YGYNO-977215; No.

of pages: 7; 4C:
Gynecologic Oncology xxx (2018) xxx–xxx

Contents lists available at ScienceDirect

Gynecologic Oncology

journal homepage: www.elsevier.com/locate/ygyno

Enhanced Recovery After Surgery (ERAS) in gynecologic oncology:

System-wide implementation and audit leads to improved value and
patient outcomes
S.P. Bisch a,⁎, T. Wells b, L. Gramlich c, P. Faris d, X. Wang d, D.T. Tran e, N.X. Thanh e, S. Glaze a, P. Chu a,
P. Ghatage a, J. Nation a, V. Capstick b, H. Steed b, J. Sabourin b, G. Nelson a
a
Division of Gynecologic Oncology, Tom Baker Cancer Centre, Calgary, Alberta, Canada
b
Division of Gynecologic Oncology, Cross Cancer Institute, Edmonton, Alberta, Canada
c
Department of Medicine, University of Alberta, Edmonton, Alberta, Canada
d
Alberta Health Services, Calgary, Alberta, Canada
e
Institute of Health Economics, Edmonton, Alberta, Canada

H I G H L I G H T S

• Implementation of ERAS gynecologic oncology guidelines results in significant clinical improvements and cost savings.
• Use of an audit system allows measurement of compliance to the individual ERAS recommendations.
• ERAS teams should strive to improve compliance to guidelines as this translates into improved outcomes.

a r t i c l e i n f o a b s t r a c t

Article history: Objective. Enhanced recovery pathways have been shown to reduce length of stay without increasing read-
Received 3 July 2018 mission or complications in numerous areas of surgery. Uptake of gynecologic oncology ERAS guidelines has
Received in revised form 28 July 2018 been limited. We describe the effect of ERAS guideline implementation in gynecologic oncology on length of
Accepted 4 August 2018
stay, patient outcomes, and economic impact for a province-wide single-payer system.
Available online xxxx
Methods. We compared pre- and post-guideline implementation outcomes in consecutive staging and
Keywords:
debulking patients at two centers that provide the majority of surgical gynecologic oncology care in Alberta,
ERAS Canada between March 2016 and April 2017. Clinical outcomes and compliance were obtained using the ERAS
Gynecologic oncology Interactive Audit System. Patients were followed until 30 days after discharge. Negative binomial regression
Cost savings was employed to adjust for patient characteristics.
Clinical outcomes Results. We assessed 152 pre-ERAS and 367 post-ERAS implementation patients. Mean compliance with ERAS
care elements increased from 56% to 77.0% after implementation (p b 0.0001). Median length of stay for all sur-
geries decreased from 4.0 days to 3.0 days post-ERAS (p b 0.0001), which translated to an adjusted LOS decrease
of 31.4% (95% CI = [21.7% - 39.9%], p b 0.0001). In medium/high complexity surgery median LOS was reduced by
2.0 days (p = 0.0005). Complications prior to discharge decreased from 53.3% to 36.2% post-ERAS (p = 0.0003).
There was no significant difference in readmission (p = 0.6159), complications up to 30 days (p = 0.6274), or
mortality (p = 0.3618) between the cohorts. The net cost savings per patient was $956 (95%CI: $162 to $1636).
Conclusions. Systematic implementation of ERAS gynecologic oncology guidelines across a healthcare system
improves patient outcomes and saves resources.
© 2018 Elsevier Inc. All rights reserved.

1. Introduction

Enhanced recovery programs aim to improve patient outcomes and

⁎ Corresponding author at: Tom Baker Cancer Centre, 1331 29th St NW, Calgary,
Alberta, T2N 4N2, Canada.
E-mail address: steven.bisch@ahs.ca (S.P. Bisch).
decrease resource utilization through implementation of evidence-
based practice. These programs go by a variety of names including
fast-track surgery, evidence-based surgery, multimodal care, and En-
hanced Recovery After Surgery (ERAS). ERAS has been successfully

https://doi.org/10.1016/j.ygyno.2018.08.007
0090-8258/© 2018 Elsevier Inc. All rights reserved.

Please cite this article as: S.P. Bisch, et al., Enhanced Recovery After Surgery (ERAS) in gynecologic oncology: System-wide implementation and
audit leads to improved value and p..., Gynecol Oncol (2018), https://doi.org/10.1016/j.ygyno.2018.08.007
2 S.P. Bisch et al. / Gynecologic Oncology xxx (2018) xxx–xxx

implemented across a wide array of surgical disciplines; the goal of this
multimodal evidence-based program is to reduce hospital length of stay
and complications through modifying the surgical stress response [1–5].
Every year over 1,000,000 women globally develop a gynecologic ma-
lignancy [6]. The mainstay of treatment involves surgical intervention. Re-
cently published ERAS guidelines have incorporated evidence-based
perioperative principles for gynecologic oncology surgery [2,3]. Consis-
tent uptake of these principles has been limited, and few centers are
performing continuous quality improvement to ensure reliable compli-
ance with ERAS recommendations [7]. Studies of ERAS that have included
gynecologic oncology surgery have previously only involved single aca-
demic centers and have shown conflicting results around the efficacy of
ERAS in decreasing length of stay, complications, and cost [8,9].
Alberta Health Services (AHS) is a publicly funded provincial
healthcare system serving a population of over 4 million people in 59
acute care facilities. Gynecologic oncology surgery for the province is
primarily performed at two tertiary referral centers. Following a suc-
cessful implementation program in colorectal surgery [10], AHS began
implementing ERAS for gynecologic oncology surgery across the prov-
ince in November 2016.
The aim of this study was to evaluate the impact of the ERAS gyneco-
logic oncology guidelines on length of stay, complication rates,
readmissions, and cost using a system-wide implementation program.
2. Methods
The AHS ERAS Implementation Program (EIP) for gynecologic oncol-
ogy began in November 2016. Implementation occurred at the two
major referral centers for gynecologic oncology surgery in the
healthcare system: Foothills Medical Centre (FMC) in Calgary, Alberta
and Royal Alexandra Hospital (RAH) in Edmonton, Alberta.
The ERAS gynecologic oncology guideline includes 20 care elements
that are followed during the pre-, intra-, and post-operative period
(Table 1) [2,3]. The AHS EIP has been described in detail elsewhere
[7,10] and includes: (i) the formation of an implementation team (in-
cluding a surgeon/medical lead, anaesthesiologist, and nurse); (ii) col-
lection of pre-ERAS (baseline) data for a minimum of 50 consecutive
staging, and 50 consecutive debulking patients; (iii) entry of data into
the web-based ERAS Interactive Audit System (EIAS) [www.
erassociety.org, ENCARE, Kista, Sweden] including data on care element
compliance; (iv) auditing of pre-ERAS data with subsequent tailored
training of those involved (to address location specific compliance defi-
ciencies and ensure international standardization); (v) preparation of
involved teams (e.g. in-service meetings with nurses, creation of elec-
tronic order sets); (vi) prospectively recruiting consecutive surgical pa-
tients and (vii) biweekly compliance auditing using EIAS with on-going
targeted training to address areas of low compliance. Although initial
versions of the EIAS were tailored to colorectal surgery, this study is
the first to utilize a version of the EIAS that is specifically programed
to include the ERAS gynecologic oncology guidelines [2,3].
The pre-ERAS cohort was comprised of patients from March 2016 to
September 2016 and consisted of 76 consecutive debulking and 76 con-
secutive staging surgical patients to ensure adequate representation of
the breadth of gynecologic oncologic surgeries. The post –ERAS cohort
consisted of all consecutive gynecologic oncology surgical patients
from November 2016 to April 2017. Surgical complexity was scored

Table 1
Description of ERAS gynecologic oncology guideline elements.

Guiding question Mode of measurement

Preop
Preadmission patient education
Avoidance of oral bowel
preparation
Oral carbohydrate treatment
Avoidance of long-acting sedative
medication
Thrombosis prophylaxis
Antibiotic prophylaxis before
incision
PONV prophylaxis administered
Did the patient get specific ERAS information preoperatively?
Did the patient receive oral bowel preparation preoperatively?
Was the patient treated with a preoperative carbohydrate-rich drink? Yes = compliant; no-contraindicated = compliant; no-any other
Did the patient get any long-acting sedative premedication after
midnight prior to surgery?
Did the patient get thrombosis prophylaxis preoperatively?
Was antibiotic prophylaxis given before skin incision?
Was PONV prophylaxis given before operation?
Yes = compliant; no = non-compliant
No = compliant; yes = non-compliant
reason = non-compliant
No = compliant; yes = non-compliant
Anticoagulant or anticoagulant and compression = compliant; no
= non-compliant
Yes = compliant; no = non-compliant
Yes = compliant; no = non-compliant

Intraop
Avoidance of epidural/spinal
anesthesia
Avoidance of systemic opioids
Upper-body forced-air heating
cover used
Avoidance of nasogastric tube use
Avoidance of resection-site
drainage
Did the patient receive intraoperative epidural? No = compliant; yes = non-compliant
Did the patient receive long-acting systemic opioids No = compliant; yes = non-compliant
intraoperatively?
Was the patient kept warm with an upper-body forced-air heating Yes = compliant; no = non-compliant
cover during the operation?
Was a nasogastric tube left in place after the operation? No = compliant; yes = non-compliant
Abdominal and pelvic drains No drain = compliant; drain = non-compliant

Postop
Prompt termination of urinary
drainage
Stimulation of gut motility
Patient weight recorded POD1
Prompt termination of intravenous
fluid infusion
Energy intake (oral nutritional
suppl) POD0
Energy intake (oral nutritional
suppl) POD1
Mobilization at all on day of
surgery
30 day follow up performed
When was urinary drainage successfully terminated?
Was the patient's gut motility stimulated?
What was patient's weight POD1 (in am)?
When was the intravenous infusion successfully terminated?
How much energy from ONS did the patient get on POD0?
How much energy from ONS did the patient get on POD1?
Did the patient mobilize at all postoperatively, on day of surgery?
Was the 30-day follow-up performed?
Removed POD1 = compliant; removed N POD1 = non-compliant
Laxatives, chewing gum = compliant; No = non-compliant
Weight gainb2 kg = compliant; weight gain ≥ 2 kg =
non-compliant
On day of operation = compliant; on day after operation =
non-compliant
≥300 kcal = compliant; b300 kcal = non-compliant
≥600 kcal = compliant; b600 kcal = non-compliant
Yes = compliant; no = non-compliant
Yes = compliant; no = non-compliant

Please cite this article as: S.P. Bisch, et al., Enhanced Recovery After Surgery (ERAS) in gynecologic oncology: System-wide implementation and
audit leads to improved value and p..., Gynecol Oncol (2018), https://doi.org/10.1016/j.ygyno.2018.08.007
 S.P. Bisch et al. / Gynecologic Oncology xxx (2018) xxx–xxx 3

using the validated scoring system described by Aletti et al. [11]. Patient The average cost per day was estimated to be $2397 among pre-ERAS
comorbidity was classified using the Charlson comorbidity index [12]. patients. As the average cost per day in the whole LOS is lower than
We removed “cancer” and “metastatic solid tumour” from the Charlson the cost per day in the first few days when most procedures/treatments
index so that we could independently assess the impact of patient char- were applied to patients and higher than the cost per day in the last few
acteristics and malignancy on our outcome measures. All patients un- days, we used the marginal cost for the days shortened by ERAS. We as-
dergoing open gynecologic oncology surgery in the province of sumed this cost equated to the “hotel cost” which was previously esti-
Alberta were eligible for inclusion. mated at 43.5% of the average cost [15]. All costs were converted to
2018 US dollars using the purchasing power parity exchange rate [16].
2.1. Patient outcome measures Mean TLOS was used for the cost analysis to allow for comparison
with most other recent publications on the subject that report in
Data on patient demographics and outcomes were entered into EIAS mean LOS and not median LOS [17].
as well as abstracted from the Alberta Discharge Abstract Database Ethics approval was obtained from the Research Ethics Boards of the
(DAD) for correlation and costing. The primary outcomes of this study University of Calgary and University of Alberta, Canada.
were acute length of stay (LOS) (number of days between surgical ad-
mission and discharge date), and total length of stay (TLOS). Total 3. Results
length of stay (TLOS) was calculated using both acute LOS and duration
of any readmissions (TLOS = acute length of stay + length of readmis- 3.1. Demographics
sion); these measures were calculated using EIAS and DAD, respectively.
Total length of stay was used for the cost-analysis to ensure that the cost Five hundred and nineteen patients were assessed between March
of readmissions was fairly represented. Secondary outcomes included 2016 and April 2017. There were 152 patients pre-ERAS and 367
complications, readmissions within 30 days of discharge, compliance patients post-ERAS implementation. Demographic, diagnostic, and
for each ERAS care element, and cost of implementation. surgical characteristics are presented in Table 2. The pre-ERAS and
post-ERAS groups were similar in age, smoking status, non-malignant
2.2. Sample size estimate comorbid disease, and ASA score (Table 2). Uterine cancer, ovarian
cancer, and benign ovarian mass were the most common diagnoses.
A confidence level of 0.05, and power of 80%, combined with an es- The final diagnosis of ovarian cancer was more common in the
timated TLOS of 6.5 days pre-ERAS and a common standard deviation of pre-ERAS cohort (39.5% versus 20.4%) and subsequently more
6 days estimated a sample size of 142 (per sample) could detect a differ- debulking surgery was represented in the pre-ERAS cohort, 50% versus
ence of 2.0 days between pre- and post-ERAS patients. 24% (p b 0.0001). All patients underwent open surgery in both cohorts
and median procedural complexity score was equal between cohorts
2.3. Data analysis (Median = 3.0 for both pre- and post-ERAS cohorts) (p = 0.1609).
There was no statistically significant difference in the distribution of
Wilcoxon tests were used to compare pre- and post-ERAS continu- low versus medium/high complexity between the pre-ERAS cohort
ous variables if their distribution was skewed; for normally distributed and the post-ERAS cohort (p = 0.1181). In the pre-ERAS cohort 94
continuous data, t-Test was employed. Chi-Square test, with post-hoc patients (61.8%) were classified as low surgical complexity and 58
Bonferroni correction were used for comparing nominal data. (38.2%) were classified as medium/high complexity. In the post-ERAS
Adjusted outcome data was analysed using a negative binomial re- cohort 253 patients (68.9%) were classified as low complexity while
gression to account for confounding factors. Data analysis was per- 114 (31.1%) were classified as medium/high complexity (Table 2).
formed using margins: Marginal Effects for Model Objects [13] software.
In this approach, LOS is treated as a count variable, and the coeffi-
cients are then expressed as the percentage increase (or decrease) in Table 2
Patient demographics and surgical characteristics.
LOS associated with each variable. Candidate confounding factors for
the models were age (≤50, 51–75, and ≥ 76), alcohol usage (Yes/No), Characteristics Pre-ERAS Post-ERAS p Value
smoker (Yes/No), BMI (≤24.9, 25–30, N30), Charlson Comorbidity (n = 152) (n = 367)
Index (0, 1, ≥2, with metastatic solid tumor malignancy removed), ma- Age (in years) 0.0743
lignancy (Yes/No, from EIAS), procedure complexity score (Low, Me- Median (IQR) 61.0 57.0
dium/High, from EIAS), and ASA score (1–2, 3–4). Variables were (52.0–68.0) (50.0–66.0)
BMI (Kg/m2), median (IQR) 27.5 29.4 0.0182
included in the model if the probability value for their association was (24.0–33.3) (24.8–35.1)
b0.25. Using stepwise elimination, variables with p N 0.25 at each step Tobacco usage, n (%) 19 (12.5) 57 (15.5) 0.3740
were removed and the remaining variables were reported. To assist in Charlson index levela 0.3983
the interpretation of the regression results, we used a marginal ap- =0 96 (63.2) 249 (67.9)
=1 42 (27.6) 81 (22.1)
proach to determine the average predicted reduction in for each patient
≥2 14 (9.2) 37 (10.1)
in the dataset. ASA group, n (%) 0.0685
Chi-square tests were used to compare the rate of having at least one ASA 1–2 103 (67.8) 275 (75.6)
complication during initial stay or in the 30-day immediate post- ASA 3–4 49 (32.2) 89 (24.6)
operative period between the pre-ERAS and post-ERAS cohorts. Main procedure, n (%) b0.0001
Debulking gynecology 76 (50) 88 (24)
Staging gynecology 76 (50) 279 (76)
2.4. Cost impact analysis Procedure complexity score class, n 0.1181
(%)
Using a hospital perspective, impact analysis was performed using Low 94 (61.8) 253 (68.9)
Medium/high 58 (38.2) 114 (31.1)
total LOS to determine differences in costs. Net cost impact accounted
Final diagnosis, n (%) b0.0001
for ERAS intervention costs of: licensing fees, additional nursing/coordi- Benign 44 (29.0) 176 (48.0)
nation costs, and maintenance of EIAS (which was estimated at $906 per Malignant 108 (71.1) 191 (52.0)
patient using AHS financial data). The difference in total LOS was esti- p-Values are from Chi-square test (for categorical variables) and Wilcoxon test (for contin-
mated as described above. Costs of inpatient hospital stay were esti- uous variables).
mated from DAD using the case mix group plus (CMG+) method [14]. a
Excluding metastatic solid tumor.

Please cite this article as: S.P. Bisch, et al., Enhanced Recovery After Surgery (ERAS) in gynecologic oncology: System-wide implementation and
audit leads to improved value and p..., Gynecol Oncol (2018), https://doi.org/10.1016/j.ygyno.2018.08.007
4 S.P. Bisch et al. / Gynecologic Oncology xxx (2018) xxx–xxx

There was no difference in the bowel resection rate between the pre- Table 4
ERAS (38.16%) and post-ERAS (31.82%) debulking cohorts (p = 0.3952). Patient outcomes.

Outcome Pre-ERAS Post-ERAS p

(n = 152) (n = 367) Valuec
3.2. Compliance
Acute length of stay
All surgery b0.0001
Twenty individual measures of compliance to ERAS guidelines were Median (IQR) 4.0 (2.0–6.0) 3.0 (2.0–4.0)
assessed pre- and post-ERAS implementation (Table 3). Mean compli- Mean (SD) 5.5 (5.9) 3.7 (3.8)
ance across all elements was 56% pre-implementation and improved Low complexity surgery 0.0002
to 77% post-implementation (p b 0.0001). Significant improvements Median (IQR) 3.0 (2.0–5.0) 2.0 (2.0–3.0)
High complexity surgery 0.0005
were seen in ERAS specific patient education (0.66% pre-ERAS to
Median (IQR) 5.0 (4.0–8.0) 3.0 (3.0–5.0)
93.46% post-ERAS, p b 0.0001) and pre-operative oral carbohydrate Total LOS (days)b b0.0001
treatment (5.26% pre-ERAS to 61.31% post-ERAS, p b 0.0001). After im- Median (IQR) 4.0 (2.0–7.0) 3.0 (2.0–4.0)
plementation, mean compliance was highest for pre-operative (85.7%) Mean (SD) 6.4 (7.5) 4.5 (5.9)
and intra-operative (80.0%) ERAS elements. Compliance in the post- Readmission LOS (days)b 0.7585
Mean (SD) 0.7 (3.0) 0.5 (2.7)
operative ERAS elements improved from 37.5% pre-ERAS to 62.5% Mortality/complications, n (%)
post-ERAS implementation (p b 0.0001). Mortality at dischargea 0 (0.0) 2 (0.5) 0.3618
Readmissions within 30 daysb 10 (6.6) 22 (6.0) 0.8011
Complication during primary stay 81 (53.3) 133 (36.2) 0.0003
3.3. Length of stay Complications after primary stay (Y/N) 56 (36.8) 127 (34.6) 0.6274

LOS – length of stay in hospital following initial operation (from EIAS).

There was a significant reduction overall in length of stay (LOS) be- a
Prior to discharge from initial/surgical admission.
tween pre-ERAS (median LOS = 4.0 days) and post-ERAS cohorts (me- b
From Discharge Abstract Database (DAD).
c
dian LOS = 3.0 days) (p b 0.0001) (Table 4). Low complexity surgery p Values by Wilcoxon test for non-normally distributed continuous variables, t-Test
for normally distributed continuous variables, and Chi-squared test for percentages.
alone demonstrated a significant reduction of median LOS from
3.0 days pre-ERAS to 2.0 days post-ERAS (p = 0.0002). Medium/high
complexity surgical cases demonstrated a reduction in LOS from
5.0 days pre-ERAS to 3.0 days post-ERAS (p = 0.0005) (Table 4).
After regression analysis for patient characteristics related to the 3.4. Analgesia, diet, fluid outcomes
outcome measure, the adjusted LOS following surgery in the post-
ERAS period was 31.4% shorter than in the pre-ERAS period (95% CI = Significant improvements were seen in other important clinical fac-
[21.7%–39.9%], p b 0.0001). This corresponds to an average marginal re- tors; median time to adequate pain control (date when pain control was
duction in acute LOS of 1.63 days (95% CI = [0.99–2.27]). The adjusted adequate on oral analgesics only) improved from 2.0 days (IQR =
reduction in total length of stay was 27.9% (95% CI = [16.0%–38.1%], p b 1.0–3.0) to 1.0 day (IQR = 1.0–2.0) following ERAS implementation
0.0001). (p b 0.0001). The time to tolerance of solid food (date when patient
was able to tolerate ≥50% solid foods consistently for 2 consecutive
meals and did not regress during the primary stay) improved from
Table 3 2.0 days (IQR = 1.0–4.0) pre-ERAS to 1.0 day (IQR = 1.0–2.0) post-
Compliance with ERAS-elements.
ERAS (p = 0.0406). Significant improvements were found in IV fluid ad-
ERAS compliance measure Pre-ERAS Post-ERAS p Value ministration following implementation in both intraoperative and post-
Total 55.43% 76.35% b0.0001 operative settings. Mean (±SD) intraoperative IV fluid volume
Pre Op Preadmission patient educationa 0.66% 93.46% b0.0001 decreased from 2136 mL (±1582 mL) to 1679 mL (±1192 mL) after
Avoidance of oral bowel preparation 95.39% 98.64% 0.0787 ERAS implementation (p b 0.0001). Mean IV fluid in the first day
Preoperative oral carbohydrate 5.26% 61.31% b0.0001
post-operatively decreased from 1965 mL (±746 mL) to 1645 mL
treatment
Avoidance of long-acting sedative 94.74% 94.82% 0.2035 (±740 mL) after ERAS implementation (p b 0.0001).
medication
Thrombosis prophylaxis 91.45% 91.01% 0.4242
Antibiotic prophylaxis before incision 92.76% 86.38% 0.1121 3.5. Complications
PONV prophylaxis administered 96.24% 96.79% 0.8743
Intra Op Avoidance of epidural or spinal 76.97% 85.83% 0.0219
anesthesia There was a significant decrease in complications during primary
Systemic opioids given 26.97% 23.16% 0.3002 stay from 53.3% pre-ERAS to 36.2% post-ERAS (p = 0.0003). This im-
Upper-body forced-air heating cover 96.05% 99.46% 0.0042 provement in complication rate remained significant after adjustment
used
for patient and surgical factors (p = 0.0292). There was no significant
Avoidance of nasogastric tube use 91.45% 98.37% 0.0001
Avoidance of resection-site drainage 93.42% 96.73% 0.0886 difference in mortality (p = 0.3618), readmission (p = 0.6159), or com-
Post Op Prompt termination of urinary 74.17% 89.23% b0.0001 plications after discharge up to 30 days (p = 0.6274) (Table 4). Table 5
drainageb demonstrates the breakdown of complications by type; there was a sig-
Stimulation of gut motility 3.95% 77.11% b0.0001 nificant difference in blood transfusion without hemorrhage (31.6% to
Patient weight recorded POD#1 N/A 51.23% N/A
15.5%, p b 0.0001), respiratory (11.2% to 5.2%, p = 0.0142), and cardio-
Prompt termination of intravenous 32.24% 58.31% b0.0001
fluid infusionc vascular (13.8% to 5.8%, p = 0.0008) complications between the pre-
Energy intake POD#0 0.00% 25.07% b0.0001 and post-ERAS cohorts. There was no significant difference in the rate
Energy intake on POD#1 0.00% 43.60% b0.0001 of anastomotic leak (0.7% pre-ERAS versus 0.3% post-ERAS, p =
Mobilization at all on day of surgery 50.00% 73.02% b0.0001
0.5190). There was a reduction in combined renal, hepatic, pancreatic,
30 day follow up performed 99.34% 99.18% 0.5372
and gastrointestinal complications (25% to 16.9%, p = 0.0331)
p-Values by Chi-squared test of pre versus post ERAS compliance. (Table 5). Among complications, there was no significant difference in
a
Patient education including written materials for home reference as specified by ERAS
protocols.
the rate of major complications (Clavien score [18] ≥3) between pre-
b
Removal of urinary catheter earlier than 24 h post-operatively. ERAS and post-ERAS cohorts during primary stay (p = 0.9735) or up
c
Discontinuation of intravenous fluids earlier than 24 h post-operatively. to 30 days postoperatively (p = 0.5194).

Please cite this article as: S.P. Bisch, et al., Enhanced Recovery After Surgery (ERAS) in gynecologic oncology: System-wide implementation and
audit leads to improved value and p..., Gynecol Oncol (2018), https://doi.org/10.1016/j.ygyno.2018.08.007
 S.P. Bisch et al. / Gynecologic Oncology xxx (2018) xxx–xxx 5

Table 5 number of steps, frequency of ambulation [21], or simply ambulating

Proportion of patients who developed at least one complication by complication group. to a communal meal with other patients [22] remains to be seen.
Pre-ERAS Post-ERAS Diff. p Value Through the implementation of ERAS care elements in gynecologic
n (%) n (%) (%) a b
oncology we were able to effect a significant reduction in post-
All complications 77 (50.66) 118 (32.15) 18.51 b0.0001 operative LOS as well as inpatient complications without an increase
Respiratory complications 17 (11.18) 19 (5.18) 6.00 0.0142 in readmission rate or outpatient complications. The absolute reduction
Infectious complications 13 (8.55) 31 (8.45) 0.10 0.9686 in LOS was even more pronounced when looking at medium and high
Cardiovascular 21 (13.82) 19 (5.81) 8.01 0.0008
complexity surgery.
RHPG complicationsc 38 (25.00) 62 (16.89) 8.11 0.0331
Anastomotic leak 1 (0.7) 1 (0.3) 0.4 0.5190 A recent study by Bergstrom et al. demonstrated no improvement in
Transfusion without hemorrhage 48 (31.58) 57 (15.53) 16.05 b0.0001 LOS post ERAS implementation but noted a significant decrease in opi-
Epidural or spinal anesthesia 7 (4.61) 9 (2.45) 2.16 0.1966 oid use in gynecologic oncology patients [9]. There are some important
related complications differences between our study and theirs that may account for the dif-
Anesthetic complications 4 (2.63) 4 (1.09) 1.54 0.1945
Psychiatric complications 0 (0.00) 2 (0.54) −0.54 0.3618
ference in results; by utilizing a structured implementation program
a
and EIAS together with biweekly compliance audits we have ensured
Difference in proportions (positive values denote reduced complication in the post-
that across a health system we were able to improve compliance in
ERAS cohort).
b
p Values are from Chi-square test. over 50% of the ERAS protocol elements. Bergstrom et al. also noted a
c
RHPG means Renal, hepatic, pancreatic and gastrointestinal. significant improvement in LOS in a subgroup with higher compliance
with ERAS elements. In contrast to Bergstrom et al. we define ERAS
compliance as avoiding epidural analgesia. Epidural analgesia has
3.6. Costs been independently shown to increase LOS [20] and effective alterna-
tives for analgesia exist [23]. The post-ERAS cohort described in
The mean TLOS per patient pre-ERAS implementation was 6.4 days Bergstrom's study had greater use of epidural than the pre-ERAS cohort
(Table 4); this was reduced by 27.9% (95%CI: 16.0% to 38.1%) or 1.79 (86.7% vs. 57.6%) and this may have attenuated improvements in LOS.
(95%CI: 0.83 to 2.93) days post-ERAS implementation after adjustment By comparison, our results are similar to a study by Modesitt et al. in
for confounders. Multiplying the LOS values with the average cost per which a significant improvement in length of stay from 3.0 days to
day ($2397) and with the percentage of the “hotel cost” (43.5%), the 2.0 days was demonstrated in major gynecology surgery following im-
gross cost-saving per patient was estimated at $1862 (95%CI: $1067 to plementation of ERAS care elements [24]. Notably, during the Modesitt
$2542). Deducting the cost of ERAS ($906 per patient) from the gross study, the rate of epidural use decreased (24.6% pre-ERAS to 4.4% post-
cost-saving, the net cost-saving per patient was estimated at $956 ERAS) and the rate of spinal analgesia increased significantly (12.3%
(95%CI: $162 to $1636). For the whole cohort (367 patients), the net pre-ERAS to 86.8% post-ERAS) following ERAS implementation [24].
cost-saving of ERAS was $350,784 (95%CI: $59,298 to $600,477). Divid- Despite having a relatively low compliance with early oral feeding
ing the gross cost-saving by the cost of ERAS, the return-on-investment overall, we demonstrate a dramatic change in practice towards early en-
ratio was estimated at 2.1 (95%CI: 1.2 to 2.8). In other words, every $1 ergy intake, which was previously not the standard of practice for al-
invested would bring $2.1 in return. most all patients in the pre-ERAS cohort. Early energy intake was
achieved by the use of oral nutritional supplements (ONS)(e.g. Ensure
Plus, Nutren 2.0) in addition to hospital meals. Although achieving
4. Discussion and maintaining adequate nutritional status is important in the periop-
erative setting, the use and amount of oral nutritional supplements in
Systematic implementation of the ERAS gynecologic oncology ERAS has not been demonstrated to independently predict outcome, es-
guidelines using the ERAS Interactive Audit System (EIAS) across a pecially given that the majority of patients are able to tolerate a full diet
large healthcare system improved compliance with ERAS care elements. earlier than previously seen (1.0 day versus 2.0 days). We recognize that
Even modest improvements in compliance (approximately 20% in- compliance with ONS intake is low but this may simply represent tran-
crease) with ERAS care elements resulted in significant improvements sition to a regular diet with lower consumption of these products. In the
in: length of stay, time to adequate pain control, time to tolerance of study by Martin et al., time to tolerating diet was independently associ-
solid foods, goal-directed fluid administration, inpatient complication ated with length of stay [25].
rates, and resource utilization. The demonstrated improvement in post-operative cardiovascular
Behavioural and systematic change towards compliance was facili- and respiratory complications are in keeping with previous studies
tated through a systems approach enabled by provincial leadership [26,27] and are likely attributable to avoidance of nasogastric tubes,
and local interdisciplinary teams [19]. Regular and thorough auditing avoidance of long acting sedatives, early ambulation, and judicious use
of compliance was possible through the use of EIAS and is a crucial of IV fluid. Myles et al. recently published an RCT purporting evidence
part of changing entrenched behaviour [19]. Despite demonstrated im- against fluid restriction in the setting of major abdominal surgery citing
provements in compliance, there were notable areas where compliance concerns around hypoperfusion and acute kidney injury [28]. Our post-
lagged behind in our study; energy intake post-operatively (use of oral ERAS cohort demonstrated lower mean administered volume of IV fluid
nutritional supplements), and avoidance of systemic opioids all demon- than the restrictive group in Myles' study but still showed improve-
strated b50% compliance post implementation. Given that clinician acu- ments in complication rates, including renal complications. Myles
men is still an essential part of ERAS programs, it is more likely that et al. reported that less than half of the patients in their trial were
compliance with these particular elements was not advisable for each treated according to ERAS principles [28]. By avoiding pre-operative de-
patient rather than dissent on the part of the treating clinician. It is hydration and by encouraging early oral intake, ERAS patients are at less
not surprising, therefore, that no defined goals for compliance for indi- risk of iatrogenic hypovolemia and require less intravenous supplemen-
vidual ERAS elements have been set. Associations have been made tation. It is also possible that the improvements in stress response
which demonstrate improved patient outcomes with increased compli- achieved by ERAS care elements [4] promote maintenance of intravas-
ance [20] however, compliance itself can also be a surrogate marker for cular volume without the need for excessive IV fluids.
patient wellbeing. As ERAS becomes standard of care across surgery, fu- Despite being cost-saving and even more effective than ERAS for co-
ture research will be necessary to determine evidence-based goals lorectal surgery (LOS reduction per patient: 1.79 vs. 1.54 days) [10,15],
within each care element rather than overarching ideals. Whether the the gynecologic oncology ERAS implementation is less cost-effective
standard of care for “early ambulation” is characterized by a specific with an ROI ratio of 2.1 compared to an ROI ratio of 3.8 [15]. The

Please cite this article as: S.P. Bisch, et al., Enhanced Recovery After Surgery (ERAS) in gynecologic oncology: System-wide implementation and
audit leads to improved value and p..., Gynecol Oncol (2018), https://doi.org/10.1016/j.ygyno.2018.08.007
6 S.P. Bisch et al. / Gynecologic Oncology xxx (2018) xxx–xxx
explanation is that compared to colorectal surgery, the average cost per
hospital day for gynecologic oncology patients was lower ($2397 vs.
$3003) but the intervention cost per patient was higher ($906 vs.
$533). Given that the number of colorectal surgeries was larger (1295
patients) [15], this would emphasize the importance of “allocative effi-
ciency” (where to invest) and “economies of scale” in health economics
[17].
Our results are similar to those demonstrated by Kalogera et al. who
studied ERAS care elements in urogynecology and gynecologic oncology
patient populations in the United States. The authors found a significant
decrease in length of stay of 4 days with no change in overall complica-
tion rate, but did find more nausea and vomiting in the ERAS cohort. Our
study did not reach the same absolute reduction in length of stay
(1.8 day versus 4.0 days) or cost-savings ($956/patient versus $7600/
patient) [8]. One explanation for this discrepancy is the difference in
unit cost between the 2 studies and their respective health systems;
the mean cost of hospital stay in the United States at the time of
Kalogera's study was $21,200 [29]. More importantly, given pre-ERAS
LOS is 12 days in the US study [8] compared to 6.4 days in ours, another
explanation is that our patients started closer to the absolute limit on
minimum LOS, leading to a smaller possible reduction that the interven-
tion could effect. Other studies involving general gynecology with
smaller starting length of stay demonstrate this concept by showing rel-
atively lower cost-savings of $149.03 [30], and $480.40 per patient
[31,32].
This study has several notable strengths. The system-wide imple-
mentation provided a diverse population across the breadth of gyneco-
logic oncology surgeries to improve validity and generalizability. The
inclusion of real-time auditing in a systematic fashion via the EIAS im-
proved compliance and maintained high-quality data. The EIAS is com-
mercially available and can be implemented in a variety of clinical
environments and has been used in a number of centers worldwide.
As a research tool, EIAS naturally facilitates standardization of results
among ERAS surgical patients on a global level thereby facilitating col-
laborative international research.
By utilizing DAD in addition to EIAS we were able to ensure the accu-
racy of EIAS data for LOS and were able to capture readmissions across
the province if a patient presented to a center other than the treating
surgical hospital. Finally, by employing experienced peri-operative
nurse clinicians to collect the data ensured accurate and appropriate
collection of clinically relevant factors.
A limitation of EIAS is that it is labour intensive and currently re-
quires manual entry by trained practitioners; the cost of this may be
prohibitive for some institutions. The use of automation and informatics
technologies, like natural language processing, would dramatically de-
crease the human resources required for such an audit system [33],
and warrants future research in the era of value-based medicine.
Another limitation of this study is the non-randomized historical co-
hort utilized in the study. Given the breadth of evidence for ERAS ele-
ments and lack of evidence for the previous standard of care, it has
been suggested that randomized trials of ERAS would not be ethically
feasible if they withheld such care in a control arm [19]. In order to en-
sure adequate representation of debulking surgeries prior to ERAS im-
plementation, we sequentially included an equal proportion of both
staging and debulking surgeries. All patients undergoing gynecologic
oncology surgery were included in the post-ERAS cohort and this re-
sulted in a different ratio of staging to debulking surgeries between
the cohorts. Since not all staging procedures result in a final diagnosis
of malignancy this ratio resulted in a greater proportion of patients
who had malignancy in the pre-implementation group. Malignancy
can be associated with greater surgical complexity and perioperative
complications including increased rates of venous thromboembolism
[34]. It is worthwhile to note, however, that surgical complexity did
not differ between the cohorts and the improvement in length of stay
remained statistically significant after adjusting for various potential
confounders, including diagnosis of malignancy.
Please cite this article as: S.P. Bisch, et al., Enhanced Recovery After Surgery (ERAS) in gynecologic oncology: System-wide implementation and
audit leads to improved value and p..., Gynecol Oncol (2018), https://doi.org/10.1016/j.ygyno.2018.08.007
5. Conclusions
Systematic implementation of ERAS gynecologic oncology guide-
lines across a large healthcare system improves patient outcomes and
saves resources. In our study we demonstrate that even modest im-
provements in compliance with ERAS protocols yields significant im-
provements in LOS and inpatient complication rates without
increasing readmission rates or outpatient complications. This imple-
mentation resulted in more than $350,000 in savings in a surgical sub-
specialty on a system-wide level, without factoring indirect-cost
savings associated with early recovery [22]. When applied to multiple
healthcare systems globally the results would be substantial.
We provide a framework and toolset to implement ERAS gyneco-
logic oncology guidelines on a large scale and emulate results demon-
strated by single center trials. With further improvements in
compliance gleaned from rigorous auditing we expect to improve com-
pliance in post-operative interventions and, subsequently, patient
outcomes.
By demonstrating not only improvements in patient outcomes as
well as cost-efficacy this study provides a strong impetus for systematic
change in the practice of perioperative gynecologic oncology.
Conflict of interest statement
None of the authors have any financial disclosures to report. Dr. Nelson is the Secretary of
the ERAS® Society.
Author contributions
Drs. Bisch, Wells, Gramlich, Faris, Wang, Tran, Thanh, Glaze, Chu,
Ghatage, Nation, Capstick, Steed, Sabourin, and Nelson were involved
in all aspects of the manuscript design and each contributed to individ-
ual sections as well as edited the final draft.
References
[1] H. Kehlet, Fast-track colorectal surgery program reduces hospital length of stay, Lan-
cet 371 (2008) 791–793, https://doi.org/10.1097/NUR.0b013e3181e3604c.
[2] G. Nelson, A.D. Altman, A. Nick, L.A. Meyer, P.T. Ramirez, C. Achtari, et al., Guidelines
for pre- and intra-operative care in gynecologic/oncology surgery: Enhanced Recov-
ery after Surgery (ERAS®) Society recommendations - Part i, Gynecol. Oncol. 140
(2016) 313–322, https://doi.org/10.1016/j.ygyno.2015.11.015.
[3] G. Nelson, A.D. Altman, A. Nick, L.A. Meyer, P.T. Ramirez, C. Achtari, et al., Guidelines
for postoperative care in gynecologic/oncology surgery: Enhanced Recovery after
Surgery (ERAS®) Society recommendations - Part II, Gynecol. Oncol. 140 (2016)
323–332, https://doi.org/10.1016/j.ygyno.2015.12.019.
[4] L. Ren, D. Zhu, Y. Wei, X. Pan, L. Liang, J. Xu, et al., Enhanced Recovery after Surgery
(ERAS) program attenuates stress and accelerates recovery in patients after radical
resection for colorectal cancer: a prospective randomized controlled trial, World J.
Surg. 36 (2012) 407–414, https://doi.org/10.1007/s00268-011-1348-4.
[5] O. Ljungqvist, M. Scott, K.C. Fearon, Enhanced recovery after surgery a review, JAMA
Surg. 152 (2017) 292–298, https://doi.org/10.1001/jamasurg.2016.4952.
[6] J. Ferlay, I. Soerjomataram, R. Dikshit, S. Eser, C. Mathers, M. Rebelo, et al., Cancer in-
cidence and mortality worldwide: sources, methods and major patterns in
GLOBOCAN 2012, Int. J. Cancer 136 (2015) E359–E386, https://doi.org/10.1002/ijc.
29210.
[7] G. Nelson, S.C. Dowdy, J. Lasala, G. Mena, J. Bakkum-Gamez, L.A. Meyer, et al., En-
hanced recovery after surgery (ERAS®) in gynecologic oncology – practical consid-
erations for program development, Gynecol. Oncol. 147 (2017) 617–620, https://
doi.org/10.1016/j.ygyno.2017.09.023.
[8] E. Kalogera, J.N. Bakkum-Gamez, C.J. Jankowski, E. Trabuco, J.K. Lovely, S. Dhanorker,
et al., Enhanced recovery in gynecologic surgery, Obstet. Gynecol. 122 (2013)
319–328, https://doi.org/10.1097/AOG.0b013e31829aa780.
[9] J.E. Bergstrom, M.E. Scott, Y. Alimi, T.T. Yen, D. Hobson, K.K. Machado, et al., Narcotics
reduction, quality and safety in gynecologic oncology surgery in the first year of en-
hanced recovery after surgery protocol implementation, Gynecol. Oncol. (2018)
https://doi.org/10.1016/j.ygyno.2018.04.003.
[10] G. Nelson, L.N. Kiyang, E.T. Crumley, A. Chuck, T. Nguyen, P. Faris, et al., Implemen-
tation of Enhanced Recovery after Surgery (ERAS) across a provincial healthcare sys-
tem: the ERAS Alberta colorectal surgery experience, World J. Surg. 40 (2016)
1092–1103, https://doi.org/10.1007/s00268-016-3472-7.
[11] G.D. Aletti, A. Santillan, E.L. Eisenhauer, J. Hu, G. Aletti, K.C. Podratz, et al., A new
frontier for quality of care in gynecologic oncology surgery: multi-institutional as-
sessment of short-term outcomes for ovarian cancer using a risk-adjusted model,
Gynecol. Oncol. 107 (2007) 99–106, https://doi.org/10.1016/j.ygyno.2007.05.032.
[12] M. Charlson, T. Szatrowski, J. Peterson, J. Gold, Validation of a combined comorbidity
index, J. Clin. Epidemiol. 47 (1994) 1245–1251.
 S.P. Bisch et al. / Gynecologic Oncology xxx (2018) xxx–xxx
[13] T.J. Leeper, Margins: Marginal Effects for Model Objects, 2018.
[14] Canadian Institute for Health Information (CIHI), CMG+, n.d. https://www.cihi.ca/
en/cmg, Accessed date: 11 June 2018.
[15] N.X. Thanh, A.W. Chuck, T. Wasylak, J. Lawrence, P. Faris, O. Ljungqvist, et al., An eco-
nomic evaluation of the Enhanced Recovery After Surgery (ERAS) multisite imple-
mentation program for colorectal surgery in Alberta, Can. J. Surg. 59 (2016)
415–421, https://doi.org/10.1503/cjs.006716.
[16] Organisation for Economic Co-operation and Development, Purchasing Power Pari-
ties 2015, http://stats.oecd.org/Index.aspx?datasetcode=SNA_TABLE4 2015,
Accessed date: 18 June 2018.
[17] A.B. Stone, M.C. Grant, C. Pio Roda, D. Hobson, T. Pawlik, C.L. Wu, et al., Implementa-
tion costs of an enhanced recovery after surgery program in the United States: a fi-
nancial model and sensitivity analysis based on experiences at a quaternary
academic medical center, J. Am. Coll. Surg. 222 (2016) 219–225, https://doi.org/
10.1016/j.jamcollsurg.2015.11.021.
[18] D. Dindo, N. Demartines, P.A. Clavien, Classification of surgical complications: a new
proposal with evaluation in a cohort of 6336 patients and results of a survey, Ann.
Surg. 240 (2004) 205–213, https://doi.org/10.1097/01.sla.0000133083.54934.ae.
[19] J. Maessen, C.H.C. Dejong, J. Hausel, J. Nygren, K. Lassen, J. Andersen, et al., A protocol
is not enough to implement an enhanced recovery programme for colorectal resec-
tion, Br. J. Surg. 94 (2007) 224–231, https://doi.org/10.1002/bjs.5468.
[20] A. Currie, J. Burch, J.T. Jenkins, O. Faiz, R.H. Kennedy, O. Ljungqvist, et al., The impact
of enhanced recovery protocol compliance on elective colorectal cancer resection:
results from an international registry, Ann. Surg. 261 (2015) 1153–1159, https://
doi.org/10.1097/SLA.0000000000001029.
[21] C. Thorup Msn, J. Hansen, M. Grønkjær, J.J. Andreasen, G. Nielsen, E.E. Sørensen,
et al., Cardiac patients' walking activity determined by a step counter in cardiac
telerehabilitation: Data from the intervention arm of a randomized controlled
trial, J. Med. Internet Res. (2016) https://doi.org/10.2196/jmir.5191.
[22] D. Gordon, A. Malhas, A. Goubran, P. Subramanian, C. Messer, D. Houlihan-Burne,
Implementing the rapid recovery program in primary hip and knee arthroplasty
in a UK state run hospital, Eur. J. Orthop. Surg. Traumatol. 21 (2011) 151–158,
https://doi.org/10.1007/s00590-010-0690-9.
[23] E. Kalogera, J.N. Bakkum-Gamez, A.L. Weaver, J.P. Moriarty, B.J. Borah, C.L. Langstraat,
et al., Abdominal incision injection of liposomal bupivacaine and opioid use after
laparotomy for gynecologic malignancies, Obstet. Gynecol. 128 (2016) 1009–1017,
https://doi.org/10.1097/AOG.0000000000001719.
Please cite this article as: S.P. Bisch, et al., Enhanced Recovery After Surgery (ERAS) in gynecologic oncology: System-wide implementation and
audit leads to improved value and p..., Gynecol Oncol (2018), https://doi.org/10.1016/j.ygyno.2018.08.007
7
[24] S.C. Modesitt, B.M. Sarosiek, E.R. Trowbridge, D.L. Redick, P.M. Shah, R.H. Thiele,
et al., Enhanced recovery implementation in major gynecologic surgeries: effect of
care standardization, Obstet. Gynecol. 128 (2016) 457–466, https://doi.org/10.
1097/AOG.0000000000001555.
[25] L. Martin, C. Gillis, M. Atkins, M. Gillam, C. Sheppard, S. Buhler, et al., Implementation
of an Enhanced Recovery After Surgery program can change nutrition care practice:
a multi-center experience in elective colorectal surgery, JPEN (2018) https://doi.org/
10.1002/jpen.1417 [Epub ahead of print].
[26] P.H.E. Teeuwen, R.P. Bleichrodt, C. Strik, J.J.M. Groenewoud, W. Brinkert, C.J.H.M. van
Laarhoven, et al., Enhanced Recovery After Surgery (ERAS) Versus conventional
postoperative care in colorectal surgery, J. Gastrointest. Surg. 14 (2010) 88–95,
https://doi.org/10.1007/s11605-009-1037-x.
[27] M. Greco, G. Capretti, L. Beretta, M. Gemma, N. Pecorelli, M. Braga, Enhanced recov-
ery program in colorectal surgery: a meta-analysis of randomized controlled trials,
World J. Surg. 38 (2014) 1531–1541, https://doi.org/10.1007/s00268-013-2416-8.
[28] P.S. Myles, R. Bellomo, T. Corcoran, A. Forbes, P. Peyton, D. Story, et al., Restrictive
versus liberal fluid therapy for major abdominal surgery, N. Engl. J. Med. (2018)
1–12, https://doi.org/10.1056/NEJMoa1801601.
[29] B. Moore, K. Levit, A. Elixhauser, Costs for Hospital Stays in the United States, 2012,
2014 (doi:NBK116744 [bookaccession]).
[30] S. Relph, A. Bell, V. Sivashanmugarajan, K. Munro, K. Chigwidden, S. Lloyd, et al., Cost
effectiveness of enhanced recovery after surgery programme for vaginal hysterec-
tomy: a comparison of pre and post-implementation expenditures, Int. J. Health
Plann. Manag. 29 (2014) 399–406.
[31] L. Narang, S. Mitchelmore, H. Byrne, Cost reduction and enhanced patient experi-
ence following the introduction of enhanced recovery programme in gynaecological
surgery, BJOG Int. J. Obstet. Gynaecol. 120 (2013) 439.
[32] M.D.J. Stowers, D.P. Lemanu, A.G. Hill, Health economics in Enhanced Recovery After
Surgery programs, Can. J. Anesth. Can d'anesthésie 62 (2015) 219–230, https://doi.
org/10.1007/s12630-014-0272-0.
[33] D.W. Bates, R.S. Evans, H. Murff, P.D. Stetson, L. Pizziferri, G. Hripcsak, Detecting ad-
verse events using information technology, J. Am. Med. Inform. Assoc. 10 (2003)
115–128, https://doi.org/10.1197/jamia.M1074.supported.
[34] S. Tateo, L. Mereu, S. Salamano, C. Klersy, M. Barone, A.C. Spyropoulos, et al., Ovarian
cancer and venous thromboembolic risk, Gynecol. Oncol. 99 (2005) 119–125,
https://doi.org/10.1016/j.ygyno.2005.05.009.