Anaesthesia 2014, 69, 540–549 doi:10.1111/anae.

12638

Original Article
The effect of a multicomponent multidisciplinary bundle of
interventions on sleep and delirium in medical and surgical
intensive care patients*
J. Patel,1 J. Baldwin,2 P. Bunting3 and S. Laha3

1 Medical Student, University of Manchester, Manchester, UK

2 Research Nurse, Critical Care Unit, 3 Consultant, Department of Anaesthesia and Critical Care Medicine, Royal
Preston Hospital, Lancashire Teaching Hospitals NHS Trust, Preston, UK

Summary
Sleep deprivation is common among intensive care patients and may be associated with delirium. We investigated
whether the implementation of a bundle of non-pharmacological interventions, consisting of environmental noise
and light reduction designed to reduce disturbing patients during the night, was associated with improved sleep and
a reduced incidence of delirium. The study was divided into two parts, before and after changing our practice. One
hundred and sixty-seven and 171 patients were screened for delirium pre- and post-intervention, respectively. Com-
pliance with the interventions was > 90%. The bundle of interventions led to an increased mean (SD) sleep efficiency
index (60.8 (3.5) before vs 75.9 (2.2) after, p = 0.031); reduced mean sound (68.8 (4.2) dB before vs 61.8 (9.1) dB
after, p = 0.002) and light levels (594 (88.2) lux before vs 301 (53.5) lux after, p = 0.003); and reduced number of
awakenings caused by care activities overnight (11.0 (1.1) before vs 9.0 (1.2) after, p = 0.003). In addition, the intro-
duction of the care bundle led to a reduced incidence of delirium (55/167 (33%) before vs 24/171 (14%) after,
p < 0.001), and less time spent in delirium (3.4 (1.4) days before vs 1.2 (0.9) days after, p = 0.021). Increases in
sleep efficiency index were associated with a lower odds ratio of developing delirium (OR 0.90, 95% CI 0.84–0.97).
The introduction of an environmental noise and light reduction programme as a bundle of non-pharmacological
interventions in the intensive care unit was effective in reducing sleep deprivation and delirium, and we propose a
similar programme should be implemented more widely.
.................................................................................................................................................................
Correspondence to: S. Laha
Email: shondipon.laha@lthtr.nhs.uk
*Presented in part at the Intensive Care Society State of the Art Meeting, London, UK, December 2012, and the Associ-
ation of Anaesthetists of Great Britain and Ireland Annual Congress, Dublin, Ireland, September 2013
Accepted: 20 February 2014

Introduction lungs and side-effects from commonly used medica-

Sleep in the intensive care unit (ICU) is often frag- tions [1, 2]. Many studies have reported abnormal
mented and of poor quality. This is largely attributed sleeping patterns and sleep architecture in this
to high levels of noise and light overnight, intense patient population. This is associated with multi-
patient monitoring, mechanical ventilation of the organ dysfunction, and is a leading cause of anxiety

540 © 2014 The Association of Anaesthetists of Great Britain and Ireland

Patel et al. | The effect of a bundle of interventions on sleep and delirium
and stress during ICU admission [2]. Increasing evi-
dence suggests a possible association between sleep
deprivation and delirium [2, 3]. Prolonged sleep
deprivation, immobility, mechanical ventilation of the
lungs, and commonly-used sedative medications all
increase the risk of developing delirium in the ICU.
The development of delirium may impact on a
patient’s recovery from critical illness, and is associ-
ated with increased morbidity and mortality in ICU
populations [3, 4].
The intensive care environment is a commonly
cited reason for poor sleep, which in turn has been
associated with the onset of delirium [1–3]. High lev-
els of noise and the continuous lighting seen in ICU
correlate with poor sleep quality and quantity [5, 6].
The intense patient monitoring required for critically
ill patients is a further cause of sleep disturbance [7].
However, only 10–30% and 10% of arousals in inten-
sive care overnight can be attributed to noise and
patient-care activities, respectively, and sleep depriva-
tion attributed to light alone is not as severe as pre-
viously speculated [6, 8]. Individual reductions in
noise, light and intensive monitoring are therefore
unlikely to impact sleep deprivation significantly [8].
Previous evidence demonstrates that addressing
several of these factors simultaneously can achieve
significant improvements in sleep quality and quan-
tity [5].
Strategies to modify the management of ICU
patients include efforts to avoid deep sedation and
prolonged mechanical ventilation, as these are known
to affect adversely sleep architecture and the risk of
developing delirium [1, 3, 9, 10]. In addition, immobil-
ity associated with ICU admission is thought to impact
negatively on sleep quality and the risk of developing
delirium [4, 11].
Studies outside of ICU have shown that it is possi-
ble to reduce the incidence of delirium through non-
pharmacological interventions [12, 13]. The use of
such strategies remains poorly investigated in ICU
environments [3]. We aimed to reduce the incidence
of sleep deprivation and delirium by collectively
addressing these risk factors through a novel, entirely
non-pharmacological bundle of interventions. The
components of this care bundle were based on prior
research from within and outside the ICU setting.
© 2014 The Association of Anaesthetists of Great Britain and Ireland
Anaesthesia 2014, 69, 540–549
Methods
This study was approved by the Lancaster Ethics Com-
mittee and our local Research and Development Com-
mittee, and written informed consent was obtained
from patients for whom ICU admission was planned
pre-operatively. Patients admitted non-electively were
approached when their clinical status was stable. Inclu-
sion and exclusion criteria for patients to provide sleep
data are summarised in Table 1. Patients with pre-
existing sleep pathologies were not studied as we deter-
mined their sleep would not reflect that of the general
ICU population. Patients with pre-existing cognitive
dysfunction and active delirium were not studied as
they could not be relied upon to provide accurate data
when using sleep questionnaires [14]. Patients with
previous admissions to ICU within the same hospital
admission were not studied due to the persisting sleep
disruption seen after discharge from ICU [15]. Neuro-
surgical patients were not studied due to the need for
hourly neurological observations overnight, which were
considered to be a barrier to sleep. Patients who had
received sedatives within the preceding 24 h were not
studied as these medications are known to impact
negatively on sleep architecture [1, 2, 9].
This was a cohort-based study, and was under-
taken in our 24-bed adult mixed surgical/medical
Table 1 Inclusion and exclusion criteria for comple-
tion of sleep questionnaires during study of a multi-
component bundle of interventions in the intensive
care unit (ICU).
Inclusion criteria
Patient > 18 years of age
Patient spending one or more nights on the ICU
Exclusion criteria
Pre-existing history of sleep pathology, severe visual or
hearing impairment, alcohol addiction or illicit drug
abuse
History of cognitive dysfunction (defined as the presence
or history of dementia, traumatic brain injury, stroke or
hepatic encephalopathy)
Previously discharged from the ICU in this hospital
admission
Neurosurgical patients
Developed delirium at any point during the study
(defined as a single positive result on the Confusion
Assessment Method for the ICU)
Received sedative medications within 24 h
541
Anaesthesia 2014, 69, 540–549 Patel et al. | The effect of a bundle of interventions on sleep and delirium

teaching hospital ICU. It was divided into two phases Table 2 Compliance with each of the components of
– before and after the implementation of the the multidisciplinary bundle of interventions during
multicomponent bundle of interventions. We initially the study period. Compliance is expressed as the num-
ber (proportion) of nights in which all staff followed
gathered baseline data regarding sleep, the environ-
the protocol out of the 26 nights sampled.
ment and the incidence of delirium on our ICU (June
2012). This was followed by a 21-day period in which Compliance
we implemented the multicomponent bundle of inter- Noise
Close all doors 25 (96%)
ventions (Table 2), during which we ran daily staff
Turn monitoring equipment to night 25 (96%)
training sessions. We then repeated the data collection mode between 23:00 and 07:00
to evaluate the impact of the bundle of interventions Reduce volumes on all telephones 25 (96%)
between 23:00 and 07:00
(July/August 2012) (Fig. 1). The study dates were cho- No non-clinical discussions around 24 (92%)
sen to minimise any impact from seasonal changes in patients’ bed spaces
Staff and visitors to speak quietly 25 (96%)
noise, light and temperature from summer to winter.
Offer earplugs to all 26 (100%)
The multicomponent bundle of interventions was patients with Richmond
designed to be multidisciplinary, and included mea- Agitation Sedation Scale score
greater than 4
sures taken to reduce noise, light and iatrogenic sleep Light
disturbance, as well as attempts to modify risk factors Dim main ICU lights between 26 (100%)
23:00 and 07:00
for delirium. The individual components and the pro-
Use bedside lighting 26 (100%)
cess itself were promoted through staff education and for patient care
training sessions run several times each day, which Offer eyemasks to all patients 25 (96%)
with Richmond Agitation
involved provision of background information relating Sedation Scale score greater
to sleep and delirium, as well as practical points than 4
Patient care
regarding how the interventions should be performed.
Group care/procedures 23 (88%)
This was supported by displaying posters in clinical where possible
and non-clinical areas. Eight senior clinical and nurs- Complete care procedures 24 (92%)
before 23:00 or delay their
ing staff were also approached to act as ‘champions’ completion until after 08:00
for the intervention; they received additional education where possible
Orientate patients regarding 26 (100%)
regarding the current evidence surrounding delirium
time, place and
prevention and training regarding the various compo- date every eight hours
nents of the intervention, and acted as a further If patients sleep poorly or 25 (96%)
have a positive result on
resource for members of staff. the Confusion Assessment
Measures to control noise on the ICU included Method for the Intensive
Care Unit, perform a
closing doors when not in use and decreasing the
medication review within 24 h
alarm noise levels on bedside monitors and the volume Set appropriate sedation 26 (100%)
of the telephones. Instructions on how to perform targets once per day
(based on the Richmond
these measures were provided in all patient areas. Sin- Agitation Sedation Scale)
gle-use earplugs (Bilsom 303 SNR 33; Howard Leight, All patients requiring mechanical 26 (100%)
ventilation of the lungs to
Villepinte, Roissy, France) were offered to all patients
be assessed daily for suitability
who had a Richmond Agitation Sedation Scale score for sedation hold or trials of
greater than 4. Measures to reduce light involved the spontaneous breathing
Hourly pain scores and 26 (100%)
implementation of a timed ‘lights off and lights on’ prompt action to optimise
schedule, performing night-time care activities with analgesia
Ensure early mobilisation 23 (88%)
bedside lighting where possible and dimming/switching
when possible and
off any monitor screens not in use. In addition, single- appropriate
use eyemasks (Flight Eyemasks; Dreaming, Zhuji City,
Zhejiang, China) were offered to all patients with a

542 © 2014 The Association of Anaesthetists of Great Britain and Ireland

Patel et al. | The effect of a bundle of interventions on sleep and delirium
Figure 1 Schematic representation of the study design and study dates.
Richmond Agitation Sedation Scale score greater than
4. Night-time routines were adjusted by dimming
overhead lights early, discouraging staff from talking
around bed spaces and encouraging them to group
patient-care activities and treatments to limit the num-
ber of individual disturbances for each patient. To
ensure orientation of patients, nurses were approached
during handover and reminded to perform this task.
We also purchased a number of clocks to allow
patients who were awake to see the time.
To monitor the incidence of delirium, our ICU
observation charts were edited to incorporate routine
Confusion Assessment Method for the ICU measure-
ment. These were presented to clinicians daily during
ward rounds by the nurses, and they were encouraged
to review medications for patients who were found to
have delirium. Similarly, sedation targets, pain scores,
and suitability for trials of spontaneous breathing or
sedation holds were introduced as observations on the
new charts. These measurements were also included
on the daily management charts used for patient
review by the clinicians. This acted as a further
prompt to raise and address relevant issues during
ward rounds. Decisions regarding early mobilisation
were also encouraged daily, and these were performed
and recorded by the physiotherapy team.
Patients who met the inclusion/exclusion criteria
were approached to complete the Richards Campbell
Sleep Questionnaire each morning during their ICU
© 2014 The Association of Anaesthetists of Great Britain and Ireland
Anaesthesia 2014, 69, 540–549
stay. This five-parameter visual analogue questionnaire
assesses sleep depth, latency, number of awakenings,
time spent awake and overall sleep quality. It has been
validated against polysomnography, the gold standard
for assessing sleep [14]. The mean score from this
questionnaire was used to estimate the sleep efficiency
index [16]. In addition, a Sleep in Intensive Care
Questionnaire was given to the patients who com-
pleted the Richards Campbell Sleep Questionnaire
immediately after discharge [14]. This questionnaire
allows patients to rate their overall sleep quality before
and during ICU admission and their levels of daytime
sleepiness. In addition, it allows patients to rate poten-
tial sleep disruptive factors on a scale of 1–10.
We measured light and sound levels on the ICU
using two CEM DT-8820 environmental meters (CEM
Shenzhen Everbest Machinery Industry Company,
Nanshan, Shenzhen, China) placed centrally for the
full duration of the study. To evaluate patients’ sleep-
ing patterns, nursing staff recorded assessments of
whether patients were asleep or awake each hour.
Nursing staff also kept a record of the total number of
patient-care activities performed overnight and the
number of times patients were woken as a result.
All patients with a Richmond Agitation Sedation
Scale score of greater than 4 were screened for delir-
ium using the Confusion Assessment Method for the
ICU at 08:00, 14:00, 18:00 and where appropriate at
02:00 by a member of the research team. This tool has
543
Anaesthesia 2014, 69, 540–549 Patel et al. | The effect of a bundle of interventions on sleep and delirium

been shown to have excellent sensitivity and specificity
for detecting delirium [3, 4].
At the time of conducting the study, there were no
reported papers published to inform a power calcula-
tion and determine a sample size. We instead screened
all appropriate patients for delirium admitted to the
ICU and approached all patients who met the inclu-
sion/exclusion criteria to provide sleep data and during
the pre- and post-intervention study phases. Indepen-
dent t-tests were used to the compare age, Acute Phys-
iology and Chronic Health Evaluation Score
(APACHE)-2 [17] and length of stay between the total
ICU population and the subpopulation who provided
sleep data in the pre- and post-intervention cohorts.
We also used the chi-squared test and Fisher’s exact
test to assess for differences in sex, admitting specialty
and reason for admission between the two overall ICU
populations and the associated subgroups who com-
pleted the questionnaires. As the pre- and post-inter-
vention cohorts contained entirely different patients,
we used the independent t-test to compare the sleep
efficiency index before and after the multicomponent
bundle of interventions. As patients were asked to
complete a copy of the Richards Campbell Sleep Ques-
tionnaire during each morning of their ICU admission,
one questionnaire per patient was selected at random
to be included in our data analysis, to reduce bias
from multiple questionnaires completed by the same
patient. We used the independent t-test and the
Mann–Whitney U-test to compare sleep quality before
ICU admission, average sleep quality during ICU
admission, patients’ rating of possible barriers to sleep,
sound and light levels and daytime sleepiness. We used
an independent t-test to identify changes in the total
hours of sleep achieved overnight, and the chi-squared
test to identify significant change in the number of
nights containing a 3-h period of uninterrupted sleep.
We used binary logistic regression to derive the odds
ratio (OR) of a patient developing delirium based on
their sleep efficiency index. Statistical analysis was per-
formed using Microsoft Excel (version 12.2.5; Micro-
soft, WA, USA) and SPSS Statistics software (version
22; IBM, Portsmouth, UK).
Results
A total of 167 patients were included before the multi-
component bundle of interventions, and a further 171
patients after it had been implemented. Fifty-nine of
these patients completed the Richards Campbell Sleep
Questionnaire during their ICU admission (30 before
the intervention and 29 after it). Overall, the baseline
characteristics of the two cohorts were similar
(Table 3). There was no difference between the two
cohorts regarding median (IQR [range]) sleep quality
before hospital admission, 8 (7–9 [5–10]) vs 8 (8–9
[5–10]), p = 0.107.

Table 3 Baseline characteristics for the two patient cohorts before and after the multicomponent bundle of interven-
tions. Values are mean (SD) or number (proportion).

Before After

Questionnaire Questionnaire
Total completed Total completed
(n = 167) (n = 30) (n = 171) (n = 29)
Age; years 60.0 (13.7) 61.9 (13.1) 60.6 (16.3) 60.5 (15.7)
Men 85 (51%) 14 (47%) 91 (53%) 16 (55%)
Speciality
Respiratory 27 (16%) 5 (17%) 20 (12%) 5 (17%)
Cardiovascular 6 (4%) 3 (10%) 8 (5%) 3 (10%)
Gastroenterology 8 (5%) 2 (7%) 9 (5%) 1 (3%)
General surgery 40 (24%) 12 (40%) 45 (26%) 9 (31%)
Urology/gynaecology 8 (5%) 6 (20%) 14 (8%) 3 (10%)
Miscellaneous 20 (12%) 2 (7%) 28 (16%) 8 (28%)
Neurosurgery 58 (35%) 0 47 (27%) 0
Elective admission 135 (81%) 26 (86%) 153 (89%) 25 (87%)
APACHE-2 score 15.0 (7.6) 14.7 (6.0) 14.2 (6.6) 13.5 (8.0)
Length of stay; days 6.3 (2.0) 6.8 (4.5) 6.3 (3.7) 6.7 (1.5)

APACHE-2, Acute Physiology and Chronic Health Evaluation Score-2.

544 © 2014 The Association of Anaesthetists of Great Britain and Ireland

Patel et al. | The effect of a bundle of interventions on sleep and delirium Anaesthesia 2014, 69, 540–549

Compliance with the multicomponent bundle of
interventions was > 90%. Implementation of the bun-
dle led to a reduction in mean (SD) night-time noise
(68.8 (4.2) dB before vs 61.8 (9.1) dB after, p = 0.002);
light (594 (88.2) lux before vs 301 (53.5) lux after,
p = 0.003); number of staff–patient interactions over-
night (33.6 (4.3) before vs 23.4 (6.6) after, p = 0.045)
and number of times patients were woken due to staff
interventions (11.0 (1.1) before vs 9.0 (1.2) after,
p = 0.003). Compared to the pre-intervention group,
there was an increase in the mean (SD) sleep efficiency
index following the implementation of the multicom-
ponent intervention, 60.8 (3.5) vs 75.9 (2.24),
p < 0.001 (Fig. 2), and an increase in sleep quality, 4
(3–5 [2–7] vs 7 (7–8 [5–9]), p < 0.001. There was also
a significant reduction in daytime sleepiness following
the implementation of the intervention bundle (6 (6–8
[1–10]) vs 3 (3–4 [1–7]), p = 0.042). In addition, there
were significant reductions in the ratings given by
patients for three key barriers to sleep identified from
the pre-intervention Sleep in Intensive Care Question-
naire data: noise (7 (6–8 [1–10]) vs 2 (1–2 [1–4]),
p < 0.001; light (5.5 (4–7 [1–10]) vs 1 (1–2 [1–4]),
p = 0.011 and nursing interventions (4 (2–7 [1–9]) vs
1 (1–2 [1–4]), p = 0.043 (Fig. 3).
Following the implementation of the intervention
bundle, patients spent more time asleep at night (6.6 h
(55%) before vs 8.6 h (72%) after, p < 0.001, and more
patient nights contained a 3-h window of uninter-
rupted sleep (32% before vs 39% after, p = 0.029).
Implementation of the bundle of interventions
resulted in a reduction in the incidence of delirium
(55/167 (33%) before vs 24/171 (14%) after,
p < 0.001), OR 0.33 (95% CI 0.19–0.57). Furthermore,
there was a reduction in the mean (SD) length of time
spent delirious (3.4 (1.4) days before vs 1.2 (0.9) days

Richards campbell sleep quesƟonnaire score

Figure 2 Richards-Campbell Sleep Questionnaire scores included in data analysis before and after implementation of
the multicomponent intervention. Each parameter carries a maximum score of 100, with higher scores representing a
positive response. White, before the intervention, grey, after the intervention. The boxes represent the first quartile,
median and third quartile for each test parameter and the whiskers represent the range. *p < 0.05.

© 2014 The Association of Anaesthetists of Great Britain and Ireland 545

Anaesthesia 2014, 69, 540–549 Patel et al. | The effect of a bundle of interventions on sleep and delirium

10

Patients rating of sleep disruption caused

8

Noise* Light* Nurse Diagnostic Recording Phebotomy Administration of

interventions* investigations observations medication

Sources of sleep disruption

Figure 3 Patient ratings of sleep disruption from a given parameter on the Sleep in Intensive Care Questionnaire.
Each parameter carries a maximum score of 10, with higher scores representing a negative response. The bars repre-
sent the median response and the lines are the standard error. White, before the multicomponent intervention; grey,
after the intervention. *p < 0.05.

after, p = 0.021). Patient reporting high sleep efficiency quality and quantity in ICU [20]. Reducing iatrogenic
index scores demonstrated a significantly reduced risk sleep disruption through grouping interventions has
of delirium (OR 0.90, 95% CI 0.84–0.97). also been suggested in combination with other strate-
gies [5, 21]. Regular re-orientation has been effective
Discussion in reducing sleep disruption and delirium when used
The multicomponent bundle of interventions increased alongside other interventions in general medical
both qualitative and quantitative measures of sleep. patients [12, 13] Regular assessment and provision of
Patients also reported that the key environmental bar- adequate analgesia is a further strategy suggested to
riers to sleep seen in the ICU had been addressed. This improve sleep in the ICU [8].
is reflected in the reductions in noise and light levels Sedative agents are widely used to treat sleep
and iatrogenic sleep disturbances overnight. Further- deprivation in the ICU. However, these are associated
more, there were a greater number of nights contain- themselves with disturbance of sleep architecture and
ing prolonged periods of uninterrupted sleep. increased risk of delirium [1, 3, 9]; benzodiazepines
Importantly, implementation of the bundle of inter- have been shown to be particularly culpable [9, 22–
ventions led to a reduction in the incidence and dura- 24]. Avoiding benzodiazepines through the use of
tion of delirium, and better sleep was associated with a alternate sedatives such as dexmedetomidine has been
reduced risk of developing delirium. shown to be effective in reducing the incidence of
The interventions used in this study incorporated delirium in ICU [25–27]. Employing daily sedation
strategies that have shown benefit in previous studies. targets allows clinicians to identify the lowest dose of
Closing doors, asking staff and visitors to speak qui- sedatives that ensures patient comfort and safety while
etly, reducing equipment volumes and the provision of reducing adverse effects sleep and delirium [4]. The
earplugs have been shown to improve sleep in ICU [2, use of spontaneous awakening trials (daily assessment
5, 18]. One study demonstrated the use of earplugs for sedation holds) and spontaneous breathing trials is
was associated with protective effects from cognitive recommended [4]. Girard et al. demonstrated that
dysfunction; however, the tool used in this study to these successfully reduced the number of days spent in
assess for delirium has not been validated in patients coma and the duration of mechanical ventilation.
whose trachea are not intubated [18, 19]. Eye masks However, this approach alone failed to reduce the
have previously shown success in improving sleep overall incidence of delirium [28]. Lastly, Needham

546 © 2014 The Association of Anaesthetists of Great Britain and Ireland

Patel et al. | The effect of a bundle of interventions on sleep and delirium
et al. and Schweickert et al. showed that early physical
therapy reduced the duration of delirium [29, 30].
When combined into a bundle of interventions,
the use of the above measures led to an increase in the
mean sleep efficiency index in our patients, represent-
ing a significant improvement in patients’ perception
of sleep. This is reflected in the findings from the Sleep
in Intensive Care Questionnaire, with patients report-
ing significantly better average sleep quality during
admission after the intervention was implemented. The
three major perceived barriers to sleep before the
intervention (noise, light and staff activities) were suc-
cessfully addressed. As previous studies demonstrate
that the key issue experienced by ICU patients is sleep
fragmentation, it is likely that the greater number of
nights containing prolonged uninterrupted periods of
sleep after implementation of the bundle of interven-
tions allowed patients to reach deeper and more
restorative stages of sleep [2, 31]. We would be inter-
ested to see if further work using polysomnography
could characterise the possible changes in sleep archi-
tecture that may be associated with this bundle of
interventions.
The bundle of interventions also resulted in a signif-
icant reduction in the incidence of delirium. Inoye et al.
and Vidan et al. have previously achieved reductions of
5.1–6.8% in non-ICU populations using non-pharmaco-
logical strategies [12, 13]. The larger reduction seen in
this study may be due to the greater number of modifi-
able risk factors associated with delirium and the higher
baseline incidence of the condition in our ICU. Previous
studies have demonstrated that delirium is associated
with increased morbidity and mortality [32]. The
reduced incidence seen in this study should therefore
theoretically benefit ICU patients, although we have not
studied this directly. Both ICU and overall hospital costs
are increased in the presence of delirium [33]. In the
light of this, our multicomponent bundle of interven-
tions could also reduce healthcare costs, although pro-
spective cost analysis is required.
The bundle of interventions also reduced the dura-
tion of time patients spent in delirium. This was likely
to have been achieved through addressing causative
and associative factors, and earlier recognition of the
condition leading to faster treatment, as well as the
effect of all the different components on duration and
© 2014 The Association of Anaesthetists of Great Britain and Ireland
Anaesthesia 2014, 69, 540–549
quality of sleep. Delays in the detection and treatment
of delirium have been shown to result in increased
mortality [34]. Implementation of a programme of
interventions, including education of staff regarding
the diagnosis and management of delirium, may
reduce length of stay in ICU, although again we have
not studied this directly.
This study indicates the potential to predict patients
developing delirium, as low sleep efficiency index scores
demonstrated an increased risk of developing delirium.
The PRE-DELERIC study developed a system that
proved highly predictive for the development of delir-
ium [35]. Based on the data gathered here, sleep ques-
tionnaires may be a useful addition to such a
predicative system. Further research involving combin-
ing sleep parameters into such a predictive system is
required before definitive conclusions are drawn.
There are some non-pharmacological strategies to
improve sleep that were not used in this bundle of inter-
ventions. Cognitive behavioural therapy, massage, thera-
peutic touch and aromatherapy have been shown by
some authors to be effective [36, 37]. These were not
included due to the lack of well-designed studies and the
need for additional equipment and trained personnel. In
addition, the safety of such measures remains unclear
[36, 37]. Sedatives used to treat sleep deprivation can
paradoxically damage sleep architecture and increase
the risk of delirium, and hence they were not incorpo-
rated into this intervention [9, 23, 24]. Conflicting evi-
dence exists surrounding the use of prophylactic
medications to prevent delirium. The MIND study dem-
onstrated no benefit from haloperidol or ziprazidone
over placebo to prevent delirium [38]. Similar findings
have been reported in two smaller studies [39, 40]. Con-
versely, Wang et al. demonstrated a lower incidence of
delirium in ICU patients receiving prophylactic haloper-
idol compared with controls [41]. However, this study
enrolled patients who were not severely ill [42]. Prophy-
lactic administration of atypical antipsychotics has been
used to prevent delirium. Prophylactic olanzapine did
reduce the overall incidence of delirium in patients after
hip surgery; however, it was also associated with
increased severity and prolonged duration of the condi-
tion compared to controls [43]. Dexmedetomidine
appears to carry a lower risk of delirium compared with
benzodiazepines; however, further research from
547
Anaesthesia 2014, 69, 540–549 Patel et al. | The effect of a bundle of interventions on sleep and delirium
well-designed randomised controlled trials are required
to fully define its use [44]. Currently, there is a lack of
evidence to support the routine use of prophylactic
medications to prevent delirium in the generalised ICU
population [42].
There are limitations to this study that must be
considered. These include the single-centre design and
the use of non-randomised cohorts, which may have
resulted in selection bias. However, we have
attempted to address this though a historical control
group sampled immediately before introducing the
bundle of interventions. The short time span between
the two phases of data collection makes it unlikely
that the two cohorts were subject to changes in prac-
tice that would impact sleep and delirium. The tools
used in this study to assess sleep relied on self-assess-
ment. Consideration was given to using an objective
measurement of sleep, but this was abandoned due to
the technical difficulties encountered when attempting
to use this technique on a wider scale. However, one
of the sleep questionnaires used has been validated
against polysomnography. The fact that staff were
aware that the study was being performed was
unavoidable, as the changes implemented through the
multicomponent intervention required their awareness
and compliance. Indeed, there is a danger that in
non-research conditions, without the additional pro-
motion and endorsement of the intervention, the ben-
efits seen in this study may not be replicated. To
assess whether this is the case, we would like to
repeat the study in the near future to assess whether
the bundle of interventions is still being used and
remains effective outside research conditions.
We have demonstrated that a multicomponent
multidisciplinary bundle of interventions, with appro-
priate education and promotion, is safe, effective and
practical for use on the ICU, and may lead to more
and better quality sleep, with an associated decrease in
delirium. Consideration should be given to rolling out
such a multidisciplinary bundle of interventions more
widely, although further research is required in this
area to examine the effect on outcomes. In addition,
the study findings encourage the development of a
universal programme to raise awareness amongst ICU
staff of the importance of delirium screening for all
ICU patients.
548
Acknowledgements
We would like to thank Dr. Margaret Hurley for her
invaluable advice regarding statistical analysis.
Competing interests
No external funding and no competing interests
declared.
References
1. Boyko Y, Ording H, Jennum P. Sleep disturbances in critically ill
patients in ICU: how much do we know? Acta Anaesthesiolog-
ica Scandinavica 2012; 56: 950–8.
2. Kamdar BB, Needham DM, Collop NA. Sleep deprivation in
critical illness: its role in physical and psychological recovery.
Journal of Intensive Care Medicine 2012; 27: 97–111.
3. Girard TD, Pandharipande PP, Ely EW. Delirium in the intensive
care unit. Critical Care 2008; 12: S3.
4. Vasilevskis EE, Ely EW, Speroff T, et al. Reducing iatrogenic
risks: ICU-acquired delirium and weakness – crossing the qual-
ity chasm. Chest 2010; 138: 1224–33.
5. Li SY, Wang TJ, Vivienne Wu SF, Liang SY, Tung HH. Efficacy of
controlling night-time noise and activities to improve patients’
sleep quality in a surgical intensive care unit. Journal of Clini-
cal Nursing 2011; 20: 396–407.
6. Weinhouse GL, Schwab RJ. Sleep in the critically ill patient.
Sleep 2006; 29: 707–16.
7. Tamburri LM, DiBrienza R, Zozula R, Rederker NS. Nocturnal
care interactions with patients in critical care units. American
Journal of Critical Care 2004; 13: 102–12.
8. Hardin KA. Sleep in the ICU: potential mechanisms and clinical
implications. Chest 2009; 136: 284–94.
9. Bourne RS, Mills GH. Sleep disruption in critically ill patients –
pharmacological considerations. Anaesthesia 2004; 59: 374–
84.
10. Cavallazzi R, Saad M, Marik PE. Delirium in the ICU: an over-
view. Annals of Intensive Care 2012; 2: 49.
11. Ugrasß GA, Oztekin SD. Patient perception of environmental
and nursing factors contributing to sleep disturbance in a neu-
rosurgical intensive care unit. Tohoku Journal of Experimental
Medicine 2007; 212: 299–308.
12. Inouye SK, Bogardus ST, Charpentier PA, et al. A multicompo-
nent intervention to prevent delirium in hospitalized older
patients. New England Journal of Medicine 1999; 340: 669–
76.
13. Vidan MT, Sanchez E, Alonso M, et al. An intervention inte-
grated into daily clinical practice reduces the incidence of
delirium during hospitalization in elderly patients. Journal of
the American Geriatrics Society 2009; 57: 2029–36.
14. Bourne RS, Minelli C, Mills GH, Kandler R. Clinical review:
sleep measurement in critical care patients: research and clin-
ical implications. Critical Care 2007; 11: 226.
15. McKinley S, Fien M, Elliot R, Elliot D. Sleep and psychological
health during early recovery from critical illness: an observa-
tional study. Journal of Psychosomatic Research 2013; 75:
539–45.
16. Richards KC. Techniques for measurement of sleep in critical
care. Focus on Critical Care 1987; 14: 34–40.
17. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a
severity of disease classification system. Critical Care Medicine
1985; 13: 818–29.
© 2014 The Association of Anaesthetists of Great Britain and Ireland
Patel et al. | The effect of a bundle of interventions on sleep and delirium
18. Van Rompaey B, Elseviers MM, Van Drom W, Fromont V,
Jorens PG. The effect of earplugs during the night on the
onset of delirium and sleep perception: a randomised con-
trolled trial in intensive care patients. Critical Care 2012; 16:
R73.
19. Van Rompaey B, Schuurmans MJ, Shortridge-Baggett LM, Trui-
jen S, Elseviers M, Bossaert L. A comparison of the CAM-ICU
and the NEECHAM Confusion Scale in intensive care delirium
assessment: an observational study in non-intubated patients.
Critical Care 2008; 12: R16.
20. Jones C, Dawson D. Eye masks and earplugs improve patient’s
perception of sleep. Nursing in Critical Care 2012; 17: 247–54.
21. Le A, Friese RS, Hsu CH, Wynne JL, Rhee P, O’Keeffe T. Sleep
disruptions and nocturnal nursing interactions in the intensive
care unit. Journal of Surgical Research 2012; 177: 310–4.
22. Mistraletti G, Carloni E, Cigada M, et al. Sleep and delirium in
the intensive care unit. Minerva Anaestesiologica 2008; 74:
329–33.
23. Pisani MA, Murphy TE, Araujo KL, Slattum P, Van Ness PH, Ino-
uye SK. Benzodiazepine and opioid use and the duration of
intensive care delirium in an older population. Critical Care
Medicine 2009; 37: 177–83.
24. Van Rompaey B, Schuurmans MJ, Shortridge-Baggett LM, Trui-
jen S, Bossaert L. Risk factors for intensive care delirium: a
systematic review. Intensive and Critical Care Nursing 2008;
24: 98–107.
25. Pandharipande PP, Pun BT, Heer DL, et al. Effect of sedation
with dexmedetomidine vs lorazepam on acute brain dysfunc-
tion in mechanically ventilated patients: the MENDS random-
ized controlled trial. Journal of the American Medical
Association 2007; 298: 2644–53.
26. Riker RR, Shehabi Y, Bokesch PM, et al. Dexmedetomidine vs
midazolam for sedation of critically ill patients: a randomized
trial. Journal of the American Medical Association 2009; 4:
489–99.
27. Maldonado JR, Wysong A, van der Starre PJ, Block T, Miller C,
Reitz BA. Dexmedetomidine and the reduction of postopera-
tive delirium after cardiac surgery. Psychosomatics 2009; 50:
206–17.
28. Girard TD, Kress JP, Fuchs BD, et al. Efficacy and safety of a
paired sedation and ventilator weaning protocol for mechani-
cally ventilated patients in intensive care (Awakening and
Breathing Controlled trial): a randomised controlled trial. Lan-
cet 2008; 371: 126–34.
29. Needham DM, Korupolu R, Zanni JM, et al. Early physical
medicine and rehabilitation for patients with acute respiratory
failure: a quality improvement project. Archives of Physical
Medicine and Rehabilitation 2010; 91: 536–42.
30. Schweickert WD, Pohlman MC, Pohlman AS, et al. Early physi-
cal and occupational therapy in mechanically ventilated,
critically ill patients: a randomised controlled trial. Lancet
2009; 373: 1874–82.
© 2014 The Association of Anaesthetists of Great Britain and Ireland
Anaesthesia 2014, 69, 540–549
31. Freedman NS, Gazendam J, Levan L, Pack AI, Scwab RJ. Abnor-
mal sleep/wake cycles and the effect of environmental noise
and sleep disruption in the intensive care unit. American
Journal of Respiratory and Critical Care Medicine 2001; 162:
451–7.
32. Ouimet S, Kavanagh BP, Gottfried SB, Skrobik Y. Incidence, risk
factors and consequences of ICU delirium. Intensive Care Med-
icine 2007; 33: 66–73.
33. Milbrandt EB, Deppen S, Harrison PL, et al. Costs associated
with delirium in mechanically ventilated patients. Critical Care
Medicine 2004; 32: 955–62.
34. Heymann A, Radtke F, Schiemann A, et al. Delayed treatment
of delirium increases mortality rate in intensive care unit
patients. Journal of International Medical Research 2010; 38:
1584–95.
35. van den Boogaard M, Pickkers P, Slooter AJ, et al. Develop-
ment and validation of PRE-DELIRIC (PREdiction of DELIRium in
ICu patients) delirium prediction model for intensive care
patients: observational multicentre study. British Medical Jour-
nal 2012; 344: e420.
36. Richards KC, Nagel C, Markie M, Elwell J, Barone C. Use of
complementary and alternative therapies to promote sleep in
critically ill patients. Critical Care Nursing Clinics of North
America 2003; 15: 329–40.
37. Richards KC, Gibson R, Overton-McCoy AL. Effects of massage
in acute and critical care. AACN Clinical Issues Advanced Prac-
tice in Acute and Critical Care 2000; 11: 77–96.
38. Girard TD, Pandharipande PP, Carson SS, et al. Feasibility, effi-
cacy, and safety of antipsychotics for intensive care unit delir-
ium: the MIND randomized, placebo-controlled trial. Critical
Care Medicine 2010; 38: 428–37.
39. Kalisvaart KJ, de Jonghe JF, Bogaards MJ, et al. Haloperidol
prophylaxis for elderly hip-surgery patients at risk for delir-
ium: a randomized placebo-controlled study. Journal of the
American Geriatric Society 2005; 53: 1658–66.
40. Page VJ, Ely EW, Gates S, et al. Effect of intravenous haloperi-
dol on the duration of delirium and coma in critically ill
patients (Hope-ICU): a randomised, double-blind, placebo-
controlled trial. Lancet Respiratory Medicine 2013; 1: 515–23.
41. Wang W, Li HL, Wang DX, et al. Haloperidol prophylaxis
decreases delirium incidence in elderly patients after noncar-
diac surgery: a randomised controlled trial. Critical Care Medi-
cine 2012; 40: 731–9.
42. Brummel NE, Girard TD. Preventing delirium in the intensive
care unit. Critical Care Clinics 2013; 29: 51–65.
43. Larsen KA, Kelly SE, Stern TA, et al. Administration of olanza-
pine to prevent postoperative delirium in elderly joint-
replacement patients: a randomized, controlled trial. Psycho-
somatics 2010; 51: 409–18.
44. Mo Y, Zimmermann AE. Role of dexmedetomidine for the pre-
vention and treatment of delirium in intensive care patient.
Annals of Pharmacotherapy 2013; 47: 869–76.
549