospital, Enfield,
UK, 2Department of Medicine, Chase Farm Hospital, Enfield, UK
Introduction: DKA is a common diabetic emergency. Recent

advances have brought significant reductions in mortality. The
JBDS DKA guideline (2010) brought several evidence-based
innovations including blood ketone monitoring, continuation of
subcutaneous basal insulin and treatment with a fixed rate insulin
infusion (FRII).
Methods: Fifty case notes were identified of patients treated for
DKA at Chase Farm Hospital over a 15 month period from the
local introduction of the new guideline (May 2011). They were
audited for comprehensive indicators of care quality and protocol
compliance.
Results: At initiation of treatment, 20% did not meet correct
diagnostic criteria, 38% did not receive intravenous fluids
within 1h of arrival and 47% of those on regimens including
subcutaneous basal insulin did not receive it. Regarding
rehydration and correction of ketoacidosis, 28% did not receive
an FRII, 36% did not have variable rate dextrose infusion started
correctly and 32% did not receive saline infusion throughout the
treatment. Regarding monitoring, no patients had capillary blood
ketones measured, only 54% had urine ketones measured at least
12 hourly, only 34% had blood gases analysed 46 hourly and
only 44% had an accurate fluid chart. No incidents of significant
harm were identified.
Conclusions: Improving and updating the management of DKA
continues to be a challenge as the new guidelines are introduced
locally. Education of doctors of all grades practising general
medicine needs to be a renewed priority. As with many
institutions, blood ketone monitoring continues to be
unavailable. We believe that rigorous utilisation of only
designated wards for diabetes care may further improve standards.
P455
Diabetes management in care homes: steps
in the right direction?
E Sloan1 and S Bellary1,2
1
Aston Research Centre for Healthy Ageing, Aston University, Birmingham,
UK, 2Diabetes and Endocrinology, Heart of England NHS Foundation Trust,
Birmingham, UK
Aims: Previous research has identified several inadequacies in

management of diabetes within care homes many of which were
highlighted in Diabetes UKs report Diabetes in care homes:
awareness, screening, training. The aim of this study was to see if
this was still the case and to identify specific areas for
improvement.
Methods: Thirty care homes in Birmingham were invited to
participate in the study. Data were collected using a standard
questionnaire based on the Diabetes UK national survey of
care homes comprising questions relating to screening, selfmanagement, care planning and local authority support. All
returned responses were analysed.
Results: Responses were received from 20 of the 30 care homes
approached. The mean percentage of residents with diabetes in the
care homes sampled was 13.7%. None of the homes screened for
164
diabetes on admission and only 5% screened residents annually.

80% of homes acknowledged providing diabetes-specific training
to staff. Residents in 95% of homes had a medical review in the
last 12 months: 70% with a GP, 20% with a diabetes specialist
nurse/nurse. 65% of homes provided support for self-management.
45% of care homes did not have individualised care plans for
residents with diabetes. 35% of managers reported poor support
and guidance from their local authority.
Conclusions: Improvements were noted in the care provided to
individuals with diabetes living in care homes in
Birmingham. Aspects relating to screening, individualised care
plans and support to care home staff still need attention.
P456
A decade in diabetes specialist services
2000 2011: the views of consultant
diabetologists and diabetes specialist
nurses amidst persistent healthcare
delivery change
CA Gosden1, T Tinati2, K Barnard1, DRR Williams3 and
RIG Holt1
1
Southampton, Southampton, UK, 2Medical Research Council, Lifecourse
Epidemiology Unit, University of Southampton, Southampton, UK, 3College
of Medicine, Swansea University, Swansea, UK
Aims: In 2000 and 2006 quantitative and qualitative data were

collected on the status of UK diabetes specialist services. Since
2006 further significant NHS changes have occurred. This study
aimed to explore the views of diabetes specialist staff on issues
impacting service delivery in 2011 and to track changes over time.
Methods: Five case study locations were purposively selected
from Diabetes UK and ABCD studies in 2000 and 2006. After
initial piloting, seven semi-structured interviews with either
consultants or diabetes specialist nurses or both were carried out
within the locations. These were recorded and transcribed.
Framework analysis was used to generate themes in a grounded
theory approach. These were compared with qualitative data from
2000 and 2006.
Results: Four interacting issues emerged: opportunities for
service redesign through reinvention and strong relationships
with key personnel to ensure success; within their own team
dynamics participants identified competent and motivated
multidisciplinary teams, with professional pride in delivering
high quality care; however, their ability to do their job was
impacted by ever decreasing resources and capacity to cope with
rising diabetes prevalence, pressures from acute medicine and
reduced staffing levels; in the current state of confusion,
participants own emotions ranged from anxiety, frustration,
uncertainty and disappointment at changes leading to unease and
low morale.
Conclusion: Those delivering services were willing to innovate
and work differently to deliver higher quality services. However,
lack of resources and feelings of frustration over persistent national
policy change have been expressed by diabetes specialist staff for
the last decade.
2014 The Authors.

P457
Online outpatient consultations: the patient
perspective
D Campbell-Richards1, J Morris1, R Sudra1, S Vijayaraghavan1,
T OShea2, T Greenhalgh3 and A Collard4
1
Diabetes and Endocrinology, Barts Health, London, UK, 2Diabetes, East
London Foundation Trust, London, UK, 3Global Health, Policy and Innovation
Unit, Barts and the London School of Medicine and Dentistry, London, UK,
4
Independent Consultant, London, UK
Background: The local diabetes service faces challenges familiar

to others including cost pressures, rising demand and poor
outcomes in the face of poor patient engagements and user
experience.
Aim: This qualitative pilot study aimed to evaluate, from the
patient perspective, if and how online consultations could deliver
more effective accessible diabetes care whilst improving patient
experience.
Methods: Using everyday technology already available in
peoples homes, online web-based appointments were offered to
all suitable patients (selected appointments), under the care of one
diabetes consultant and nurse specialist. Patient perspectives were
captured from a purposively selected sample of patients, with 19
semi-structured telephone interviews and four focus groups. A
further 28 patients completed a semi-structured questionnaire
(response rate 51%). Focus group and interview data were
transcribed and grouped into topics and themes by two
researchers, and related to questionnaire findings.
Results: Patients considered online appointments: saved time;
more convenient; cheaper (significant for a minority); felt more in
control of the consultation which was more focused; influenced
their self-management and decision making; quality of care was at
least as good as face to face; over half would be more likely to
attend them. Patients who benefited the most were those in
employment/education, younger patients and those with mobility
issues.
Conclusions: Online appointments can provide more effective/
accessible care and improve patient experience, with implications
for do not attend rates. Further research has commenced to look
at the potential to impact patient self-management and diabetes
control.
P458
The development of a health
community-wide blood glucose testing strip
formulary
H Hardman1, EA Hackett2, D Jackson3, J James2, J Kaur3,
R Love2, P Navti4, R Oldridge5, T Rengaranjan2, L Walker2 and
KS Higgins2
1
Leicestershire Medicines Strategy Group, Leicester, UK, 2Department of
Diabetes and Endocrinology, University Hospitals of Leicester NHS Trust,
Leicester, UK, 3West Leicestershire Clinical Commissioning Group,
Loughborough, UK, 4East Leicestershire and Rutland Clinical Commissioning
Group, Leicester, UK, 5Leicester City Clinical Commissioning Group, Leicester,
UK
Introduction: Blood glucose testing strips (BGTS) that are 33%

cheaper than market leaders are now available. The prescribing of
less expensive BGTS could save 1.25 million annually within
Leicestershire and Rutland.
Aim: To develop a BGTS formulary which considers quality,
patient feedback and cost effectiveness to ensure robust decision
making.
2014 The Authors.

DIABETICMedicine
Methods: Stakeholders from the health community agreed

quality criteria that BGTS and compatible meters must meet for
inclusion in the formulary. Samples of the BGTS and meters were
requested from manufacturers. Those received were reviewed
against the quality criteria. Patient feedback and preference was
obtained for the shortlisted BGTS and meters.
Results: The quality criteria used related to the revised ISO 15197
accuracy standard, meter memory, no coding required, results in
mmol/l only and meter activation on strip insertion. Eleven BGTS
met the quality criteria. People with diabetes were subsequently
invited to evaluate these BGTS and meters. The three BGTS with
the highest feedback scores and lowest acquisition cost were
included as formulary choices: GlucoLab, Mylife Pura and
TRUEyou.
Summary: GlucoLab, Mylife Pura and TRUEyou BGTS offer
quality testing and are cost effective. If 50% of patients are
prescribed a formulary BGTS, the estimated annual savings are
625,000.
P459
Improved attendance at postnatal oral
glucose tolerance testing following service
transfer from secondary to primary care
setting
SL Palin1, ACJ Robinson1, J Pankhania1, N Baxendale1, S Bett1,
A Muotune2, N Singh2, L Helsby3 and S Liversedge3
1
Diabetes Department, Royal Bolton Hospitals NHS Foundation Trust, Bolton,
UK, 2Obstetric Department, Royal Bolton Hospitals NHS Foundation Trust,
Bolton, UK, 3Bolton Clinical Commissioning Group, Bolton, UK
Aims: Audit of attendance at oral glucose tolerance testing

(OGTT) 6 weeks post delivery showed attendance to be <50%
despite various measures to improve attendance. The service
was transferred into primary care to assess if this improved
attendance.
Methods: From April 2012, at the last scheduled antenatal clinic
visit at 36 weeks, we advised women with gestational diabetes they
would be invited for a 6 week postnatal OGTT at their general
practice. Practices were made aware of the change from hospital
based OGTT by letter, with arrangements for local enhanced
scheme payments. A standard letter was sent to practices following
delivery as well as the hospital discharge summary. Our diabetes
specialist midwife met with our project coordinator from primary
care to review patient attendance to ensure patients were not
missed. Our project coordinator checked attendance at OGTT.
Non-attenders were highlighted to the practice for further
appointments to be offered.
Results: In all, 115 women were identified as requiring postnatal
OGTT. 75 (65%) OGTTs were undertaken. Seventeen did not
undergo OGTT but did have HbA1c testing. Twenty-three (20%)
did not undergo any postnatal glucose assessment, with 12 patients
failing to attend, four no longer registered with the practice, six not
being offered follow-up and one not followed up due to difficulty
with identification at the practice.
Conclusion: Transfer of postnatal OGTT from secondary care to
primary care in this pilot project over 1 year showed improved
although suboptimal attendance. We have demonstrated higher
compliance with postnatal OGTT than others and propose
primary care based postnatal glucose assessment be considered
nationally.
165
DIABETICMedicine
P460
People with common mental health
problems and diabetes receive better
surveillance of diabetes related conditions
and equal surveillance of their diabetes in
primary care
AP McGovern, N Munro, T Chan, S Jones and S De Lusignan
Department of Healthcare Management and Policy, University of Surrey,
Guildford, UK
Aims: People with mental health illness are less likely to engage
with healthcare services. Pay-for-performance diabetes targets in
the UK should minimise the impact of this effect on diabetes
management. Here we report the impact of psychiatric conditions
on failure to receive regular check-ups for diabetes and related
conditions.
Methods: A cohort of people with diabetes (N = 35,502) from
the Quality Intervention in Chronic Kidney Disease (QICKD) trial
was followed up, over a period of 2.5 years, using routinely
collected primary care data. A logistic regression analysis was
performed to identify the impact of mental health on the
proportion of patients receiving HbA1c, cholesterol, renal
function and monofilament foot checks. Demographic factors,
comorbiditiesand variability between primary care practices were
adjusted for.
Results: In the cohort studied, 2,042 (5.8%) people had a
recognised affective disorder and 122 (0.3%) a recognised
psychotic disorder. People with affective disorders were slightly
more likely to have cholesterol measurements (odds ratio 1.36,
95% CI 1.08 1.70) and renal function checks (odds ratio 1.37,
95% CI 1.16 1.63). No association was found between psychotic
disorders and cholesterol measurement or renal function checks.
No association was found between mental health illness and
HbA1c and monofilament foot checks.
Conclusions: The presence of recognised psychiatric illness was
found to have no negative effect on the rates of assessment of
diabetes and improved surveillance of renal function and
cholesterol in those with affective disorders.
P461
The role of technology to improve resource
utilisation and support diabetes
self-management in clinical practice: an
audit of a virtual insulin pump clinic
B Mumford, V Oldham, D Lee and N Agarwal
Diabetes Centre, Cwm Taf Local Health Board, Merthyr Tydfil, UK
Background: Insulin pump therapy offers people with diabetes

greater flexibility with self-management, individualisation of
treatment and improved metabolic control, and has therefore
become an essential component of diabetes management.
However, delivering a highquality insulin pump service requires
greater resources from both the multidisciplinary healthcare
professional team and patients themselves.
Aims: To evaluate the effectiveness and outcomes of the first 12
months experience of a virtual insulin pump clinic.
Methods: A virtual clinic was introduced to support the local
pump service. Communication with the healthcare team was
supported by internet-based insulin pump therapy management
software. Data were collected on metabolic profiles, patient
satisfaction, hypoglycaemia and cost of clinical care.
166
Results: Over a 12 month period, 25 patients joined the virtual

clinic after commencing insulin pump therapy. Improvements
were noted in HbA1c (mean 82.6mmol/mol pre-pump vs
67.2 post-pump; p < 0.001), hypoglycaemia awareness and
hospitalisation rates secondary to acute diabetic complications.
An estimated 40,000 savings in cost were made from reducing
acute admissions and secondary care. Additionally, patients
reported greater satisfaction with their diabetes care, improved
compliance with healthcare advice and reduced absence from
individual workplaces for diabetes-related illness and clinic
appointments.
Summary: A virtual clinic allows effective management of
patients with diabetes. Furthermore, it provides increased
flexibility for patients and healthcare professionals to
communicate effectively, supports self-management, and reduces
cost by both improving clinical outcomes and resource utilisation.
P462
A data integration methodology for
targeting people with diabetes with key
adverse outcomes: the Wolverhampton
Interface Care Knowledge Empowered
Diabetes (WICKED) Project
SMR Gillani, MR Holland and BM Singh
Wolverhampton, UK
Introduction: Equity of access to diabetes healthcare is confused

within horizontal and vertical integration. The former is access to
systematic healthcare for all, the latter strives to ensure those
progressively at risk or greater need get appropriate care. It is hard
to integrate such complex care. In systematic models of diabetes
care there are many outcomes other than the standard clinical
markers. In WICKED, our local model of integrated care, three
(of five) outcomes are access; hospital readmissions; severe
hypoglycaemia. We describe our method of targeting such
patients as a preliminary step in providing and governing specific
and integrated care.
Method: From our integrated diabetes register, we identified all
patients (over 15 months) with a failed access; severe
hypoglycaemia; 3 non-elective hospital admissions. We identied
people who fall in one, two or all of these categories.
Results: Of 16,644 registered patients, 1,547/815 (2,362, 14%)
had partial/complete access failure; 470 (3%) had 3 non-elective
admissions; 143 (1%) had emergency department attendances
with severe hypoglycaemia; 1,378 (8%) fell into one or more
categories, 50 (0.3%) into two, none in all three. They were older
(p < 0.002) and had higher deprivation (p < 0.001), Type 1
diabetes prevalence (p < 0.001) and non-Caucasian ethnicity
(p < 0.001).
Conclusion: In WICKED, on the basis of access, risk, patient
centricity and targeted service delivery, our objective to identify
those with the stated care concerns was achieved. The need now
is to ensure effective focused integrated care from multiple
organisations, professions and disciplines. With targeted
multi-modal identification, an opportunity is afforded to audit
and performance manage for better outcomes.
2014 The Authors.

P463
How do people with diabetes in
England describe their experiences of
primary care: where can we improve?
CAM Paddison, CS Saunders, GA Abel, RA Payne and
M Roland
Cambridge Centre for Health Services Research, University of Cambridge,
Cambridge, UK
Aims: To describe the primary care experience of people with

diabetes in England, including those with additional comorbidities,
and to use this information to inform improvements in diabetes
care.
Methods: Using data from 906,578 responders to the 2012
General Practice Patient Survey (England), including 85,760 with
self-reported diabetes, we employed logistic regressions controlling
for age, gender, ethnicity and social deprivation to analyse patient
experience using seven items covering three domains of primary
care (access; continuity; communication).
Results: People with diabetes, on average, reported better
experience on six out of seven primary care items compared with
people without diabetes (adjusted differences controlling for age,
gender, ethnicity and deprivation 0.9% 3.2%, p < 0.001). Those
with diabetes and additional comorbid long-term conditions were
more likely to report worse experiences, particularly for access to
primary care appointments, than patients with diabetes alone
[compared with patients with no long-term conditions: odds
ratio 1.15 (95% CI 1.08 1.23) diabetes alone, odds ratio 0.84
(95% CI 0.79 0.90) diabetes plus three or more conditions].
Conclusions: People with diabetes in England report primary care
experiences that are at least as good as those without diabetes for
most domains of care. However, improvements in care are needed:
these include improving the patients ability to see their preferred
doctor, involving patients in care decisions, and improving
primary care particularly for diabetes patients with comorbid
long-term conditions.
Acknowledgement: This study was funded by Diabetes UK.
P464
Development of a care planning tool to
facilitate structured, informed, patient
empowered and led diabetes care: the
Wolverhampton Interface Care
Knowledge Empowered Diabetes (WICKED)
Project
SMR Gillani and BM Singh
Wolverhampton, UK
Introduction: Structured diabetes care is effective in both primary

and specialist care. Conventionally, compliance models of care
empower healthcare professionals to take a lead role in care
delivery. Patient empowerment focuses on self-management,
shared responsibility and decision making where patients drive
their own care. We have devised a tool My diabetes, My
information, My plan to facilitate the process of patient driven
care.
Methods: In a three-cycle process involving clinicians and
patients, we developed a document containing individualised
information mapped against all core diabetes care processes. It
incorporated an adult reflective approach driving a Do, Review,
Learn, Apply process to improve insight, and it emphasised
2014 The Authors.

DIABETICMedicine
engagement, enablement and empowerment in the direction of

patient driven care. It was piloted in 50 consecutively attending
patients. Its utility was assessed by a structured questionnaire (16
questions).
Results: Of 50 patients, 37 responded (14 females, age 59 12,
seven Type 1 diabetes, 23 on insulin, 29 Caucasians). More than
80% positively reported this document to be understandable;
informative; helpful; and it promoted being in control/charge of the
diabetes. 90% never had such information before; they wanted it
again and said they would use it in consultation with professionals.
91% of people will use this information to take better care for their
diabetes.
Conclusion: Provision of a structured and individualised diabetes
report can inform, activate and empower people with diabetes to
improve their engagement. Use of such document in routine
consultations can provide a systematic way to address individual
patients needs and facilitate the care planning process in diabetes.
P465
The introduction of patient centric, patient
driven, care planning consultations in
specialist hospital clinics: the
Wolverhampton Interface Care
Knowledge Empowered Diabetes (WICKED)
Project
BM Singh and SMR Gillani
Wolverhampton, UK
Introduction: Diabetes care planning is a systematic way of

establishing a partnership between people with diabetes and
healthcare services. This partnership is based on principles of
patient centredness, empowerment and collaboration. Thus, we
piloted a systematic patient-driven care planning consultation
process in our routine specialist diabetes clinics.
Methods: A structured document based on all key care processes
was given to and completed by patients prior to their consultation.
Each domain was assessed and prioritised by the patient using an
adult reflective model. The completed document was used to
structure the clinic consultation and then the clinic letter (sent to
the patient, cc GP). Patient and doctor outcomes were assessed by
post consultation questionnaires.
Results: Of 148 patients, 101 responded(55 male, age 60 12
years, duration of diabetes 12 9 years, 67 on insulin, three new
and 98 review patients, 63% were Caucasians, 14% Asian, 4%
Afro-Caribbean and 19% unknown ethnicity). In 80%, globally,
the process scored very good, as in all other domains (n = 8)
including concerns addressed. Ten of 12 doctors rated the
structured patient driven consultation process good/very good,
reporting increased patient engagement, shared decision making,
communication, but felt it was more time consuming though
worthwhile. Crucially it increased insight into patients needs.
Conclusion: In patient-centric consultations, patients did use the
opportunity to assess, categorise and prioritise their health
concerns with a high degree of satisfaction; health professionals
found that of benefit; this was implementable within busy routine
specialist diabetes clinics.
167
DIABETICMedicine
P466
Review of HbA1c results >120mmol/mol as
patients may require urgent assessment of
request for diagnosis of Type 2 diabetes
RP Dowd1, RA Round2, CL Mason2, PG Nightingale3,
SG Ghosh1, W Hanif1, J Webber1, R Cramb2,4 and SE Manley2,4
1
Diabetes Centre, University Hospitals Birmingham NHS Foundation Trust,
Birmingham, UK, 2Clinical Laboratory Services, University Hospitals
Birmingham NHS Foundation Trust, Birmingham, UK, 3Wellcome Trust
Clinical Research Facility, University Hospitals Birmingham NHS Foundation
Trust, Birmingham, UK, 4Division of Medical Sciences, University of
Aims: HbA1c 48mmol/mol/6.5% was introduced in the UK for

diagnosis of diabetes in asymptomatic people aged >18 years after
publication of an expert position statement in 2012 based on
WHO recommendations from 2011. We aim (i) to review
HbA1c >120mmol/mol/13.1%/estimated
average
glucose
18.3mmol/l reflecting marked hyperglycaemia and requiring
urgent assessment and (ii) to identify daily cases of newly
diagnosed diabetes with high HbA1c, investigate their aetiology
and inform requestors accordingly as glucose >15mmol/l is
regularly phoned in these circumstances.
Methods: Lists of high HbA1c measured on Tosoh G8 IE HPLC
analyser were obtained from the laboratory data system between 1
January 2013 and 2 September 2013 with patients reviewed on
laboratory and clinical information systems.
Results: Of 68,737 HbA1c obtained, 380(0.6%) were high. With
68(18%) duplicates omitted, 312(82%) patients aged 53(16) years,
mean (SD), were identified with 169(54%) male and 80 (26%)
South Asian. HbA1c, median (interquartile range), was130 (125
139)mmol/mol/14.0% (13.6%14.9%) with maximum 192mmol/
mol/19.7%/28.7mmol/l. Requests originated from GPs 230(74%),
outpatients 24(8%) or wards 58(19%). Of these cases 246(79%)
had previously been diagnosed with diabetes and 66(21%) were
possibly new diagnoses with 43(14%) from GPs, three (1%)
outpatients and 20(6%) wards. South Asians with high results
were significantly younger than others (p = 0.048), aged 50(15)
years with 41(51%) being male.
Summary: Approximately one in 200 HbA1c results
were >120mmol/mol with 20% being possibly new diagnoses of
diabetes. The cases identified included young South Asians,
patients with diabetic ketoacidosis and pancreatitis, steroid or
antipsychotic induced diabetes.
P467
Diabetes care for vulnerable older patients:
the views of professionals working in care
homes and domiciliary care using a focus
group format
CJ Fox1, CR Gillespie2, A Kilvert1 and AJ Sinclair3
1
Research and Development Unit, Northampton General Hospital,
Northampton, UK, 2Diabetes Centre, Royal Derby Hospital, Derby, UK,
3
Institute of Diabetes for Older People, Beds and Herts Postgraduate Medical
School, Luton, UK
Aims: A recent Care Quality Commission report highlighted

deficiencies in care for vulnerable older people. Those with
diabetes are particularly at risk because of the lack of knowledge
and training amongst carers, a problem acknowledged both in the
UK and globally. To identify the cause of the problem, the Institute
of Diabetes for Older People (IDOP) explored the concerns of
healthcare professionals of all grades, working in care homes and
in the domiciliary sector.
168
Methods: We recruited participants through the quality

monitoring team of NHS Northamptonshire. A session was
devoted to each of the following professional groups: managers/
senior nurses, healthcare assistants and carers from local nursing
and residential homes and from domiciliary care. Using a focus
group format, 13 questionsbased on the 2010 Diabetes UK
document Good clinical practice guidelines for care home
residents with diabeteswere discussed by each group. Comments
were documented as bullet points on flip charts, confirmed
formally by the group.
Results: The main concerns identified were as follows: poor
communication with hospital, primary care and mental health
services;lack of training among staff in diabetes and dementia;
training needs were not met by available e-learning modules;
inadequate support provided by the local diabetes specialist team.
Conclusions: The Northampton focus groups have identified
several issues of national importance. These are paralleled in the
national audit of diabetes care in care homes carried out by IDOP
and the Association of British Clinical Diabetologists on behalf of a
national stakeholder group.
P468
Using the UK Diabetes Attitudes, Wishes and
Needs (DAWN2TM) data to help guide
consultations between healthcare
professionals and people with diabetes
R Brice1 and RIG Holt2
1
Whitstable Health Centre, NHS Canterbury and Coastal, Coastal Clinical
Commissioning Group, Canterbury, UK, 2Faculty of Medicine, University of
Southampton, Southampton, UK
Aims: To use data from the DAWN2TM study to help guide

consultations between healthcare professionals (HCPs) and people
with diabetes (PWD).
(Type 1 diabetes, 81; Type 2 diabetes, 419) and 281 HCPs (GPs,
120; specialists, 81; nurses/dietitians, 80) in the UK completed the
DAWN2 survey. These data are unweighted and therefore cannot
be extrapolated across the UK.
Results: PWD most commonly worried about hypoglycaemia
(58% Type 1 diabetes; 32% Type 2 diabetes), nocturnal
hypoglycaemia (60%; 26%), weight (60%; 63%) and dietary
restrictions (47%; 56%). HCPs commonly stated that the
following from PWD would aid consultations: telling them how
they might best support them in managing their diabetes (83%
91%) and preparing questions in advance of the consultation
(83%89%). Fewer than half of PWD reported that their
healthcare team listened to how they would like to do things
(46%; 40%); an even smaller proportion were helped to set goals
to improve management of their diabetes (28%29%) and make
plans to achieve those care goals (27%31%).
Conclusions: HCPs might better support PWD if they were
informed of or directly enquired about their most common
worries. In HCPs view, consultations might be more useful if
PWD prepared questions in advance and stated how they could
best be supported. As care planning and individualised goals
comprise NICE quality standards for diabetes, it is concerning that
so few PWD in the DAWN2 study recalled being set goals/plans to
achieve goals; this needs to be given greater weight in
consultations.
2014 The Authors.

P469
Reduction in prescribing costs for diabetes:
the 3 year results of First Diabetes, an
integrated diabetes care organisation in
Derby
RD Rea1,2, K Farrell1,3 and S Fitton4
1
First Diabetes, Derby, UK, 2Diabetes and Endocrinology, Derby Hospitals NHS
Foundation Trust, Derby, UK, 3Willington Surgery, Derby, UK, 4Medicines
Management, Southern Derbyshire Clinical Commissioning Group, Derby, UK
Aims: First Diabetes was set up in 2009 as an integrated diabetes

care organisation. It is a partnership between a local acute hospital
and five GP practices. The service is based on an integrated IT
system, joint clinical governance, care planning, a year of care
budget and ongoing clinical partnerships. The aim of this study
was to determine the effect of an integrated care partnership on the
diabetes drug expenditure of the five GP practices compared with
national expenditure.
Methods: All diabetes prescriptions between 2009 and 2012 were
collated via the CCG prescribing data website. The changes in
prescribing for total diabetes expenditure were expressed as the
cost per prescribing unit (PU). The total insulin items and cost
were expressed per ASTRO-PU and the oral diabetic drugs were
expressed as items and cost per cost-based STAR-PU.
Results: The total diabetes prescribing cost per standard PU fell
from 1,855 to 1,760 compared witha national increase from
1,975 to 2,389. There was a fall in insulin costs from 119 to
101 compared with a national rise from 110 to 112. Oral
antidiabetic expenditure remained flat (192 to 199) compared
with a large national rise (334 to 495).
Conclusions: Prescribing costs for diabetes are rising
nationally. In a 3 year outcome study of an integrated diabetes
partnership, total diabetes prescribing costs, cost of insulin and
cost of oral antidiabetic drugs fell resulting in significant financial
savings. The quality of care in First Diabetes, as measured by
HbA1c, reduction in acute hospital admissions and patient
experience was also improved.
P470
Interactions between people with diabetes
and healthcare professionals in the UK:
results from the Diabetes Attitudes, Wishes
and Needs (DAWN2TM) study
N Kanumilli1, R Brice2, G Hall3 and RIG Holt4
1
South Manchester Clinical Commissioning Group, Northenden Group
Practice, Northenden, UK, 2NHS Canterbury and Coastal Clinical
Commissioning Group, Whitstable Health Centre, Whitstable, UK, 4Faculty of
Medicine, University of Southampton, Southampton, UK
Aims: This descriptive sub-analysis of the DAWN2TM study

examines the interactions between people with diabetes (PWD)
and healthcare professionals (HCPs) in the UK.
(Type 1 diabetes, 81; Type 2 diabetes, 419) and 281 HCPs (GPs,
120; specialists, 81; nurses/dietitians, 40) in the UK completed the
DAWN2 survey. These data are unweighted and therefore cannot
be extrapolated across the UK.
Results: Between 56% and 73% of PWD said they try/would like
to make improvements in their diet, physical activities and
maintaining a healthy weight; overall, HCPs reported that 65%
75% of the patients they treat currently need improvement in those
areas. Overall, 23%24% PWD said they have always/most of
the time been asked by an HCP about problems with medications
2014 The Authors.

DIABETICMedicine
and their effects; however, 65%85% of HCPs reported that they

always/most of the time ask their patients of such problems. Less
than half of PWD (40%46%) reported that their healthcare team
listened to how they would like to do things; however, 72%93%
of HCPs reported that they have done so always/most of the
time. The resource PWD most wanted more of over the next 12
months was ongoing advice outside regular office visits (30%
38%), while HCPs most wanted to offer a diabetes telephone help
hotline (46%56%).
Conclusions: There were a number of areas where the perceptions
and opinions of PWD and HCPs differed regarding management
and needs; to ensure optimal outcomes, it is important that
consultations are collaborative and that expectations/goals are
aligned.
Acknowledgement: On behalf of the DAWN2 Study Group.
P471
A stepwise approach to releasing capacity in
diabetes outpatient clinics: right care, right
time, right place
SL Woodman, KE Fayers, J Murphy and C Atkinson
Foundation Trust, Lyndhurst, UK
Aims: As the burden of long-term conditions continues to rise,

new approaches to the delivery of traditional outpatient care will
be required. Since the West Hampshire Community Diabetes
Service was launched in 2010, numbers of diabetes patients have
risen by 20%. We have devised a stepped approach to the patient
journey. This audit assessed the likely numbers of patients who
would be suitable for group intervention at first encounter.
Methods: In all, 307 referrals were triaged by diabetes nurses at our
single point of access in July/August 2013. Each patient received a
telephone call from the diabetes specialist nurse (DSN). Patients were
clinically grouped around themed interventions.
Results: Sixty-three DSN clinics comprised 20% insulin titration,
30% general diabetes update, 12% complication management,
12% meter download, 26% other. Thirty-twodietitian clinics
comprised 60% carbohydrate awareness advice, 30%
carbohydrate counting and 10% other. DSNs and dietitians
concluded that 48% and 52% of respective referrals could be
managed in a group setting. Reasons for unsuitability for group
education included being elderly, having learning needs, having a
complex medical history, anxiety and patient choice. According to
our stepped pathway these patients would be offered individualised
decision making support (one to one) by members of the
multidisciplinary team, including consultant review as appropriate.
Conclusion: Group interventions are appropriate for many
patients. Peer support is a useful tool to help patient
engagement. It is likely that diabetes group interventions (outside
traditional Type 1 and Type 2 education) are under-used across the
NHS. A national stepped model could improve quality and
efficiency of care.
169
DIABETICMedicine
P472
Reduction in acute hospital admissions for
patients with diabetes: the 4 year results of
First Diabetes, an integrated diabetes care
organisation in Derby
B Eaglesfield1,2, R Butterfield3, K Farrell1,4 and RD Rea1,2
1
First Diabetes, Derby, UK, 2Department of Diabetes and Endocrinology,
Derby Hospitals NHS Foundation Trust, Derby, UK, 3Department of
Information, Derby Hospitals NHS Foundation Trust, Derby, UK, 4Willington
Surgery, Derby, UK
Aims: First Diabetes was set up in 2009 as an integrated diabetes

care organisation. It is a partnership between a local acute hospital
and five GP practices. The service is based on an integrated IT
system, joint clinical governance, care planning, a year of care
budget and ongoing clinical partnerships. The aim of this study
was to determine the effect of an integrated care partnership on the
admission rates of patients with a primary and secondary code of
diabetes.
Methods: All acute admissions for patients with diabetes between
2009 and 2012 were collated via HES data. Admissions were
divided into those with a primary code of diabetes and those with
diabetes as a secondary code. The admissions from First Diabetes
were compared with those from the rest of the county where there
is no integrated diabetes organisation.
Results: The total number of unplanned hospital admissions for
patients with a primary code of diabetes remained unchanged
between 2009 2010 and 2012 2013 (37 to 39). This was
consistent with the unchanged number of county admissions
(143 to 141). The number of unplanned admissions for patients
with a secondary code of diabetes fell by 21% in First Diabetes
practices (536 to 426) compared with a 9% fall in the county
(1865 to 1699).
Conclusion: There was a significant reduction in emergency
hospital admissions for patients with a secondary code of
diabetes in First Diabetes. This reduction was over twice as great
as that seen in the rest of the county which do not have an
integrated diabetes organisation.
P473
Review of prescribing patterns for two
clinical commissioning groups diabetes
population with an HbA1c greater than
64mmol/mol in primary care
RC Dallmeyer1, A Hodgkinson2, D Sennik3 and L Briant1
1
Trust, London, UK, 2Medicines Management, NHS Lambeth Clinical
Commissioning Group, London, UK, 3Medicines Management, NHS
Southwark Clinical Commissioning Group, London, UK
Aims: Glycaemic control of HbA1c less than 64mmol/mol

remains in the bottom national quartile for people with diabetes
in the boroughs of Lambeth and Southwark despite many years of
focus and support. The National Institute of Clinical Excellence
(NICE) diabetes pathway algorithm clearly identifies first, second
and third line therapies but local individual patient data show
antidiabetic medicines are started late and titrated slowly to
individual HbA1c targets. We tested this finding at a population
level across two CCGs.
Methods: We developed general practice clinical system searches
to identify the number of people with diabetes with an HbA1c
greater than 64mmol/mol that were on no medication, first line,
second or third line treatment options as defined by NICE. We
170
tested and refined the searches in several general practices and

supported their implementation at practice and clinical
commissioning group (CCG) levels working closely with local
practice system experts.
Results: We collected data from 82 (87%) practices and 23,945
people with diabetes in the boroughs of Lambeth and
Southwark. At 31 March 2013, 32% of people with diabetes
had an HbA1c more than 64mmol/mol. Of those not controlled,
4% were prescribed no medication, 13% were only prescribed
metformin, 3% were only prescribed a sulfonylurea, 23% were
only prescribed metformin and a sulfonylurea, 41% were
prescribed insulin and 16% were prescribed pioglitazone, a
gliptin or a glucagon-like peptide-1 analogue.
Conclusions: Improvements can be made at each stage of the
NICE prescribing algorithm. Optimising first and second line
therapy could increase glycaemic control of HbA1c less than
64mmol/mol by up to 9%.
P474
Use of a structured situationbackground
assessmentrecommendation (SBAR)
approach to clinic letter dictation to
improve communication between primary
and secondary care
J King1 and A Robinson1,2
1
Diabetes and Endocrinology, Royal Bolton Hospital, Bolton, UK, 2Bolton
Diabetes Centre, Royal Bolton Hospital, Bolton, UK
Aims: To compare general practitioner (GP) satisfaction with

clinic letters dictated from Bolton Diabetes Centre using a
five-question survey before and after implementation of a
structured SBAR approach to letter dictation.
Method: A total of 181 satisfaction surveys were sent out with
clinic letters from Bolton Diabetes Centre to primary care
providers in March 2012 and again in March 2013 after the
implementation of the SBAR technique to letter dictation. GPs
marked each letter out of 5 for five different areas of satisfaction on
both occasions. Results were analysed using an unpaired Students
t-test
Results: Significant improvements were made in GP satisfaction in
all areas assessed when training registrars were trained to use the
SBAR technique to dictate letters compared with satisfaction
before training. Mean overall letter satisfaction was 3.60 out of
5 [95% confidence interval (CI) 2.807 4.493] before the
introduction of the SBAR technique and 4.44 (95% CI
3.647 5.233) after the introduction of SBAR in letter dictation.
No significant difference was seen in the consultant group with
mean overall satisfaction 4.36 (CI 4.260 4.468) before the use of
SBAR and 4.35 (CI 4.244 4.452) after. A degree of improvement
was seen in the diabetic specialist nurses group with mean
satisfaction 4.20 (CI 4.112 4.280) prior to SBAR and 4.50 (CI
4.427 4.580) after its introduction.
Conclusion: The quality and clarity of clinic letters written by
registrars and doctors in training can be improved by training
them to use a structured SBAR approach to clinic letter dictation.
2014 The Authors.

P475
Non-elective readmissions to an acute
hospital in people with diabetes: causes and
the potential for avoidance: the
Wolverhampton Interface Care Knowledge
Empowered Diabetes (WICKED) Project
U Aziz, SMR Gillani, D Blundell, R Nithyananda and BM Singh
Wolverhampton, UK
Introduction: Managing people with diabetes is a health priority

worldwide. Cost benefit attempts at avoiding non-elective
admissions (NEA) have had some success. To develop an NEA
avoidance service, we audited multiple NEA in those with diabetes.
Method: All people with diabetes who had 3 NEA to our
hospital over 18 months were identified (n = 387); 97 patients
were randomly selected and retrospective data were collected.
Results: Of 97 patients, 50 were male; aged 69 16 years; 85
Type 2 diabetes; 47 on insulin. Of the whole group, according to
standardised criteria, 92% were justified although 28% were
deemed avoidable. Conditions contributing to admissions included
significant comorbidities (92% 2; in 70% amongst them 10% had
4 comorbidities); palliative care (9%); mental health disorders
(3%); dependency (30%); poor compliance (6%). Only 14%
(n = 14) were due to diabetes amongst whom 57% (n = 8) were
deemed avoidable. Thus 20 (out of 28) patients, being 70% of
avoidable admissions, were not diabetes related.
Conclusion: The majority of readmissions were not diabetes
related and comorbidities have a significant contribution.
WICKED aspires to be an integrated model of diabetes
healthcare, which is patient needs centric. Thus it is clear that
the provision of an admissions avoidance service to people with
diabetes cannot be glucose or diabetes centred. Clearly this
necessitates
multi-professional,
multidisciplinary,
multiorganisational working and vertical integration of services. But
does it not have further implications for specialist diabetes teams
providing acute care, emphasising acute/general medicine skills
within the consultant body and in our training programmes?
P476
The Royal Free Continuous Subcutaneous
Insulin Infusion (CSII) Service: a comparison
of patch pump vs traditional pump showing
different population demographics,
increased patch pump uptake in pregnancy
and equal efficacy at reducing HbA1c
N Gostelow1,2, L Williamson1,2, J Holland1, M Vanderpump1
and M Rosenthal1
1
Department of Diabetes and Endocrinology, Royal Free London NHS Trust,

London, UK, 2Department of Medicine, University College London, London, UK
Background: The Royal Free Diabetes Service offers the choice of

patch pump (PP) vs traditional CSII (TP). Unlike TP, PP is smaller
and attaches directly to the skin with an integrated cannula. There
are currently no studies comparing the efficacy and indications for
use of PP with TP.
Aims: To examine the differences in demographics and outcomes
between patients using PP and TP.
Methods: Data were collected retrospectively from patient notes,
clinic letters and databases.
Results: Fifty-two out of 188 patients chose PP; of those,
male 33%, female 67%, mean age 41.9, median 40.0, mean
2014 The Authors.

DIABETICMedicine
duration of diabetes 21.8. TP: male 47%, female 53%, mean

age 38.2, median 32.0, mean duration of diabetes 20.2.
Indications were as follows. Persistently elevated HbA1c:
PP = 59.1%, TP = 44.6%. Disabling hypoglycaemia PP = 40.1%,
TP = 37.4%. Hypoglycaemia unawareness PP = 9.1%, TP = 9.8%.
Improving glycaemic control in pregnancy: PP = 13.6%,
TP = 9.8%. Non-NICE approved indications were PP, 0.02%
(recorded as lifestyle), TP, 14.3%. Reducing HbA1c: all patients
with HbA1c > 8.5% mean pre-pump HbA1c/mean post-pump
9.86%/8.64%.
Mean
reduction SEM = 1.22% 0.124,
p < 0.005. There was no significant difference between PP and TP
in reducing HbA1c.
Conclusion: PP is as effective as TP in lowering HbA1c. More
women chose a PP. There were no differences in indication for
starting CSII between groups. PP was more often used in pregnancy
than TP supporting anecdotal evidence that it is easier to
commence and support users. As no clinical benefit of PP against
TP has been shown, further studies into treatment satisfaction or
quality of life are now required to guide clinical decision making.
P477
Specialist community diabetes services
make significant improvements to patient
clinical outcomes in Lambeth and
Southwark
H Ward1, J Sharpe2, M Chamley3 and J Collins1
1
Diabetes Modernisation Initiative, Guys and St Thomass NHS Foundation
Trust, London, UK, 2Southwark Community Diabetes Service, Guys and St
Thomass NHS Foundation Trust, London, UK, 3Lambeth Diabetes
Intermediate Care Team, Crowndale Medical Centre, London, UK
Aims: Specialist community diabetes clinics were established in

Lambeth and Southwark in 2012. A key aim of the clinics was to
improve the clinical outcomes of patients with diabetes in the
boroughs. Patients are seen by specialist multidisciplinary clinical
teams, including diabetes specialist nurses, consultant
diabetologists, dietitians and GPs with special interests in
diabetes. In addition, for those patients requiring it, mental
health interventions are available.
Methods: The Diabetes Modernisation Initiative collected data
on patient clinical outcomes from the two community services.
These data were based on all patients seen between April 2012 and
March 2013, and comprised 793 patients in Lambeth and 343 in
Southwark. The data were analysed to find the reductions achieved
by patients seen in community clinics, and the number of patients
achieving their individualised targets.
Results: Our analysis showed that the average reduction in
HbA1c for patients attending community clinics in Southwark
was 14mmol/mol (from first appointment to time of discharge) and
in Lambeth the average reduction was 11mmol/mol. Where
suitable, some patients were set targets for HbA1c reduction. In
Lambeth, of the 317 patients with an individualised target, 23%
achieved this. In Southwark, 34% of the 155 patients with an
individualised target achieved it.
Conclusions: The model of care patients receive in the Lambeth
and Southwark community diabetes clinics is leading to excellent
achievement of better diabetes control. Patients seen in clinics
are achieving large and unprecedented reductions in HbA1c,
suggesting the model of care is suitable for improving patient
clinical outcomes.
171
DIABETICMedicine
P478
Diabetes services in Lambeth and
Southwark have worked together to reduce
hospital diabetes outpatient caseloads by
ensuring patients are cared for in the most
appropriate setting
H Ward1, J Sharpe2, M Chamley3, C Gayle4, S Thomas5 and
J Collins1
1
Intermediate Care Team, Crowndale Medical Centre, London, UK, 4Diabetes
and Endocrinology, Kings College Hospital NHS Foundation Trust, London,
UK, 5Diabetes and Endocrinology, Guys and St Thomass NHS Foundation
Trust, London, UK
Aims: In August 2011, business cases were approved by NHS

Lambeth and NHS Southwark to expand the existing community
diabetes services offered in each borough. The new services were
tasked with wide remits a key aim was to ensure patients were
seen in the most appropriate care setting, particularly focusing on
reducing hospital outpatient caseloads.
Methods: To achieve this aim, the services set up specialist
community diabetes clinics, for patients requiring a level of
intervention unavailable in primary care and patients no longer
requiring intervention in an acute setting. In addition, the
community services worked to upskill primary care, e.g. by
providing virtual clinics, where patients on a practices diabetes
register are reviewed virtually by diabetes specialist nurses or
consultant diabetologists in conjunction with primary care staff,
and a future plan of management for each patient is put in place.
Results: Secondary care caseloads have reduced by 15% and 11%
respectively at the two acute trusts since 2011 2012. In addition,
514 fewer patients are being seen once or twice in hospital clinics,
which could suggest fewer less-complex patients are being
managed in secondary care. Primary care initiated referrals to
secondary care have reduced significantly in the same time period:
by 56% in Lambeth and by 20% in Southwark.
Conclusions: It is possible to significantly reduce hospital
outpatient caseloads. The community diabetes services in
Lambeth and Southwark have done so successfully, by providing
an alternative care setting for patients, both through community
diabetes clinics and by upskilling primary care.
P479
Creating improvements in foot health and
increasing foot assessments in Southwark
and Lambeth
M Ferdinand1, C Gayle2, J Doherty3 and J Collins1
1
Diabetes Modernisation Initiative, Guys and St Thomas Charity, London,
UK, 2Diabetes, Kings College Hospital, London, UK, 3General Practice,
Southwark, London, UK
Aims: To address the feedback from patients and primary care

clinicians regarding an unclear foot health pathway, to reduce
variation seen and increase foot assessments, and to develop a local
foot health referral pathway based best practice.
Method: A clinical project team was brought together to develop
the Lambeth and Southwark diabetes foot care pathway. Diabetic
foot activity was analysed across the pathway in order to inform
the work. The team also incorporated best practice (i.e. the
Putting Feet First pathway by Diabetes UK). The pathway was
launched for clinicians in autumn 2012. In addition, two training
172
events were aimed at GPs and practice and community nurses to

improve recognition of foot health problems, to launch the
pathway, and to provide practical support from podiatrists. The
events were attended by 70 primary care staff in total, including 23
GPs and 21 practice nurses.
Results: In 2012 2013 there was an increase in the number of
foot assessments performed/recorded of 1.9% in Lambeth and
2.9% in Southwark (increase of 2,545 patients). In 2011 2012 the
variation in the percentage of patients recorded with a foot check
on the registry ranged from 21.7% to 94.2% across practices; in
2012 2013 this variation narrowed to 44.8% 96.8%.
Conclusions: The knowledge sharing, clinical guidance and
upskilling of primary care, which occurred through learning
events, disseminating a local foot health pathway and analysing
diabetic foot activity, has created improvements in the number of
foot checks and reduced variation seen amongst practices.
P480
SCI-Diabetes: a population-based shared
electronic record for diabetes
SG Cunningham1, RG Morris2, C Flach2, WM Urquhart2,
A Taylor2, I Dickson3, J Walker4 and A Emslie-Smith5
1
Clinical Technology Centre, University of Dundee, Dundee, UK,

SCI-Diabetes Collaboration, NHS Tayside, Dundee, UK, 3Springwell Medical
Centre, NHS Lothian, Edinburgh, UK, 4St Johns Hospital, NHS Lothian,
Livingston, UK, 5Tayside Diabetes Managed Clinical Network, NHS Tayside,
Dundee, UK
Aims: Scottish Care Information: Diabetes Collaboration

(SCI-DC) is the NHS Scotland diabetes technology programme,
providing patient management solutions for healthcare teams
delivering diabetes services. The flagship system SCI-DC Network
was decommissioned in 2013 following 13 years of live use.
Capitalising on our extensive experience, we created its
replacement: SCI-Diabetes. We aimed to analyse initial uptake
and levels of user engagement following implementation.
Methods: A technology refresh and redesign was commissioned
in 2008 following the identification of requirements beyond the
capacity of the legacy system, including new specialty and clinical
support modules. Following development, health boards migrated
incrementally following acceptance testing and data validation
checks. SCI-Diabetes was fully implemented in all 14 NHS
Scotland health boards by May 2013. Audit trails between 1May
and 30 September 2013 were analysed for usage trends.
Results: SCI-Diabetes records data on 258,570 people diagnosed
with diabetes in Scotland (4.9% prevalence). 11.2% have Type 1
diabetes; 88.2% Type 2; 0.6% other types. During the 5 month
period analysed, 3,422 distinct users accessed the system (684.4/
month); 106,593 logins were successfully completed (21,318.6/
month; 31.1/user); 10,116,648 audit trail entries were created
following user activity (2,023,329.6/month; 2,956.4/user; 94.9/
login).
Conclusions: A comprehensive, centralised, shared computer
system for diabetes care now exists for use by healthcare
professionals of all specialties across NHS Scotland. The system
has high levels of user engagement and is a core part of day-to-day
patient management. SCI-Diabetes is a world-leading diabetes
information system with the potential to be extended to benefit
other parts of the UK and beyond.
2014 The Authors.

P481
Lambeth and Southwark community
diabetes services have worked to enable
primary care clinicians to better manage
patients with diabetes
J Collins1, M Chamley2, J Sharpe3 and H Ward1
1
Intermediate Care Team, Crowndale Medical Centre, London, UK
Aims: Business cases were approved by NHS Lambeth and NHS

Southwark to expand the existing community diabetes services
offered in each borough in 2012 2013. The new services were
tasked with wide remits a key aim was to improve knowledge
and management of diabetes in primary care.
Methods: Virtual clinics were provided to each practice (patients
are reviewed virtually by practice staff, diabetes specialist nurses
and consultant diabetologists) and joint clinics, when a member of
the community diabetes team will attend a diabetes clinic held in
the practice, with practice staff and patients in attendance.
Feedback on these clinics was gathered from primary care using
questionnaires. All practices were assigned a member of
the community team for advice and support, and staff are
available daily for telephone and email advice. In each borough
10 practices with the lowest performance in key clinical outcomes
were identified and offered intensive support. This support
included extra joint clinics, education sessions, reviews of
administrative procedures and more. In order to measure the
teams successes in supporting these practices, 2011 2012 Quality
and Outcomes Framework performances were compared with
2012 2013 results.
Results: The practices identified for intensive support in each
borough in 2012 2013 achieved a greater improvement in
biological outcome results than the borough average in
particular in Lambeth patient blood pressure control improved
by 9.3% in the identified practices and in Southwark detection
improved by 12.6%. Virtual clinics are valued by primary care
who reported that they are important learning tools.
Conclusions: The community teams were successful in supporting
and upskilling primary care to better manage diabetes in
2012 2013.
P482
A sustainable model of integrated
community care: 2 year outcome data
KE Fayers, S Woodman, C Atkinson and J Murphy
Aims: Our consultant-led community diabetes service was

launched in 2010. Care comprises clinical reviews (dietitian,
nurse,
consultant),
a
telephone
helpline,
education
programmes for patients (Type 1/Type 2) and healthcare
professionals. We aim to improve patient self-management and
deliver high quality clinical care closer to home through better
integration across both primary and secondary care.
Method: Patients referred to the service receive a diabetes nurse
triage telephone call within one working day. Patients are
appropriately directed to education or clinical appointments. We
are able to use remote technology to access our clinical database
(integrated with a local teaching hospital trust) from any
community location with 3G connectivity. Clinical records are
2014 The Authors.

DIABETICMedicine
uploaded to the Hampshire Health Record (a depositary for

primary care records). Retinal screening data are available via a
secure internet link. Access to Diasend enables us to download
bolus advisor meters, thereby offering intensive insulin adjustment
in the community setting. We reviewed 2 year data pre/post service
(2009 2010 to 2011 2012).
Results: Numbers of patients on the diabetes register rose from
19,975 to 21,848 (9.3% increase). New to follow-up ratios
reduced by 33%. The proportion of patients achieving Quality
and Outcomes Framework glycaemic targets improved
(HbA1c < 59mmol/mol = 51.3% 70%, HbA1c < 64 mmol/
mol = 76.4% 79%,
HbA1c < 75mmol/mol = 88.1% 90%).
Admission rates for hypoglycaemia fell by 8%. Type 1/Type 2
education feedback was very positive (97%/99% would
recommend the course to family or friends).
Conclusion: An integrated community diabetes service promotes
self-management and enhanced service delivery within primary
care, reduces new to follow-up ratios and provides a sustainable
model of care.
P483
The effectiveness of telephone consultation
in patients with diabetes
P Joseph, K Leong, M Banerjee and KS Leong
Diabetes Department, Wirral University NHS Foundation Trust, Bebington, UK
Background: Achieving optimal glycaemic control is a key

clinical objective in diabetes management. There is mixed
evidence for the effectiveness of the diabetes specialist nurse
(DSN) in improving glycaemic control. Telephone contact is one of
the main resources utilised by DSNs for follow-up but the
effectiveness is questionable.
Method: A retrospective review of the local diabetes database
was performed to review patients receiving DSN telephone calls
between September and October 2012. Data collection included
number of calls, number of weeks calls made, HbA1c, weight,
insulin dose and oral agents at baseline and 6 months following
contact.
Results: A total of 108 patients were contacted over the 2 month
period of which 56 (51.9%) were male and 52 (48.1%) were
female. Mean age of males was 63.1 14.2 years and of females
was 63.6 13.9years. Twenty-six patients (24.1%) had Type 1
diabetes and 82 patients (75.9%) had Type 2 diabetes. Patients had
a median of five calls (range 1 27 calls) over a median of 13 weeks
(range 1 26 weeks). Nine patients were on oral hypoglycaemic
agents (OHAs) only and 99 patients were on either insulin therapy
alone or a combination of insulin and OHAs. HbA1c improved
significantly from 78 (43 140) to 72 (41 132)mmol/mol at 6
months, p = 0.0001, for insulin patients; however, there was no
significant improvement in HbA1c, p = 0.400, for non-insulin
patients.
Conclusion: DSN telephone support for insulin patients
approximately every 18 days over 13 weeks significantly
improved HbA1c at 6 months. Longer follow-up is required to
determine if this reduction is sustained.
173
DIABETICMedicine
P484
Impact of UK guidelines on clinical
prescribing in patients with Type 2 diabetes:
a comparative effectiveness analysis of
liraglutide vs sitagliptin in the UK
P McEwan1, ML Evans2, ME Nyeland3, R Skovgaard3,
A Richards3, EQ Bergan3, D Demuth4, L Garcia Alvarez4,
A Muthutantri4 and UJ Ploug3
1
Swansea Centre for Health Economics, Swansea University, Cardiff, UK,

General Medicine, Diabetes and Endocrinology, Llandough Hospital,
Penarth, UK, 3Market Access, Novo Nordisk A/S, Sborg, Denmark,
4
Real-World Evidence Solutions and Health Economics and Outcomes
Research, IMS Health, London, UK
2
Aim: This study aimed to explore which treatment targets for

patients with Type 2 diabetes might be influencing clinical
prescribing in the UK, where a number of guidelines are
recommended. Comparisons were made between incretin-based
therapies liraglutide and sitagliptin and are reported.
Methods: The Clinical Practice Research Datalink (CPRD) was
used to conduct a retrospective database analysis. Patients (18
years) diagnosed with Type 2 diabetes and prescribed liraglutide or
sitagliptin between July 2009 and July 2012 were included.
Outcomes included proportions of patients achieving HbA1c
treatment targets of 7.5% [Quality and Outcomes Framework
(QOF) guidelines], <7% (American Diabetes Association
guidelines) and <6.5% (NICE guidelines) after 6 months of
therapy.
Results: A total of 287 liraglutide and 2,781 sitagliptin patients
were identified at index, with a mean (SD) age of 55.7 (10.6) and
62.0 (11.0) years and mean (SD) HbA1c of 8.8% (1.9) and 8.6%
(1.5), respectively. Twice as many liraglutide patients (15.7%)
reached HbA1c below 6.5% vs sitagliptin patients (8.2%). A
higher proportion of liraglutide patients achieved HbA1c 7.5%
and <7% (46.3% and 29.3%, respectively) in comparison with
sitagliptin patients (45.1% and 22.8%, respectively). When
controlling for confounding (e.g. age, gender, baseline HbA1c
and body mass index), liraglutide patients were significantly more
likely to achieve HbA1c targets (7.5%, odds ratio 1.67; <7%,
odds ratio 2.08; <6.5%, odds ratio 2.36; p < 0.05 in all cases) than
sitagliptin patients after 6 months of therapy.
Conclusions: Liraglutide offers improved glycaemic control over
sitagliptin at all treatment targets utilised in UK clinical practice.
The impact of QOF targets on therapy choice requires further
evaluation.
Southwark and only 51% of their diabetes population received

all nine care processes.
Methods: We developed general practice clinical system searches
to identify the number of people with diabetes that had received all
of the nine care processes as defined by NICE and tested the
correlation with glycaemic control of HbA1c less than 64mmol/
mol.
Results: We collected data from 82 (87%) practices and 23,945
people with diabetes in Lambeth and Southwark. At 31 March
2013, 7,714 (32%) of people with diabetes had received all nine
care processes in the previous 12 months. Five out of nine of the
care processes are being received by more than 80% of the diabetes
population. The care processes received by fewer people with
diabetes were smoking status 64%, urinary albumin 66% and feet
checks 68%. There was no correlation between population
completion of the nine care processes and HbA1c less than
64mmol/mol.
Conclusions: Improvements can be made in the percentage of
people with diabetes for each of the nine care processes with focus
on recording and reach of smoking status, urinary albumin and
foot checks having the greatest impact. Completion of the nine
care processes does not appear to impact the diabetes population
glycaemic control.
P486
Comparative achievements in
treatment-related Quality and Outcomes
Framework targets for Type 2 diabetes
patients receiving dual therapy with either
a sodium glucose co-transporter-2 inhibitor,
a dipeptidyl peptidase-4 inhibitor or a
sulfonylurea in dual therapy
1
Beerse, Belgium
Aims: The National Institute of Clinical Excellence (NICE)

recommends that people with diabetes should receive nine key
tests annually: weight, blood pressure, smoking status, blood
glucose, urinary albumin, serum creatinine, cholesterol and eye and
feet checks. The National Diabetes Audit was completed by 46
practices in 2010 2011 in the boroughs of Lambeth and
Objectives: The UK NHS treatment targets in Type 2 diabetes are

defined through the Quality and Outcomes Framework. This
analysis compares achievement of glycaemic control, blood
pressure (BP) and total cholesterol (TC) with three agents in
combination with metformin: the sodium glucose co-transporter-2
(SGLT-2) inhibitor canagliflozin, the dipeptidyl peptidase-4
(DPP-4) inhibitor sitagliptin and the sulfonylurea (SU)
glimepiride in dual therapy.
patient level data from two large randomised head-to-head trials
(modified intent-to-treat populations, n = 2,733), a logistic
regression model, including baseline patient-covariates, was fitted
to compare the number of patients reaching an HbA1c target
of <7.5%, BP of 140/80mm Hg and TC of 5mmol/l after
treatment for 52 weeks with canagliflozin 100mg or 300mg once
daily, sitagliptin 100mg once daily, or glimepiride titrated from
1mg to 6 8mg per day.
Results: For HbA1c <7.5%, 80.5%, 75.3%, 78.2% and 75.1%
of patients achieved the goal with canagliflozin 300mg, 100mg,
sitagliptin and glimepiride respectively (baseline HbA1c 7.9%).
The percentage of patients achieving the goal of 140/80mm Hg
was 69.9%, 67.6%, 62.8% and 54.7% for canagliflozin 300mg,
canagliflozin 100mg, sitagliptin and glimepiride respectively
(baseline systolic BP 129.0mmHg; dyastolic BP 78.4mmHg). TC
5mmol/l was achieved by 20.4%, 25.1%, 26.6% and 31.1% for
174
2014 The Authors.

P485
Are the nine care processes for diabetes
being received by people with diabetes in
the boroughs of Lambeth and Southwark
and does this impact on glycaemic control at
a population level?
RC Dallmeyer, L Briant, A Manya and J Collins
Trust, London, UK
canagliflozin 300mg, 100mg, sitagliptin and glimepiride

respectively (baseline TC0.268mmol/dl).
Conclusions: More patients achieved their HbA1c and BP targets
with canagliflozin 300mg, than with canagliflozin 100mg,
sitagliptin 100mg or glimepiride. However, fewer patients
achieved their TC targets when treated with canagliflozin than
with sitagliptin or glimepiride.
P487
Tracking achievement in diabetes-related
local policy targets supporting the Outcome
Framework using the diabetes management
information tool (DMIT)
G Thompson1, L Stein2, S Williamson3 and I Girod1
1
Outcomes Research, Janssen-Cilag Ltd, High Wycombe, UK, 2Medical
Affairs, Janssen-Cilag Ltd, High Wycombe, UK, 3External Affairs,
Janssen-Cilag Ltd, High Wycombe, UK
Objectives: The Outcome Framework drives improvements in

quality and outcome measurements throughout the NHS by
encouraging a change in culture and behaviour in order to
improve efficiencies. This key policy is supported by various
local policies and incentives which set targets to area teams (ATs),
clinical commissioning groups (CCGs), trusts and/or general
practices (GPs).
Methods: An analysis of published English NHS data [Hospital
Episode Statistics database, Quality and Outcomes Framework
(QOF) and GP Practice Prescribing Chemical-level Data] has been
conducted using DMIT for the years 2010 2013. NHS targets
analysed at trust and CCG level were percentage of emergency
admissions due to hypoglycaemia or diabetic ketoacidosis (DKA).
Achievement of the nine care processes related QOF indicators was
also analysed at GP and CCG level. Examples of Leicestershire and
Lincolnshire AT and East Leicestershire and Rutland CCG were
used.
Results: The proportion of emergency admissions as part of the
total number of admissions associated with diabetes has not
changed between 2011 2012 and 2012 2013 (above 80% for
both AT and CCG), although the proportion of emergency
admissions related to DKA and hypoglycaemia amongst all
emergency admissions increased at both AT and CCG level.
Achievement of QOF indicators related to the nine care processes
varies greatly between GP practices, and more specifically for
HbA1c and blood pressure.
Conclusions: The DMIT can be used to track achievement of
some of the main NHS targets related to the management of
diabetes, and allows commissioners to share best practice and
provide targeted support to low performing trusts and practices.
P488
Impact of the Integrated Care Pilot on
HbA1c, cholesterol and systolic blood
pressure levels in patients with diabetes
DIABETICMedicine
in ICP practices. Trends in HbA1c, cholesterol and SBP levels over

time were measured after controlling for a wide range of patient
and clinical factors.
Methods: The period analysed extended from 1 April 2004 until
30 April 2013. A multilevel model with patient as a random effects
level was used to allow for multiple tests and codes clustered in
individual patients. A general estimating equation function was
used as we have repeated observations over time.
Results: ICP patients (n = 8,494) were exposed to the ICP, on
average, for 5.9 months. Ethnicity, deprivation, test year and Type
2 diabetes were risk factors for a higher HbA1c. We also found
that HbA1c decreased significantly, by 1.13mmol/mol per year of
exposure to ICP. Moreover, cholesterol increased significantly, by
0.04mmol/l, and SBP by 0.3mm Hg per year of exposure to ICP.
Cholesterol increase was associated with risk factors such as age,
gender, Type 2 diabetes, multi-morbidity, duration in ICP and test
year. We identified risk factors for higher SBP age, Mixed and
Black/Black Caribbean ethnicity, deprivation and multi-morbidity.
Conclusion: After controlling for most important confounders,
there is now evidence of a significant fall in HbA1c values in ICP
practices.
P489
Teleconsultation is an effective method of
supporting diabetes care in a remote and
rural setting
S Philip1, N Milne2, C Page2, C McCallum2, A Kelday2,
T Gilmour2, J Onubi3 and M Safdar3
1
NHS Grampian, Aberdeen, UK, 2NHS Orkney, Kirkwall, UK, 3University of
Aberdeen, Aberdeen, UK
Introduction: Providing specialist support for people with

long-term conditions in remote and rural areas can be
challenging. A teleconsultation service was set up for delivering
specialist diabetes care for the Orkney Islands in August 2008.
There are currently 1,035 people on the diabetes register and 13
general practice surgeries are the main providers for diabetes care.
We conducted an audit of the patients that were supported by
the teleclinic service between 2008 and 2013.
Methods: Data were collected from the clinic administration
system and the regional diabetes register.
Results: Ninety-two people with diabetes (54 Type 1, 30 Type 2,
8 gestational) were reviewed at the teleclinic, aged between 15 and
84 years. The number of patients reviewed has increased from 32
patients in 2008 to 56 patients (118 appointments offered) in
2012. In addition a further 43 patients were discussed via
multidisciplinary meetings using video-conferencing with GPs in
the isles. The main reasons for referral included poor glycaemic
control, hypoglycaemia, adolescents, gestational diabetes
and insulin pump therapy. This model of care has a high degree
of acceptability by both patients and healthcare professionals.
Conclusion: Our experience over the last 5 years suggests that
teleconsultation is an effective method of supporting diabetes care
in remote areas.
S Adomaviciute, H Watt, M Soljak, J Car and A Majeed

Department of Primary Care and Public Health, Imperial College London,
London, UK
Aims: The Integrated Care Pilot (ICP) is an innovative programme

designed to improve the coordination of care for people with
diabetes and people over 75 years in Inner North West London
(INWL). We compared HbA1c, cholesterol and systolic blood
pressure (SBP) levels between ICP patients with diabetes and a
comparator group of patients with diabetes but without care plans
2014 The Authors.

175
DIABETICMedicine
P490
Understanding equity of management of
diabetes using primary care data in an inner
city borough
H Dodhia1, S Thomas2 and M Chamley2
1
Lambeth and Southwark Public Health, Southwark Council, London, UK,

Diabetes and Endocrinology, Guys and St Thomas Foundation Trust,
London, UK
2
Objectives: To use anonymised primary care patient level data in

an inner London borough to audit against standards of diabetes
management for glycaemic and cardiovascular control regression
models to compare compliance to standards between different
patient characteristics.
Methods: Lambeth DataNet data, uploaded on 31 March
2012, were obtained from 49/50 practice diabetes registers. The
following variables were gathered: patient ID, age, sex, ethnicity,
HbA1c level, body mass index (BMI), smoking status, practice
ID, locality, comorbidity, date and value of last blood pressure
measurement if within 9 months (of March 2012), index of
multiple deprivation score. Diabetes management was analysed
against guideline standards for (i) diabetes checks; (ii) diabetes
values for glycaemic and cardiovascular control. Group differences
in these were reviewed by age, sex, ethnicity, index of multiple
deprivation score, duration of diabetes and risk indicators
(smoking, BMI, cholesterol) using regression modelling.
Results: There were 12,880 people with diabetes on the
registers. Over 80% lived in the worst two quintiles for
deprivation. A current record for record for glycaemic control
was achieved for 87% with Type 1 and 94% with Type 2 diabetes.
Glycaemic control (HbA1c between 48 and 58mmol/mol) was
achieved for 40% with Type 1 and 60% with Type 2
diabetes. Factors associated with poor glycaemic control were
age <65 years and longer duration of diabetes as well as poor
cardiovascular control.
Conclusions: Primary care data can be used to provide an equity
audit profile to inform targeted strategies for improving quality of
care.
P491
Improving management of patients with
steroid related hyperglycaemia by
introduction of a management protocol:
results of a prospective audit
EJ McCracken, IR Wallace, T Allen, C Megaw, C Norcross and
AG Nugent
Diabetes Education Unit, Belfast City Hospital, Belfast, UK
Aim: Steroids are commonly used in the management of many

medical conditions, with hyperglycaemia a known side effect.
Recognition and management is suboptimal. We propose a
treatment protocol for steroid induced and exacerbated
hyperglycaemia, and present results of an audit following
introduction of this protocol.
Methods: We recorded baseline renal and liver function,
indication for steroids, type and duration of steroid treatment for
17 patients referred to the diabetes team with either steroid
induced hyperglycaemia or worsening of pre-existing diabetes
control. Hyperglycaemia was managed as per the proposed
algorithm, first using gliclazide and then insulin depending on
capillary blood glucose readings.
176
Results: Three of 17 patients were managed with dietary input

alone. 7/17 commenced gliclazide, 4/17 insulinand 3/17 required
both gliclazide and insulin. Of the 10 patients prescribed gliclazide,
five different dosing regimens were identified, from 40mg once
daily to 160mg twice daily. Two cases required maximal dose
gliclazide without insulin. A further patients gliclazide was
discontinued because of abnormal liver function tests due
to progression of an underlying medical condition. Blood glucose
control remained variable between patients, but all
symptomatically improved. Blood glucose readings ranged from
5 to 17mmol/l.
Conclusion: We recommend implementation of a protocol for
management of steroid induced and exacerbated hyperglycaemia.
Appropriate treatment must be individualised, aiming to achieve
symptomatic control rather than tight glycaemic control, factoring
in the underlying disease process and the risks and benefits of
medications used to control glycaemia. While some patients will
require insulin, this is not necessarily suitable for all.
P492
The use of electronic prescribing to reduce
insulin prescribing errors: audit on
fast-acting insulin prescribed at bedtime at
Kings College Hospital
DCD Hope1, V Austin2, B Fidler3 and OG Mustafa1
1
Kings Insulin Safety Group; Department of Diabetes, Kings College
Hospital, London, UK, 2Kings Insulin Safety Group; Pharmacy, Kings College
Hospital, London, UK, 3Kings Insulin Safety Group; Information Technology,
Kings College Hospital, London, UK
Introduction: Nocturnal hypoglycaemia is hazardous as patients

may be unaware of the warning symptoms. Electronic Prescribing
and Medications Administration (EPMA) went live in our hospital
in November 2008 incorporating an insulin order-set in 2009 to
streamline prescribing and reduce insulin prescribing errors.
Aims: To calculate the overall number of bedtime fast-acting
insulin prescriptions, the indications for these prescriptions and the
number of associated hypoglycaemic episodes.
Methods: A spreadsheet was generated containing bedtime
fast-acting insulin prescriptions (November 2009 to November
2012). Electronic patient records were reviewed retrospectively to
determine the indications for each prescription and the number of
associated hypoglycaemic episodes.
Results: Over 3 years there were 4,420 fast-acting insulin
prescriptions using the order-set, 128 of these were bedtime
fast-acting insulin prescriptions. 57 (45%) of bedtime prescriptions
were indicated (e.g. for enteral/parenteral feeding). There were 71
(55%) bedtime prescription errors (34% prescribed in addition to
the intended threetimes daily regimen, 4% carried on from the
medication history and 62% unidentified indications). NovoRapid
was the most frequently prescribed insulin (mean 13units; range
1 55units). Eleven hypoglycaemic episodes (15%) were associated
with prescription errors.
Summary: There is potential to reduce the number of bedtime
insulin prescription errors. The electronic prescribing order-set was
modified, removing the ability to prescribe bedtime fast-acting
insulin and incorporating options for common indications
identified in this audit. We intend to analyse the unidentified
indications insulin errors and anticipate that the changes described
will reduce the number of non-indicated bedtime fast-acting insulin
prescriptions and reduce night-time hypoglycaemia.
2014 The Authors.

Clinical care and other categories posters: therapies
DIABETICMedicine
Clinical care and other categories posters:

support services
P493
The impact of a dedicated specialist service
on clinical outcomes in patients with
lipodystrophy and/or severe insulin
resistance
CL Adams, CL Hames, C Jenkins-Liu, J Harris, RM Williams,
DB Dunger, S ORahilly, RK Semple, DB Savage and AJ Stears
Cambridge University Hospitals NHS Foundation Trust, Cambridge, UK
Aims: Lipodystrophy and other severe insulin resistance (SIR)

syndromes are very rare causes of diabetes frequently associated
with delayed diagnosis and severe metabolic disease. In order to
address these problems we have established a highly specialised
multidisciplinary national service, commissioned by NHS
England. Here we report an audit of the first 2 years of this
service on clinical outcomes.
Methods: A retrospective audit was performed of the 86 patients
(74 female, 12 male) with SIR and/or lipodystrophy seen in the SIR
clinic between April 2011 and March 2013. Thirty-two patients

have a confirmed genetic diagnosis (5 BSCL2, 2 AGPAT2, 9
LMNA, 11 PPARG, 5INSR). Thirteen patients are taking leptin,
two rhIGF1 and five Humulin R U500 insulin. Median age was
35.0 (range 3.0 72) years.
Results: There was a reduction in median (interquartile range)
HbA1c between the first clinic visit and at 6 12 months follow-up
from 71.0 (46.5 85.0)mmol/mol to 55.0 (47.0 73.5)mmol/mol,
p = 0.05, n = 25. There was also a reduction in median HbA1c in
patients who were above target (HbA1c > 53mmol/mol) at
baseline from 83.0 (71.3 90.5)mmol/mol to 70 (56.0 82.5)
mmol/mol, p = 0.02, n = 16. In patients with fasting
triglycerides above target (>1.7mmol/l) at baseline there was a
reduction in median (interquartile range) triglycerides at 6 12
months from 3.7(2.4 5.4)mmol/l to 2.8(1.4 4.5)mmol/l,
p = 0.06, n = 15. Patient satisfaction is high with a mean service
evaluation score of 8.9/10.
Conclusions: The provision of a dedicated national service for
patients with lipodystrophy and/or SIR, staffed by a team with
expertise in these rare disorders, has already had a positive impact
on metabolic control with high patient satisfaction.
Clinical care and other categories posters: therapies

P494
Patients with Type 2 diabetes treated with
analogue insulin require 40% more insulin
to achieve the same level of HbA1c than
those on non-analogue insulin
M Shepherd1,2, B Shields1, B Knight1,2, S Weinand-Barnett1,
CJ Fox3, TJ McDonald4 and AT Hattersley1
1
National Institute for Health Research, Exeter Clinical Research Facility,
University of Exeter Medical School, Exeter, UK, 2Research and Development,
Royal Devon and Exeter NHS Foundation Trust, Exeter, UK, 3Diabetes/
Endocrinology, Northampton General Hospital Trust, Northampton, UK, 4Clinical
Biochemistry, Royal Devon and Exeter NHS Foundation Trust, Exeter, UK
Refer to Oral A44
P495
Greater postprandial glucose control across all
three daily meals with liraglutide compared
with placebo when added to metformin and
rosiglitazone in patients with Type 2 diabetes:
posthoc analysis of Liraglutide Effect and
Action on Diabetes (LEAD)4
SC Bain1, M Shomali2, L Blonde3, CB Svendsen4, H Thomsen4
and B Zinman5
1
Diabetes, Singleton Hospital, Swansea, UK, 2MedStar Union Memorial
Hospital, Baltimore, Maryland, USA, 3Ochsner Medical Center, New Orleans,
Louisiana, USA, 4Liraglutide Diabetes Medical Affairs, Novo Nordisk A/S,
Sborg, Denmark, 5Leadership Sinai Centre for Diabetes, University of
Toronto, Toronto, Canada
2014 The Authors.

Aim: To examine the postprandial glucose (PPG) effects of liraglutide

add-on to metformin plus rosiglitazone across three daily meals.
Methods: Using 26week data from LEAD-4, comparing
liraglutide 1.2mg (n = 178), 1.8mg (n = 178) and placebo
(n = 177) each combined with metformin and rosiglitazone in
patients with Type 2 diabetes, this posthoc analysis determined the
proportion of patients below the American Diabetes Association
(ADA) PPG target (<10mmol/l) 90min after each individual meal
and after all three meals using seven-point self-measured blood
glucose (SMBG) profiles. The percentage of time spent below PPG
target at baseline and week 26 was compared using 15h
post-dosing SMBG.
Results: At week 26, more patients had PPG <10mmol/l with
liraglutide 1.2mg (60%) and 1.8mg (56%) vs placebo (26%)
across all three meals; the odds of achieving the PPG target was
greater for liraglutide 1.2mg (4.29) and 1.8mg (3.54) vs placebo
(p < 0.001 for both). Additionally, liraglutide-treated patients
were more likely to be below PPG target 90min after individual
meals (breakfast, lunch, dinner) vs placebo (p 0.001 for each
meal). At baseline, patients in all treatment groups were below
PPG target for <50% of the 15h timeframe, while at 26 weeks
patients receiving liraglutide (1.2mg and 1.8mg) spent significantly
greater time with PPG < 10mmol/l vs placebo (83% and 85% vs
61%, respectively; p < 0.001), with no significant difference
between liraglutide doses.
Conclusion: In patients with Type 2 diabetes receiving metformin
plus rosiglitazone, the odds of being below ADA PPG target across
all three meals were significantly greater after addition of
liraglutide vs placebo.
177
DIABETICMedicine
P496
The effect of glucagon-like peptide 1
agonist therapy on weight and HbA1c: how
important is dietetic input?
R Booth
Nutrition and Dietetics, University Hospital Southampton NHS Foundation
Trust, Southampton, UK
Aim: Glucagon-like peptide 1 (GLP-1) agonists are known to

improve glycated haemoglobin (HbA1c) and help achieve weight
loss. This study evaluates the effectiveness of dietetic input on
weight and HbA1c reduction in a group of subjects taking a GLP-1
analogue compared with a group of subjects taking a GLP-1
analogue that did not receive dietetic input.
Method: Data were collected retrospectively from notes and
laboratory results from patients known to the diabetes outpatient
team. Thirty patients taking the analogue for less than 12 months
were identified and included in the study. Group 1 (n = 15)
received dietetic input, had a start weight of 109.6kg 27.8kg
and started the treatment with an HbA1c of 83mmol/
mol 20.8mmol/mol. Group 2 (n = 15) did not receive dietetic
input; they had a start weight of 103kg 16.2kg and an HbA1c of
75.6mmol/mol 17.3mmol/mol. The outcome measures were
change in weight and HbA1c over a 12 month period.
Results: The change in weight and HbA1c for each group was
analysed and compared. In group 1 the weight significantly
reduced by 6.5kg 3.7kg compared with a weight reduction of
0.7kg 3.5kg in group 2 (p < 0.0001). The reduction in HbA1c
for group 1 was significantly reduced by 16.3mmol/
mol 22.2mmol/mol compared with 3.6mmol/mol 8.6mmol/
mol in group 2 (p < 0.05).
Conclusion: This study found that patients receiving dietetic
input had statistically significant weight and HbA1c reduction
whilst on GLP-1 treatment. The data highlight the importance of
dietetic input as part of a patients ongoing treatment for diabetes.
P497
An investigation into the efficacy of
glucagon-like peptide 1 (GLP-1) analogues in
obese patients with Type 1 diabetes
MK Shah, M Danjuma, T Saeed, and EG Khan
Diabetes Centre, Nobles Hospital, Douglas, Isle of Man, UK
Aim: This study determined the efficacy of liraglutide therapy on

weight and glycaemic control in patients with Type 1 diabetes.
This was not intended as a research project ab initio.
Methods: Twelve obese patients (body mass index > 28kg/mm2),
with poorly controlled Type 1 diabetes [mean HbA1c 71
12mmol/mol (8.64% 1.04%)], on conventional treatment with
insulin were given liraglutide (1.2mg od) after informed consent
(2009 2011). At the time of analysis 11 patients (seven men, four
women) had received liraglutide therapy for a mean duration of
11 2 months. Weight, HbA1c and 24h insulin requirements
(before and after treatment) were ascertained retrospectively.
A retrospective approval from the ethics committee has been
obtained.
Results: Baseline and follow-up characteristics of the study
population were classified with pair-t tests. The mean age of the
study population was 46 years (range 25 66), with a mean
duration of diabetes of 20.5 years (range 4.2 43.2 years). There
was a statistically significant difference in the mean weight
(5.05kg) between baseline and 11.09 2.01 months following
treatment with liraglutide (confidence interval 1.302 8.807,
178
p = 0.01). There was no significant difference in mean HbA1c or

insulin requirements,however.
Conclusion: Despite obvious limitations of this study, therapy
with GLP-1 analogues such as liraglutide appears to be well
tolerated and does result in significant weight loss in obese patients
with Type 1 diabetes. This, however, needs to be replicated in large
prospective studies.
P498
Liraglutide add-on to metformin improves
postprandial glucose control compared with
metformin monotherapy in patients with
Type 2 diabetes: posthoc analysis of the
Liraglutide Effect and Action in Diabetes
(LEAD)2 study
JW Stephens1, G Fulcher2, DR Matthews3, MA Nauck4,
CB Svendsen5, M Donsmark5 and A Garber6
1
Endocrinology, College of Medicine, Swansea University, Swansea, UK,

Diabetes, Endocrinology and Metabolism, Northern Clinical School, Sydney,
Australia, 3Oxford Centre for Diabetes, Endocrinology and Metabolism,
Churchill Hospital, Oxford, UK, 4Diabetes, Diabeteszentrum, Bad Lauterberg,
Germany, 5Liraglutide Diabetes, Medical Affairs, Novo Nordisk A/S, Sborg,
Denmark, 6Diabetes, Baylor College of Medicine, Houston, USA
2
Aim: To examine multi-meal postprandial glucose (PPG) effects of

liraglutide in patients with Type 2 diabetes.
Methods: Posthoc analyses of LEAD-2, comparing liraglutide
1.2mg (n = 241), liraglutide 1.8mg (n = 242), placebo (n = 122)
and glimepiride (n = 244) add-on to metformin, determined the
proportion of patients below American Diabetes Association
(ADA) PPG targets (<10mmol/l) 90min after individual meals,
and after three meals overall, at week 26 using seven-point
self-measured blood glucose (SMBG) profiles. Time spent below
PPG target at baseline and week 26 was compared using 15h
post-dosing SMBG.
Results: More patients had PPG < 10mmol/l across all three
meals with liraglutide 1.2mg (38%) and 1.8mg (37%) vs
glimepiride (27%) or placebo (15%); odds of remaining below
PPG target were higher with liraglutide 1.2mg and 1.8mg vs
placebo (odds ratio = 3.51, 3.37, respectively; p < 0.001) and
glimepiride (odds ratio = 1.68, p = 0.019, 1.61, p = 0.033,
respectively). Patients were likelier to be below target after
individual meals with liraglutide vs placebo (p < 0.001,
individual meals), but the difference vs glimepiride was not
significant at most meals. Time spent below PPG target during
15h post-dosing at week 26 increased to a greater extent from
baseline (~50%) in patients receiving liraglutide 1.2mg (71.6%),
liraglutide 1.8mg (73.4%) or glimepiride (69.9%) vs placebo
(44.0%, p < 0.001 for all). After 26 weeks, mean PPG increments
for the three meals overall were not significantly different
(liraglutide 1.2mg, 0.40mmol/l; liraglutide 1.8mg, 0.56mmol/l;
glimepiride, 0.44mmol/l; placebo, 0.44mmol/l).
Conclusion: The odds of being below ADA PPG target across
three daily meals were greater with liraglutide add-on to
metformin vs placebo or glimepiride.
2014 The Authors.

Clinical care and other categories posters: Type 1 diabetes
P499
Dipeptidyl peptidase-4 inhibitor
(DPP-4i) prescribing in persons with Type 2
diabetes with and without renal
impairment from the Clinical Practice
Research Datalink (CPRD)
DIABETICMedicine
subjects with diabetes and renal impairment are exposed to

unlicensed doses of these agents.
P500
AD Thomas1 and A Scowcroft2
Local audit on the use of glucagon-like

peptide (GLP-1) agonists in the treatment of
Type 2 diabetes: NICEly does it?
HMC Leong, S Williams and Y Nathan
Department of Diabetes, Lewisham Hospital, London, UK
Medical and Scientific Affairs, Boehringer Ingelheim Ltd, Bracknell, UK,

Market Access, Pricing and Outcomes Research, Boehringer Ingelheim Ltd,
Bracknell, UK
Introduction: A significant proportion of persons with Type 2

diabetes have concomitant renal impairment. This group are at
increased risk of hypoglycaemia. The DPP-4i class of agents are
licensed for treatment of Type 2 diabetes. Due to the
predominantly renal route of excretion, sitagliptin, saxagliptin
and vildagliptin all require dose adjustments in subjects with renal
impairment. As linagliptin is excreted via the enterohepatic system
it is exempt from dose adjustment.
Aims: The aim of this study was to investigate DPP-4i
prescribing in line with licence requirements.
Methods: All patients in the CPRD database with diagnosed Type
2 diabetes treated with DPP-4i during the course of 2012
were included. Furthermore all subjects with established and
documented diagnosis of renal impairment and/or a
documented estimated glomerular filtration rate investigation
result of less than 60ml/min/1.73m2 were identified.
Results: A total of 7,664 subjects with diabetes and treated with
DPP-4i were identified; 6,307 were treated with sitagliptin, 941
with saxagliptin, 257 with vildagliptin and 159 with linagliptin.
20% of sitagliptin patients, 25% of saxagliptin patients and 21%
of vildagliptin patients had been prescribed maximal doses of the
agents despite documented evidence of renal impairment.
Conclusion: Renal impairment is a common comorbidity in
persons with Type 2 diabetes. The DPP-4i class of agents are
commonly used for glucose control. Dose adjustment of all DPP-4i
with the exception of linagliptin in subjects with moderate and
severe renal impairment is a licence requirement. Our study
demonstrates that dose adjustment of DPP-4i does not always
happen in clinical practice and thus a significant proportion of
Aim: GLP-1 agonists can be used as triple or dual therapy in the

treatment of Type 2 diabetes. Our aim was to assess whether local
practice adheres to current NICE guidance (NCG 248, 203 and
87).
Methods: This was a retrospective audit of electronic patient
records of patients from January to December 2011. Patients
undergoing triple therapy should have an HbA1c 7.5% with a
body mass index (BMI) 35. Continuation beyond 6 months
requires a reduction in HbA1c of at least 1% and a weight loss of
at least 3% of initial body weight. Dual therapy is used in patients
where hypoglycaemic agents are contraindicated and should
achieve HbA1c reduction 1% by 6 months.
Results: Thirty-eight patients were started on GLP-1 agonists.
89.5% used liraglutide 1.2mg and the rest used exenatide. 90.9%
of patients had an HbA1c 7.5%. 45.5% of patients had an
HbA1c reduction of 1% and 81.8% achieved a weight reduction
of at least 3% of initial body weight beyond 6 months of triple
therapy. In dual therapy, 13.6% had recorded intolerance to
hypoglycaemic agents and 53% had at least 1% reduction in
HbA1c beyond 6 months. 39.4% of patients were on GLP-1
agonist and insulin therapy as an off-licence treatment.
Conclusions: Local practice does not strictly adhere to NICE
guidance, revealing off-licence use of GLP-1 agonists. However,
conjecture exists regarding their therapeutic niche. Case-by-case
individualised clinical judgement may be a suitable alternative. A
calculator identifying patients not fulfilling existing criteria will be
introduced and patients reassessed. We propose a re-audit in 8
months.

P501
Are antibodies to ion channels and other
neural antigens present in adults with
coexistent epilepsy and Type 1 diabetes?
A pilot study
I Idris1, B Lang2, T Moloney2, S Howell3 and P Maddison4
1
Division of Medical Sciences and Graduate Entry Medicine, University of
Nottingham, Derby, UK, 2Nuffield Department of Clinical NeuroSciences,
John Radcliffe Hospital, Oxford, UK, 3Clinical Neurology, Royal Hallamshire,
Sheffield, UK, 4Clinical Neurology, Queens Medical Centre, Nottingham, UK
Background: The high association of glutamic acid decarboxylase

(GAD) neuronal antibodies with Type 1 diabetesled us to
hypothesise that the co-occurrence of Type 1 diabetes and
epilepsy is likely to occur by more than just chance.
Methods: Twenty-five patients with Type 1 diabetes and coexistent
epilepsy (DM/EP) and 36 patients with Type 1 diabetes without
2014 The Authors.

epilepsy (DM) were recruited. All casenotes were reviewed to

confirm clinical details of diagnosis. Serum samples in all patients
were tested for voltage gated potassium channel antibodies, GAD
antibodies, leucine-rich glioma inactivated 1 protein (LGI-1)
antibodies (cell-based assay) and contactin-associated protein-2
(CASPR2) antibodies (cell-based assay). All assays were performed
in duplicate, and cell-based assays were scored by more than
one operator.
Results: All patients had unprovoked seizures, unrelated to
hypoglycaemic attacks. We found a non-significant greater
number of patients with positive antibody levels in the DM/EP
group compared with the DM group; DM/EP 13/25 (52%) had
positive neuronal antibodies, compared with 10/36 (27.7%) in
the DM group (p = 0.06, Fishers exact test). Antibodies were
most commonly detected in patients with complex partial
seizures, with or without secondary generalisation (8/12, 67%)
compared with generalised tonic clonic seizures, not otherwise
179
DIABETICMedicine
specified (2/6, 33%), and idiopathic generalised seizures (3/7,

43%).
Discussion: We found that a greater number of patients with
coexistent Type 1 diabetes and epilepsy had antibodies to neuronal
antigens compared with a control group of patients with solely
Type 1 diabetes. Most DM/EP patients harbouring neuronal
antibodies had complex partial seizures rather than idiopathic
generalised epilepsy.
P502
Atrial fibrillation, heart failure and
ischaemic heart disease in elderly people
M Srikantharajah1, MJ Chapman1 and K Nirantharakumar2,3
1
School of Clinical and Experimental Medicine, University of Birmingham,
Birmingham, UK, 2School of Health and Population Sciences, University of
Birmingham, Birmingham, UK, 3Institute of Digital Healthcare, University of
Warwick, Warwick, UK
Aim: People with Type 1 diabetes are living longer than previous
decades. However, many live with multiple comorbidities. Limited
data exist on cardiac comorbidities in the elderly with Type 1
diabetes. Here we look at the burden of cardiac comorbidities and
their association with risk factors in this population.
Methods: We did a cross-sectional analysis of a UK based general
practice database (the Health Improvement Network) for the year
2008. We used descriptive statistics (percentages) and logistic
regression [reporting odds ratio (OR) and 95% confidence
intervals (95%CI)] to study associations with risk factors such as
smoking, age, gender, hypertension, duration of diabetes and
glycaemic control (HbA1c).
Results: There were 997 elderly people with Type 1 diabetes
identified in the database. 42 (4.2%) had atrial fibrillation (AF),
263 (26.4%) had ischaemic heart disease (IHD) and 60 (6%) had
heart failure (HF). AF was significantly associated with age (OR
1.08, 95%CI 1.02 1.14) and HF (OR 3.32, 95%CI 1.29 8.54);
heart failure was significantly associated with male gender (OR
1.93, 95%CI 1.02 3.68) and IHD (OR 13.61, 95%CI
6.80 27.23); and ischaemic heart disease was significantly
associated with smoking (OR 1.52, 95%CI 1.09 2.13) and
duration of diabetes (OR1.02, 95%CI 1.01 1.03).
Conclusion: There is a high burden of cardiac morbidities in
elderly people with Type 1 diabetes. The benefit of optimising
therapy in an elderly population with Type 1 diabetes with the aim
of reducing adverse cardiac outcomes needs to be further studied
and recommendations implemented.
P503
Factors associated with glycaemic control in
adult patients with Type 1 diabetes treated
with insulin pump therapy
1,2, M T Maecki1,2 and T Klupa1,2
B Matejko1, J Skupien
1
Department of Metabolic Diseases, Jagiellonian University Medical College,
Krakow, Poland, 2University Hospital, Jagiellonian University, Krakow, Poland
Aims: Continuous subcutaneous insulin infusion (CSII) by insulin

pump seems to improve glycaemia and quality of life compared
with conventional insulin therapy in Type 1 diabetes. However,
while many people with Type 1 diabetes achieve excellent
glycaemic control, some others cannot reach recommended goals.
In a retrospective analysis, we searched for factors associated with
glycaemic control in people with Type 1 diabetes treated with
insulin pump therapy.
180
Methods: Data from 192 patients (133 women and 59 men)

treated with CSII at the Department of Metabolic Diseases,
University Hospital, Krakow, Poland, were analysed. Sources of
information included medical records, memory read-outs from
insulin pumps and data from glucose meters. Univariate,
multivariate linear and logistic regression analyses for the
association with HbA1c level were performed.
Results: The mean age of the subjects was 28.9 years, the mean
duration of Type 1 diabetes 14.6 years, mean body mass index
23.5kg/m2. The mean HbA1c level for the entire group was 7.4%.
In the multivariate linear regression analysis HbA1c correlated
with the mean number of daily blood glucose measurements
(b = 0.31, p = 0.0001), number of hypoglycaemic episodes per
100 blood glucose measurements (b = 0.24, p = 0.0015), age at
the examination (b = 0.20, p = 0.0084) and continuous glucose
monitoring system use (b = 0.18, p = 0.0195). Multivariate
logistic regression analysis for reaching a therapeutic target of
HbA1c < 7.0% showed that the independent predictors of
achieving this goal included the same four variables.
Conclusions: In a large clinical observation, we identified
patient-related and technological factors associated with
glycaemic control in adult CSII treated Type 1 diabetes subjects.
P504
Factors associated with insulin requirement
in adult patients with Type 1 diabetes
treated with personal insulin pumps
B Matejko1, MT Maecki1,2 and T Klupa1,2
1
Department of Metabolic Diseases, Jagiellonian University Medical College,
Krakow, Poland, 2University Hospital, Jagiellonian University, Krakow, Poland
Aims: Continuous subcutaneous insulin infusion (CSII) is thought

to be the most physiological method of insulin delivery in people
with Type 1 diabetes. However, it is not clear which variables play
a crucial role in insulin requirement in CSII treated patients. The
purpose of this analysis was to investigate retrospectively factors
affecting insulin dosing in Type 1 diabetes patients treated with
personal insulin pumps.
Methods: Data from 145 patients treated with CSII at
Jagiellonian University Hospital (south Poland region, 99 women
and 46 men) were analysed. Sources of information about the
patients included available medical records, memory read-outs
from insulin pumps, blood glucose meter data, the results of
biochemical analyses and surveys. The correlation analysis was
used to study relationships between various variables. Additionally
a multivariate regression analysis was performed for the
association with mean daily insulin requirement per kilogram of
body weight.
Results: The mean age of patients was 28.6 years, mean body
mass index (BMI) 23.1kg/m2 with an average duration of diabetes
13.4 years. Mean HbA1c level for the whole studied group was
7.3%. The use of a bolus calculator function tended to correlate
with lower insulin requirements (p = 0.0137). Yet, no significant
difference in CGMS group was observed (p = 0.3784). In
multivariate regression analysis, insulin dose per kilogram of
body weight was independently and significantly associated with
age (p = 0.0000), BMI (p = 0.0001), diabetes duration
(p = 0.0029), coefficient of variability (p = 0.0008) and with
percentage of basal insulin (p = 0.0041).
Summary: This study demonstrates that both patient-related and
2014 The Authors.

technology-related variables have a strong influence on insulin

dosing in CSIItreated individuals.
P505
Changes in age at diagnosis and prevalence
of positive family history in patients with
Type 1 diabetes over five decades
AP Brooks and JSW Chong
Diabetes Specialist Service, Hampshire Hospitals NHS Foundation Trust,
Winchester, UK
Aims: To use database information on 613 adults (313 males, 300

females) with Type 1 diabetes to examine trends in age at diagnosis
and prevalence of family history over the five decades 1961 2010.
Methods: A secondary care diabetes clinic serving patients in 12
postcodes ran from1983 to 2010. Each year patients routinely
attended once or twice and clinical data were collected. Simple
arithmetical analysis has been applied looking for changes in age at
diagnosis and family history of diabetes in first degree relatives.
Results: Age at diagnosis of Type 1 diabetes is not normally
distributed, but means increased progressively through decades
1961 1970 (18.5 years); 1971 1980 (18.3); 1981 1990 (23.4);
1991 2000 (28.2); to 2001 2010 (30.9). The percentages
diagnosed between ages 0.0 19.9 and 20.0 39.9 years old were,
by decade, 65.2 vs 27.3; 61.0 vs 30.5; 46.2 vs 38.0; 33.1 vs 45.9;
and 22.8 vs 54.4. Overall 14.8% patients were 40.0 59.9 years
old. A family history of diabetes in a first degree relative was found
in 40 males (12.8%) and 33 females (11.0%). Prevalence by
decade varied from 7.6% (1971 1980) to 17.1% (1981 1990)
when there were 18 males and nine females in 158 patients.
Conclusions: Mean age at diagnosis of Type 1 diabetes has
increased in this Anglo-Saxon population. Type 1 diabetes is not a
rare disease over 40 years of age (1 in 7 patients) which is
important in management of new onset diabetes. Up to 1 in 8
males and 1 in 9 females may have a first degree relative with
diabetes.
DIABETICMedicine
IDF (2005) criteria. The UKPDS risk engine was used to calculate
the cardiovascular risk.
Results: 100 patients with Type 1 and 100 with Type 2 diabetes
were reviewed. 70% of patients with Type 1 diabetes and 94% of
patients with Type 2 diabetes had MetS by all three criteria. This
shows that MetS is more common in Type 2 (p < 0.05) but also
highly prevalent in Type 1 diabetes. Comparing no criterion for
MetS vs three criteria, the cardiovascular risks were Type 2
diabetes 9.20% vs 19.02% (p = 0.05), Type 1 diabetes 3.23% vs
13.17% (p < 0.01).
Conclusions: MetS is highly prevalent in Type 1 diabetes and
therefore screening for MetS should be routine practice. Metformin
has been shown to be insulin sparing, improve glycaemic control
and promote weight loss. Our data support more proactive
screening for MetS and use of aggressive lifestyle and therapeutic
measures to optimise cardiovascular risk in these patients.
P507
A 6 year retrospective observational study:
autoantibodies to glutamic acid
decarboxylase (GAD) and insulinoma
antigen 2 (IA-2) in patients newly diagnosed
with adult diabetes presenting with
diabetic ketoacidosis
I Wallace1,2, S Chan1, M Naidu1, B Secret2, G Braatvedt1,2,3 and
M Khanolkar1,2
1
Department of Medicine, Auckland City Hospital, Auckland, New Zealand,

Auckland Diabetes Centre, Greenlane Clinical Centre, Auckland, New
Zealand, 3School of Medicine, University of Auckland, Auckland, New
Zealand
2
Aims: Metabolic syndrome (MetS) is an important risk factor for

cardiovascular disease in Type 2 diabetes. With limited data on
Type 1 diabetes, our aims are (1) to compare the prevalence of
MetS in patients with Type 1 vs Type 2 diabetes; and (2) to
investigate whether the cardiovascular risk of Type 1 diabetes is
higher if they have MetS.
Methods: Data from patients were collected in the diabetes
outpatient clinic. Every third patient with Type 2 and second
patient with Type 1 diabetes were selected for the study. The
exclusion criteria included patients whose body mass index could
not be calculated and/or they had other abdominal pathology.
MetS was identified using the WHO (1999), ATP III (2001) and
Aim: Diabetic ketoacidosis (DKA) is a common presentation of

diabetes, with most patients having Type 1diabetes; however, some
patients with Type 2diabetes also present with DKA. Antibodies to
GAD and IA-2 may be used to help define the type of diabetes. Our
aim was to evaluate the frequency of GAD and IA-2 antibody
positivity in newly diagnosed adult subjects with diabetes
following their initial presentation with DKA.
Methods: Adult subjects (15 years) admitted with a diagnosis of
DKA at Auckland City Hospital from January 2007 to December
2012 were identified by retrospective review of medical records.
For 54 subjects the initial presentation with diabetes was a DKA
episode. Due to lack of antibody data five subjects were excluded
from further analysis.
Results: 45% were of European descent, 41% were aged under 30
years at the time of presentation and 55% were male. 28% were
negative for both antibodies. Those of European descent were
more likely to have positive antibody titres compared with
non-Europeans (GAD 77% vs 59%, IA-2 54% vs 25%, either
antibody 81% vs 62%). All subjects of European descent testing
negative at presentation were aged over 40 years.
Conclusions: We highlight differences in frequency of GAD and
IA-2 positivity in newly diagnosed diabetes subjects presenting
with DKA. In our multi-ethnic catchment population a significant
proportion of patients presenting with DKA may have a ketosis
prone Type 2diabetes and this state may be more frequent than
previously thought.
2014 The Authors.

181
P506
Type 1 diabetes + metabolic syndrome: need
for a different management paradigm?
N Othonos1, V Patel1,2 and P Saravanan1,2
1
Academic Department of Diabetes and Endocrinology, George Eliot
Hospital, Nuneaton, UK, 2Division of Metabolic and Vascular Health, Warwick
Medical School, University of Warwick, Coventry, UK
DIABETICMedicine

P508
Hyperglycaemia in acute illness and the
subsequent risk of Type 2 diabetes
DA McAllister1, KA Hughes2, N Lone1 and NL Mills2
1
UK, 2Centre for Cardiovascular Science at the University of Edinburgh,
University of Edinburgh, Edinburgh, UK
Aims: Hyperglycaemia during a hospital admission is common in

people without known diabetes and is associated with adverse
outcomes. However, the subsequent risk of Type 2 diabetes is
unknown. We linked a national register of patients with diabetes
(SCI-DC), a national hospitalisation database and regional
biochemistry results databases in order to describe the
association between admission venous glucose and subsequent
3year risk of Type 2 diabetes.
Methods: Patients aged 40 years or older with an emergency
admission to hospital between 2004 and 2008 were included.
Prevalent diabetes and incident diabetes were identified via SCI-DC
and patients with prevalent diabetes (diagnosed on or before 30
days after the date of discharge from hospital) were excluded. The
predicted 3year risk of Type 2 diabetes by admission glucose, age
and sex was obtained from logistic regression models.
Results: 87,076 (70.5%) of patients had glucose measured on
admission. The 3year risk of Type 2 diabetes was <1% for a
glucose 5mmol/l increasing linearly to approximately 15% at
15mmol/l. The 3year risks at 7mmol/l and 11.1mmol/lwere 3%
and 10% respectively. One in 4 and 1 in 40 patients had glucose
levels above these respective cut-points. Risk of diabetes was also
associated with age and sex but not with specialty (medical vs
surgical). Only 49.1% of patients with admission glucose
11.1mmol/l had a subsequent glucose measure within 3 months
of discharge from hospital.
Conclusions: Information about glucose levels during a hospital
admission can be used to estimate risk of subsequent diabetes and
inform guidelines for follow-up of people with hyperglycaemia.
Acknowledgement: On behalf of the Scottish Diabetes Research
Network Epidemiology Group.
with metformin and a sulfonylurea: the sodium glucose

co-transporter-2 inhibitor (SGLT-2) canagliflozin and dipeptidyl
peptidase-4 inhibitor (DPP-4) sitagliptin.
patient level data from two large randomised trials (modified
intent-to-treat populations, n = 1,224), and a GEElogistic
regression analysis, including baseline patient covariates, were
performed comparing the number of patients reaching an HbA1c
target of <7.5%, BP of 140/80mm Hg and TC of 5mmol/l after
52weeks of treatment with canagliflozin 100mg or 300mg once
daily or sitagliptin 100mg once daily.
Results: For HbA1c < 7.5%, 74.0% of patients achieved the goal
with canagliflozin 300mg, 67.5% with canagliflozin 100mg and
50.8% on sitagliptin 100mg (baseline HbA1c, 8.1%). The
percentage of patients achieving the goal of 140/80mm Hg was
66.5%, 67.0% and 50.4% for canagliflozin 300mg, 100mg and
sitagliptin respectively (baseline systolic BP, 131mmHg; diastolic
BP, 79mmHg). TC 5mmol/l after 52week treatment was achieved
by 61.1% of patients with canagliflozin 300mg, 62.5% on
canagliflozin 100mg and 68.3% on sitagliptin (baseline TC,
4.71mg/dl).
Conclusions: More patients achieved their HbA1c and BP targets
with both doses of canagliflozin than with sitagliptin. However,
fewer patients achieved their TC targets when treated with
canagliflozin than with sitagliptin.
P510
Impact of linagliptin therapy in patients
with Type 2 diabetes and suboptimal
glycaemic control
AL Holland, S Gururaj and I Lawrence
Department of Diabetes and Endocrinology, Leicester Royal Infirmary,
Leicester, UK
Objectives: The UK NHS treatment targets for Type 2 diabetes

patients are defined through the Quality and Outcomes
Framework. Determining how anti-hyperglycaemic agents
(AHAs) can help achieve these targets is important to optimise
achievement of patients goals. This analysis compares
achievement of glycaemic control, blood pressure (BP) and total
cholesterol (TC) targets achieved with two AHAs in combination
Aim: To audit the impact of linagliptin 5mg daily upon glycaemic

control (HbA1c) and weight reduction (body mass index, BMI),
while monitoring renal function and sideeffect profile over a 6
month period in Type 2 diabetes.
Methods: We undertook a retrospective audit comparing
clinical parameters before and after linagliptin therapy (5mg
once daily, n = 44) in a specialist diabetes service. Patients with
Type 2 diabetes on at least one other antidiabetic medication,
but still having a baseline HbA1c of >7.5%, were included.
Clinical parameters studied included a reduction in HbA1c
of >0.5%, maintenance of renal function (eGFR) and BMI
reduction over 4 6months using electronic patient records and
biochemistry.
Results: Linagliptin improved glycaemic control (mean HbA1c
9.4% pre vs. 8.2% post) with 47% of patients achieving 0.5%
reduction. BMI (mean 34.7 pre vs. 32.5 post) reduced by 2.2kg/m2
and a mean increase in GFR of 2.35 ml min-1 was observed. Side
effects included oedema (n=3), gastrointestinal disturbance (n=3),
and musculoskeletal pain (n=2). Two patients ceased the
medication due to lack of efficacy, and 1 patient was lost to
follow up. Combined with insulin (n=27), Linagliptin resulted in a
mean HbA1c reduction of 1.2% and a BMI reduction of 2kg/m2.
Pre-existing renal impairment did not impact on efficacy.
Conclusions: Introducing linagliptin into a pre-existing regimen
of antidiabetic medications resulted in improved glycaemic control
and modest weight loss in Type 2 diabetes patients attending a
182
2014 The Authors.

P509
Comparative achievement of treatment
targets for patients with Type 2 diabetes
receiving triple therapy with the sodium
glucose co-transporter-2 inhibitor
canagliflozin and the dipeptidyl peptidase-4
inhibitor sitagliptin in combination with
metformin and a sulfonylurea
1
Beerse, Belgium
specialist clinic. There was no deterioration in renal function, and

indeed a modest improvement. No patients reported an increased
incidence of hypoglycaemia.
P511
Risk assessment tools for detecting those at
high risk of Type 2 diabetes: a systematic
review
SR Barber1, MJ Davies2, K Khunti2 and LJ Gray1
1
2
Department of Health Sciences, University of Leicester, Leicester, UK,

Diabetes Research Centre, University of Leicester, Leicester, UK
Aim: To describe and evaluate risk assessment tools which detect

those with undiagnosed impaired glucose regulation or at high risk
of diabetes using HbA1c (with or without Type 2 diabetes).
Methods: Tools were identified through a systematic search of
PubMed and EMBASE for articles which developed a risk tool for
those currently at high risk of diabetes. Data were extracted using a
standardised data extraction form.
2014 The Authors.

DIABETICMedicine
Results: Seventeen tools met the inclusion criteria from 11 papers.

Eleven tools were derived using logistic regression and six using
decision trees. Age, body mass index, family history of diabetes and
hypertension were the most frequently included variables. Those
developed using logistic regression included a median of 12
variables (range 419); the decision trees included a median of
three variables (range 25). The median internal C-statistic of those
developed using logistic regression was 0.70 (range 0.690.74)
compared with 0.72 (range 0.670.75) for those developed using
decision trees. The size of the datasets used and the number of
events per variable were acceptable in all the tools. Missing data
were not discussed for 41% of the tools, 10 (91%) of the logistic
scores categorised continuous variables and external validation
was carried out for only seven (39%).
Conclusions: Detecting those at high risk of Type 2 diabetes using
a non-invasive method is quick and cheap and gives the
opportunity for prevention. This review shows that scores
developed using decision trees have similar utility to scores
developed using logistic regression with fewer variables.
183

A Case of Haemochromatosis and Diabetes A Missed Opportunity

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

A Case of Haemochromatosis and Diabetes A Missed Opportunity

Încărcat de

Drepturi de autor:

Formate disponibile

Refer to Page 80- Poster 176

Clinical care and other categories posters: audit

Background: Disorders of sleep have been linked to obesity in the

awakenings. Central obesity was assessed using waist

Clinical care and other categories posters: audit

Aims: To investigate degrees of glucose intolerance amongst

2014 The Authors.

Central Manchester Foundation Trust, Manchester, UK, 2Partington Health

Objectives: Type 2 diabetes is usually a result of obesity and is

Abstracts of the Diabetes UK Professional Conference, 57 March 2014

Aims and objectives: NICE supports using insulin pump therapy

Aims: To assess whether the initiation and continuation of

liraglutide, a positive outcome in terms of weight loss and patients

Background: Intensive risk factor management is recommended

Objectives: The clinical benefits of influenza vaccination in older

2014 The Authors.

Clinical care and other categories posters: audit

vaccination against hospitalisation for acute cardiovascular and

Aim: To assess the change in the proportion of people with

2014 The Authors.

Conclusion: Over recent years there have been substantial

Background: Inpatients with diabetes is a common problem

Background: Successive National Diabetes Inpatient Audits

Abstracts of the Diabetes UK Professional Conference, 57 March 2014

identifies where an individual resides during inpatient care. Patients

Diabetes and Endocrinology, Royal Sussex County Hospital, Brighton, UK,

Aim: The Joint British Diabetes Societies (JBDS) guidelines

Clinical care and other categories posters: cardiovascular

1.28, p = 0.007] and CAD (OR 1.36, p = 0.001) after adjustment

Objective: Patients with chronic liver disease (CLD) are said to

2014 The Authors.

Clinical care and other categories posters: case reports

ethnicity, 121,535 (77.4%) had no CVD at baseline (2005 or

Introduction: Hyperglycaemia is associated with higher mortality

Clinical care and other categories posters: case reports

CL Hames, C Jenkins-Liu, J Harris, C Adams, RK Semple,

Department of Diabetes and Endocrinology, Cambridge University Hospitals

A Salam1, T Vakilgilani1, A Chambers2, A Hine2 and WM Kong1

Refer to Oral A55

Refer to Oral A53

A Juszczak1,2, G Thanabalasingham1,2, M Khonsari3,

APB Tasker1,2, V Alexander2, HL Mitchell2, A Jollands2,

Refer to Oral A56

Refer to Oral A54

2014 The Authors.

Abstracts of the Diabetes UK Professional Conference, 57 March 2014

Department of Diabetes, Kings College NHS Foundation Trust, London, UK,

Introduction: We describe two cases of non-alcoholic fatty liver

A 89-year-old woman was admitted with a 3 week history of

Aims: Acute inflammation and plantar foot deformities are

A 35-year-old lady, who presented with an infected, necrotic

2014 The Authors.

Clinical care and other categories posters: case reports

control [HbA1c 133mmol/mol (14.3%)] and underweight (weight

The High Secure Pychiatric Hospital treats mentally unwell people

2014 The Authors.

A 64-year-old lady with known Type 2 diabetes on NovoRapid

Diabetes, Kings College Hospital, London, UK, 2Neurosurgery, Kings

Introduction: Brain abscess is an unusual complication that may

Abstracts of the Diabetes UK Professional Conference, 57 March 2014

most brain abscesses and S. milleri (present in the oral cavity

haemochromatosis. He was started on 6 weekly testosterone

L Veryard and RB Poole