Smith et al.

Perioperative Medicine (2019) 8:4

https://doi.org/10.1186/s13741-019-0114-2

RESEARCH Open Access

Enhanced recovery after surgery (ERAS)

program for lumbar spine fusion
Justin Smith1* , Stephen Probst1, Colleen Calandra2, Raphael Davis2, Kentaro Sugimoto1, Lizhou Nie3,
Tong J. Gan1 and Elliott Bennett-Guerrero1

Abstract
Background: There is a paucity of literature regarding the implementation of enhanced recovery after surgery (ERAS)
protocols for open lumbar spine fusions. We implemented an ERAS program for 1–2-level lumbar spine fusion surgery
and identified areas that might benefit from perioperative interventions to improve patient satisfaction and outcomes.
Methods: This institutionally approved quality improvement (QI) ERAS program for lumbar spine fusion was designed
for all neurosurgical patients 18 years and older scheduled for 1 or 2 level primary lumbar fusions. The ERAS bundle
contained elements such as multimodal analgesia including preoperative oral acetaminophen and gabapentin,
postoperative early mobilization and physical therapy, and a prophylactic multimodal antiemetic regimen to decrease
postoperative nausea and vomiting. No fluid management or hemodynamic parameters were included. Pre-ERAS and
post-ERAS data were compared with regard to potential confounders, compliance with the ERAS bundle, and
postoperative outcomes.
Results: A total of 230 patients were included from October 2013 to May 2017. The pre-ERAS phase consisted of 123
patients, 11 patients during the transition period, and 96 serving as post-ERAS patients. The pre-ERAS and post-ERAS
groups had comparable demographics and comorbidities. Compliance with preoperative and intraoperative
medication interventions was relatively good (~ 80%). Compliance with postoperative elements such as early physical
therapy, early mobilization, and early removal of the urinary catheter was poor with no significant improvement in
post-ERAS patients. There was no significant change in the amount of short-acting opioids used, but there was a
decrease in the use of long-acting opioids in the post-ERAS phase (14.6 to 5.2%, p = 0.025). Post-ERAS patients required
fewer rescue antiemetic medications in the recovery room compared to pre-ERAS patients (40 to 24%). There was no
significant difference in postoperative pain scores or hospital length of stay between the two groups.
Conclusions: Implementing an ERAS bundle for 1–2-level lumbar fusion had minimal effect in decreasing length of
stay, but a significant decrease in postoperative opioid and rescue antiemetic use. This ERAS bundle showed mixed
results likely secondary to poor ERAS protocol compliance. Going forward, this QI project will look to improve post-
operative ERAS implementation to improve patient outcomes.
Keywords: ERAS, Lumbar fusion, QI spine, Spine surgery

* Correspondence: justinksmithmd@gmail.com
1
Department of Anesthesiology, Stony Brook University Medical Center, 101
Nicolls Rd, Stony Brook, NY 11794, USA
Full list of author information is available at the end of the article

© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Smith et al. Perioperative Medicine (2019) 8:4 Page 2 of 9

Introduction care bundle for all lumbar fusion patients. Table 1 out-
Back pain and spinal disorders are one of the most com- lines the protocol.
monly encountered medical problems facing the health- Among the gaps of the existing care protocol, patient
care system. Approximately two thirds of the population communication was observed to be an important issue
will suffer from low back pain (LBP) at some point in to address since some patients were receiving inconsist-
their lifetime, and it is estimated that the USA spends ent recommendations concerning medications to be
over $100 billion annually in direct and indirect costs re- taken on the day of surgery, and expectations following
lated to LBP (Deyo and Weinstein 2001; Dagenais et al. surgery varied between patients. It was also found that
2008). A lumbar spinal fusion may improve LBP and patients with chronic pain were not identified prior to
help many improve their quality of life. While outcomes surgery, and there was a need for better pain manage-
after spinal fusion are generally good, many patients ex- ment postoperatively for all patients. Earlier involvement
perience adverse events such as superficial and deep of physical therapy and social work was another area for
wound infections, deep vein thrombosis, pseudarthrosis, improvement to decrease delays in mobilization and dis-
urinary tract infections, transient ischemic attacks, and charge. To address these issues, our goals included: im-
continued pain following surgery (Proietti et al. 2013). proving preoperative patient education, decreasing case
In recent years, “fast track” surgery or enhanced recovery cancelations, decreasing hospital LOS, decreasing
bundles have been developed in many surgical specialties to PONV, decreasing postoperative pain, and decreasing
decrease hospital length of stay (LOS) and decrease peri- postoperative opioid use.
operative morbidity. Kehlet first introduced the enhanced To improve patient communication, a standardized
recovery model in 1997 as a multimodal, evidence-based education packet was given to patients in the neurosur-
plan to improve patient care in the perioperative period gical clinic prior to surgery. This included information
(Kehlet 1997). Since then, many ERAS strategies have about the surgery, expectations, support services, man-
shown the effectiveness of enhanced recovery bundles in agement of diabetes, and smoking cessation among
improving patient outcomes. These ERAS bundles have other things. To decrease case cancelations, the neuro-
been used successfully in colorectal surgery, radical cystec- surgery office administrative staff had an increased role
tomies, major pelvic surgery, and pancreaticoduodenec- in communicating with the surgeon and operating room
tomies to name a few (Gustafsson et al. 2013; Daneshmand staff to ensure proper scheduling of cases and to notify
et al. 2014; Nygren et al. 2013; Lassen et al. 2013). To date, care providers of the need to adhere to the ERAS proto-
there have been very few reports of the implementation of col. At preoperative services, education was reinforced
ERAS bundles that focused on improving patient outcomes and patients underwent a laboratory workup, history
in open lumbar fusions and spinal surgery (Wainwright et and physical prior to surgery. As a means to decrease
al. 2016; Blackburn et al. 2016). In this study, we hypothe- PONV, all patients were scheduled to receive dexa-
sized that the ERAS protocol would decrease: case cancel- methasone 8 mg IV after induction of anesthesia, and for
ations, postoperative nausea and vomiting (PONV), length patients with more than 2 risk factors (Apfel et al. 2012),
of stay, postoperative pain, and postoperative narcotic use. oral aprepitant 40 mg was added in the preoperative
holding area. Patients were routinely given intraoperative
Methods ondansetron 4 mg IV for PONV prophylaxis. The proto-
ERAS protocol development and implementation col did not include intraoperative fluid or hemodynamic
This ERAS QI protocol was implemented at Stony parameters. To decrease postoperative pain and opioid
Brook University Hospital in Stony Brook, New York, use, a multimodal analgesia regimen was included
after receiving institutional approval. The program in- (Mathiesen et al. 2013). On the day of surgery, the pa-
cluded neurosurgical patients 18 years and older planned tient was reassessed by the anesthesia care provider and
to undergo a 1 or 2 level lumbar spinal fusion. This QI given acetaminophen 975 mg PO and gabapentin 900 mg
project excluded patients who were pregnant, age < 18, PO in the preoperative holding area. All patients were
or planned for a revision of a previous fusion. given an intake form during their initial consultation
The departments of neuroanesthesia and neurosurgery with the neurosurgeon. Patients who marked that they
at our institution worked together to review neurosur- were “on chronic and current benzodiazepines or opi-
gery spinal fusion cases to identify interventions that oids” on this form were considered high risk for pain.
would address unmet postoperative patient care goals. The acute pain service was to be made aware of all
The protocol was divided into preoperative, intraopera- high-risk pain patients, and they were to receive keta-
tive, and postoperative interventions. While some of the mine 30 mg IV with the induction of anesthesia (Loftus
bundle elements were already common practice for lum- et al. 2010). There were no specific guidelines for intra-
bar fusion procedures at our institution and were in- operative opioid use. Prior to leaving the recovery room
cluded in the ERAS protocol, there was no standardized a fentanyl, morphine, or hydromorphone, PCA was
Smith et al. Perioperative Medicine (2019) 8:4 Page 3 of 9

Table 1 Lumbar spinal fusion ERAS protocol Table 1 Lumbar spinal fusion ERAS protocol (Continued)
Stage Location Action Stage Location Action
Preoperative Neurosurgical ▪ 1–2-level lumbar fusion patients ▪ Acetaminophen 975 mg PO, gabapentin
clinic visit identified to be included in the project 900 mg PO prior to OR
▪ Neurosurgery Spine Booking Checklist ▪ High-risk patients administer 40 mg
completed and surgery scheduled with aprepitant for PONV
the comment “lumbar fusion ERP”
OR for surgery ▪ Antibiotics will be dosed and given
▪ Patient given information letter and less than 1 h prior to incision and re-
materials including diabetes education dosed as appropriate
and smoking cessation
▪ All patients regardless of diabetic status
▪ Patients received a “pain will receive dexamethasone 8 mg IVP
questionnaire” and if any patient selects
“on chronic and current ▪ High-risk patients with known chronic
pain may receive ketamine 30 mg with
benzodiazepines or opioids,” they will
induction
be considered high risk for pain and the
acute pain service will be made aware ▪ Follow intra-op protocol as prescribed
Booking office ▪ The following appointments are including hourly attending anesthesia
made: to neurosurgeon communication
§ Should include progress of surgery,
§ Preoperative services
fluid status, hemodynamics, pressure
§ OR for surgery point evaluations, EBL
§ Postoperative wound check (14 PACU ▪ Assess patient temperature
days) and surgical (30 days) follow-up
▪ All patients receive PCA and
▪ A surgical packet will be sent to the methocarbamol 1500 mg PO or IV
patient with the following:
▪ Pain liaison will visit the patient if
§ Cover letter explaining the contents identified as “high-risk pain”
and what to expect at POS and PSA
Postoperative Floor ▪ Nursing staff will notify neurosurgery if
§ Instructions for taking or stopping there are any medically necessary
medications deviations from the protocol.
§ Directions to pre-operative services ▪ PT/SW/NS to meet Monday–Friday to
and ambulatory surgery unit review patients’ discharge progress
§ Postoperative discharge instructions POD#0 ▪ All patients will receive stool softeners
and laxatives delineated in the power
§ Social support services information
plan
§ Discharge needs assessment
▪ Diet as tolerated
form—to be returned pre-operatively
▪ Continue IVF’s until tolerating good
§ Pre-operative pain questionnaire-to
oral intake
be returned pre-operatively
▪ Reinforce incentive spirometry
▪ Assistant will add the lumbar fusion
ERP checklist to the surgical packet that POD#1 ▪ Discontinue Foley catheter at 0600
is sent to pre-operative services
▪ Celecoxib 200 mg Q12H PO, gabapentin
Preoperative ▪ History and physical-required 300 mg Q8H PO, and acetaminophen
services 975 mg Q6H PO to be continued for
▪ Anesthesia consult-required and to be
1 week
performed by an anesthesiologist,
qualified MD, or physician extender ▪ Acute pain assessment for the
transition to oral medications
▪ Required testing includes T&S, CBC,
PT/PTT, INR, UA, and for diagnosed ▪ Mobilize out of bed and reinforce
diabetic patients HgbA1C incentive spirometry
▪ HgbA1c of ≥ 9 will postpone the ▪ PT evaluation for rehabilitation needs
surgical date by at least 2 months and
will be re-evaluated prior to rebooking ▪ SW/Case management evaluation for
support services and discharge planning
▪ Incentive spirometry, OSA and CPAP
education, NPRS education POD#2 ▪ If appropriate discontinue surgical
drain and continue to mobilize
▪ Pre-operative antibiotic ordered—
▪ Discharge if appropriate
Ancef 2 g (3 g if > 120 kg), Clindamycin
900 mg, or Vancomycin 15 mg/kg Follow Up Neurosurgical ▪ Follow-up phone call post-discharge
Clinic Visit day 1
Perioperative Ambulatory ▪ If identified as a “high-risk pain” the
surgery unit “pain liaison” will visit the patient prior ▪ Wound check visit 2 weeks after
to OR discharge
Smith et al. Perioperative Medicine (2019) 8:4 Page 4 of 9

Table 1 Lumbar spinal fusion ERAS protocol (Continued) cases as a measure to track case cancelations. Length of
Stage Location Action stay was determined based on a patient’s admission time
▪ Regularly scheduled follow-up at 1 and discharge time as recorded in the EMR for the en-
month, 3 months, 6 months, and 1 year counter number that correlated to the surgical admis-
The data in italics are the important interventions for measured outcomes sion for lumbar fusion. PCA use and all other analgesic
POS preoperative services, PSA presurgical admission, OSA obstructive sleep medications used were recorded for postoperative day 0
apnea, CPAP continuous positive airway pressure, NPRS numeric pain rating
scale, PONV postoperative nausea/vomiting, IVP IV push, EBL estimated blood through postoperative day 3, at which time it was ex-
loss, PCA patient-controlled analgesia, PT physical therapy, SW social work, pected that a majority of patients would be discharged.
NS neurosurgery
Medication use at each postoperative visit was also re-
corded. Patient pain was measured based on an 11-point
started, and methocarbamol 1500 mg IV or PO was numerical pain rating scale (NPRS) in which patients
given to manage pain. Patients also received a rate their pain ranging from 0 (no pain) to 10 (worst im-
non-opioid regimen for 7 days including celecoxib 200 aginable pain). The NPRS has been shown to be effective
mg Q12H PO, gabapentin 300 mg Q8H PO, and acet- at showing pain improvement in patients with low back
aminophen 975 mg Q6H PO (Doleman et al. 2015). To pain when the NPRS shows a difference of greater than
decrease hospital LOS, the postoperative interventions 2 points (Childs et al. 2005). The NPRS was measured
included early physical therapy (PT) and social work in- prior to surgery, each day during their hospital stay, and
volvement on postoperative day 1 with early at each postoperative clinic visit. During this project,
mobilization. Each patient was contacted the day after each NPRS measurement was multiplied by 10 to create
discharge by phone and then seen in the office for a a pain score range of 0 (no pain) to 100 (worst imagin-
2-week wound check visit. Following the wound check able pain) to simplify the analysis of pain scores.
visit, each patient was scheduled for regular follow-up at
3-, 6-, and 12-month intervals.
Statistical methods
A statistician (co-author LN) performed all statistical
Data collection
analyses. Categorical variables were computed using the
The pre-ERAS patients (historical control group) were
Monte Carlo simulations of exact p values from Pear-
identified through the electronic medical record (EMR)
son’s chi-squared tests. For continuous variables, p
and included patients who underwent 1–2-level lumbar
values were computed using the Wilcoxon rank sum test
fusion surgery between October 23, 2013, and Septem-
if normality checks using the Shapiro-Wilk test failed.
ber 9, 2015. A transition phase after the ERAS protocol
Otherwise, two-sample t tests were used. If variances
was started lasted from September 10, 2015, to Novem-
from pre-ERAS and post-ERAS were found to be un-
ber 5, 2015. During this transition phase, staff and physi-
equal using the F test, then the Satterthwaite
cians were educated on the protocol and became
two-sample t test was used. Otherwise, the pooled
familiar with its steps to improve compliance. Regular
two-sample t test was used. As this was a QI project,
meetings of all ERAS team leads and members were
there was no formal hypothesis testing, nor any formal
held. The entire ERAS protocol was made available in
sample size calculation.
the EMR to be used as a reference when needed, and re-
minders were integrated into the intraoperative
anesthesia EMR. After the transition phase, post-ERAS Results
patients underwent surgery between November 9, 2015, Demographics
and May 3, 2017. Overall, 230 eligible surgical patients were included.
All preoperative, intraoperative, and postoperative data There were 123 patients in the pre-ERAS group
were collected from the EMR for both the pre-ERAS and (23-month period), 11 patients in the transition phase
ERAS groups and entered into a database. We followed (2-month period), and 96 patients in the post-ERAS
patients for up to 1 year from their surgical date. This re- group (18-month period). The pre-ERAS and post-ERAS
port includes data collected from the preoperative period group patients were not significantly different with re-
through patients’ first postoperative visit within 30 days of gard to potential confounders such as demographics and
discharge from the hospital (Table 5). comorbidities except for the rate of obstructive sleep
apnea (4.2% in the post-ERAS group and 12.9% in the
Prespecified quality metrics pre-ERAS group, see Table 2).
Prespecified quality metrics included the number of case Perioperative characteristics were also similar between
cancelations, the incidence of postoperative nausea and the two groups (Table 3). Postoperative mobility and
vomiting (PONV), opioid usage, postoperative pain, and complications were not statistically different between the
length of stay. The protocol planned to use rescheduled pre-ERAS and post-ERAS groups.
Smith et al. Perioperative Medicine (2019) 8:4 Page 5 of 9

Protocol compliance movement was started to improve efficiency, reduce

Compliance with the ERAS bundle was mixed (Table 4). morbidity, improve patient experience, and decrease
On the day of surgery, post-ERAS patients received oral cost. Our spine ERAS QI protocol was implemented to
gabapentin and acetaminophen with 78% and 81% com- improve post-operative patient care and surgical out-
pliance rates, respectively. Dexamethasone administered comes similar to what has been done with ERAS in
after induction of general anesthesia increased from other surgical specialties.
4.8% of pre-ERAS patients to 27% of post-ERAS patients There are few reports of ERAS in spine surgery, with
(p < 0.0001, Table 3). There was only a small increase in most publications focusing on the implications and feasibil-
the number of patients receiving ketamine in the ity of implementing a spine ERAS program (Wainwright et
post-ERAS group, which was not significant. All patients al. 2016; Blackburn et al. 2016; Ali et al. 2018). Blackburn et
were to receive 1500 mg methocarbamol and a PCA in al. implemented an ERAS bundle for all elective spine sur-
the recovery room. The rate of methocarbamol use im- geries including some lumbar fusions (Blackburn et al.
proved in the post-ERAS group (44% of pre-ERAS ver- 2016), and the other spinal ERAS projects have investigated
sus 62% in post ERAS groups, p = 0.0137). Prior to endoscopic lumbar fusions and correction of scoliosis
ERAS, 93% of patients were already receiving a PCA so (Wang et al. 2017; Gornitzky et al. 2016; Muhly et al. 2016).
the slight increase to 94.8% was not surprising. Based on these other reports, our ERAS protocol was more
pragmatic and focus on improving postoperative outcomes
Outcomes by decreasing case cancelations, incidence of PONV, post-
After the implementation of ERAS, we observed a statis- operative pain, postoperative opioid usage, and length of
tically significant (p = 0.0125) decrease in the adminis- hospital stay.
tration of rescue antiemetics in the recovery room for By focusing on improved communication between the
nausea (40% pre-ERAS versus 24% post-ERAS) (Table 4). clinic staff, surgeons, anesthesiologists, and operating
With regard to opioid administration, in the pre-ERAS room scheduling staff, we hoped to decrease delays and
group, 7% of patients required a PCA after 24 h, whereas case cancelations. We attempted to measure case cancel-
this decreased to 0% post-ERAS. Long-acting opioids in- ations by determining which lumbar fusions had been
cluded OxyContin, MSContin, Methadone, or any other rescheduled, but we found that this was unreliable due
opioids that are designated as an extended-release. to case cancelations that occurred for unrelated reasons
Short-acting opioids were defined as any opioid not in- such as illness and other patient factors. It should be
cluded as an extended-release or long-acting opioid. No mentioned that the neuroanesthesiologists and neuro-
significant differences were observed in the number of surgeons remarked that after the ERAS implementation,
patients requiring short-acting opioids postoperatively or patients were better prepared on the day of surgery, and
at the first postoperative visit. However, we observed a there were fewer delays due to problems with surgical
benefit with regard to long-acting opioids with lower use consents, waiting for labs, retrieving missing preopera-
at the first postoperative visit after implementation of tive evaluations, and medication histories.
ERAS (14.6% pre-ERAS vs. 5.2% post-ERAS, p = 0.0253, Pain control was another focus of the ERAS protocol.
see Table 5). The percentage of patients receiving anti- Our ERAS protocol included oral non-opioid medica-
convulsants at discharge increased from 22 to 67% in tions preoperatively to help reduce opioid needs postop-
the pre-ERAS and post-ERAS groups, respectively. This eratively. The acute pain service was to be made aware
increase was expected since it reflects the use of gaba- of all high-risk pain patients, and they were to receive
pentin postoperatively as part of the ERAS bundle. The ketamine 30 mg IV with the induction of anesthesia. All
length of stay between the pre-ERAS and post-ERAS patients were to receive a PCA postoperatively, and
groups decreased by 5 h from 96 to 92 h; however, this 1500 mg of methocarbamol in the PACU for early pain
did not reach statistical significance (p = 0.1372). Postop- control. Gabapentin, celecoxib, and acetaminophen were
erative pain scores were similar in both groups. The also added to the postoperative pain regimen which is
NPRS at the post-discharge follow-up visit was also not reflected in the increased number of patients on
statistically different between the two groups. anti-convulsant medications in the postoperative period.
A multimodal approach to pain treatment has been
Discussion shown to decrease postoperative opioid use and de-
Our results did show a small improvement in the incidence creased time to mobilization in spine surgery (Mathiesen
of PONV and a decrease in the use of long-acting opioid, et al. 2013). The Post-ERAS patients showed decreased
but no differences in postoperative pain, short-term opioid postoperative opioid use with significantly fewer patients
use, and length of stay were found. using long-acting opioids postoperatively, and none of
Patient care often lags behind the most recent the post-ERAS patients required a PCA after 24 h. The
evidence-based practice recommendations. The ERAS ERAS interventions did not change the number of
Smith et al. Perioperative Medicine (2019) 8:4 Page 6 of 9

Table 2 Demographic and co-morbidities Table 3 Perioperative data

Pre-ERAS Post-ERAS *p Pre-ERAS (n = Post-ERAS *p
(n = 123) (n = 96) value 123) (n = 96) value
Number of patients 123 96 Preoperative pain score (NPRS) 63.7 (24.7) 69.5 (22.3) 0.0969
Gender, # and % male 53 (43.1%) 48 (50.0%) 0.3395 Pre-operative medications given
orally on the day of surgery prior
Age 60.3 (12.9) 61.3 (13.3) 0.7308 to OR
BMI 29.7 (5.5) 29.7 (4.8) 0.7783 • Gabapentin 900 mg 4 (3.3%) 75 (78.1%) <.0001
Coronary artery disease, #(%) 15 (12.2%) 10 (10.4%) 0.8286 • Acetaminophen 975 mg 3 (2.4%) 78 (81.3%) <.0001
Hypertension, #(%) 68 (55.3%) 50 (52.1%) 0.6715 • Aprepitant 40 mg 1 (0.8%) 2 (2.1%) 0.5786
Asthma, #(%) 13 (10.6%) 7 (7.3%) 0.4834
Number of levels fused
Chronic obstructive pulmonary disease, 8 (6.5%) 7 (7.3%) 1.0000
•1 71 (57.7%) 54 (56.3%) 0.8906
#(%)
•2 52 (42.3%) 42 (43.8%) 0.8879
Diabetes mellitus—non-insulin 9 (7.3%) 15 (15.6%) 0.0794
dependent, #(%) Estimated blood loss > 300 mL 24 (19.5%) 33 (34.4%) 0.0152
Diabetes mellitus—insulin dependent, 4 (3.3%) 2 (2.1%) 0.7009 Blood products given in OR, 1 (0.8%) 1 (1.0%) 1.0000
#(%) #(%)
History of cerebrovascular accident, 4 (3.3%) 5 (5.2%) 0.5096 Colloid given in OR 13 (10.6%) 8 (8.3%) 0.6480
#(%)
Volume of crystalloid (mL) 2070.9 (736.2) 1761.5 (739.2) 0.0003
Anxiety, #(%) 22 (17.9%) 17 (17.7%) 1.0000
Drain placed 118 (95.9%) 96 (100.0%) 0.0720
Depression, #(%) 18 (14.6%) 10 (10.4%) 0.4078
Dexamethasone 8 mg IV 6 (4.9%) 26 (27.1%) <.0001
Kidney disease, #(%) 10 (8.1%) 4 (4.2%) 0.2775 intraoperative
Liver disease, #(%) 2 (2.1%) 0.1926 Ketamine 30 mg on induction 2 (1.6%) 5 (5.2%) 0.2497
Obstructive sleep apnea, #(%) 16 (13.0%) 4 (4.2%) 0.0326 Duration of surgery (min) 184.7 (85.4) 212.2 (140.4) 0.3093
Alcohol abuse, #(%) 2 (1.6%) 0.5024 Intraoperative dural tear 4 (3.3%) 1 (1.0%) 0.3959
Tobacco abuse, #(%)
• Yes, NOT within the last 6 months 39 (31.7%) 41 (42.7%) 0.1199 patients using short-acting opioids in the postoperative
• Yes, within the last 6 months 25 (20.3%) 15 (15.6%) 0.3850 period, but it should be recognized that this was a quali-
Substance abuse, #(%) tative measurement and the morphine equivalents of all
opioids used were not collected as part of the data. Des-
• Yes, NOT within the last 6 months 2 (1.6%) 4 (4.2%) 0.4307
pite the decrease in long-acting opioid use, there was no
• Yes, within the last 6 months 3 (2.4%) 2 (2.1%) 1.0000
difference in pain scores between the pre-ERAS and
ASA PS class ≥ 3 63 (51.2%) 50 (52.1%) 1.0000 post-ERAS groups.
Non-surgical treatments utilized PONV is often a problem in patients following surgery.
• Physical therapy 51 (41.5%) 23 (24.0%) 0.0104 In addition to the regular antiemetic regimen that is given
• Acupuncture 17 (13.8%) 8 (8.3%) 0.2928 to most patients including intraoperative ondansetron, the
ERAS protocol included dexamethasone 8 mg after induc-
• Chiropractic 27 (22.0%) 14 (14.6%) 0.2222
tion and oral aprepitant for high-risk patients preopera-
• Epidural or facet injections 78 (63.4%) 51 (53.1%) 0.1310
tively. Despite a small increase of 4 to 27% compliance
Preoperative pain medication (last 30 days) between the pre- and post-ERAS patients receiving dexa-
• NSAIDs 76 (61.8%) 48 (50.0%) 0.0992 methasone intraoperatively and only 2 patients receiving
• Opioids, short-acting (immediate- 57 (46.3%) 47 (49.0%) 0.7852 aprepitant in the post ERAS group, there was still a sig-
release) nificant decrease in the patients requiring a rescue anti-
• Opioids, long-acting (extended-re- 4 (3.3%) 2 (2.1%) 0.7021 emetic in the recovery room. With improved compliance,
lease, e.g., OxyContin, MSContin, this may improve further in the future.
Methadone)
Decreasing LOS helps to reduce costs and is an im-
• Anticonvulsants 30 (24.4%) 29 (30.2%) 0.3720
portant outcome measure in many enhanced recovery
• Antidepressants 31 (25.2%) 29 (30.2%) 0.4464 protocols (Gustafsson et al. 2013; Nygren et al. 2013;
• Benzodiazepines 18 (14.6%) 12 (12.5%) 0.6964 Lassen et al. 2013). There are many factors that affect
• Muscle relaxants 24 (19.5%) 11 (11.5%) 0.1350 LOS. Preoperative comorbidities are not the sole con-
• Acetaminophen 42 (34.1%) 34 (35.4%) 0.8826
tributor to LOS, and the most significant factors that
prolong LOS are postoperative events such as bleeding,
drains, late mobilization, and delayed discharge to
Smith et al. Perioperative Medicine (2019) 8:4 Page 7 of 9

Table 4 Postoperative data Table 4 Postoperative data (Continued)

Pre-ERAS Post-ERAS *p Pre-ERAS Post-ERAS *p
(n = 123) (n = 96) value (n = 123) (n = 96) value
PACU temperature on arrival 35.8 (5.1) 36.5 (0.4) 0.4443 • Opioids, long acting 28 (22.8%) 5 (5.2%) 0.0002
(extended-release, e.g.,
Antiemetic given in PACU 49 (39.8%) 23 (24.0%) 0.0125 OxyContin, MSContin,
Methocarbamol 1500 mg given in 54 (43.9%) 59 (61.5%) 0.0137 Methadone)
PACU
• Anticonvulsants 30 (24.4%) 66 (68.8%) <.0001
NPRS before PACU discharge 50.0 (22.3) 43.7 (23.9) 0.0925
• Antidepressants 22 (17.9%) 10 (10.4%) 0.1311
Total patients with PCA 115 91 (94.8%) 0.7803
• Benzodiazepines 34 (27.6%) 29 (30.2%) 0.7643
(93.5%)
• Muscle relaxants 112 86 (89.6%) 0.8108
• Morphine, # (% of total PCA) 56 (45.5%) 38 (39.6%) 0.4209
(91.1%)
• Hydromorphone, # (% of total PCA) 55 (44.7%) 51 (53.1%) 0.2236
• Acetaminophen 94 (76.4%) 82 (85.4%) 0.1236
• Fentanyl, # (% of total PCA) 5 (4.1%) 2 (2.1%) 0.4634 Postoperative day #3
PCA duration≥ 24 h 9 (7.3%) 0.0116 NPRS pain scores
Urinary catheter duration ≥ 24 h 42 (34.1%) 30 (31.3%) 0.6565 • Minimum pain score 8.9 (14.0) 15.6 (19.1) 0.0158
Day of surgery medications given • Maximum pain score 69.1 (22.6) 68.0 (21.4) 0.499
• NSAIDs 4 (3.3%) 10 (10.4%) 0.0533 Postoperative day #3
medications given
• Opioids, short-acting 123 95 (99.0%) 0.4343 (for patients not yet
(immediate-release) (100.0%) discharged)
• Opioids, long-acting (extended-re- 6 (4.9%) 0.0345 • NSAIDs 7 (5.7%) 7 (7.3%) 0.7846
lease, e.g., OxyContin, MSContin,
Methadone) • Opioids, short-acting 91 (74.0%) 65 (67.7%) 0.3713
(immediate-release)
• Anticonvulsants 22 (17.9%) 55 (57.3%) <.0001
• Opioids, long-acting 22 (17.9%) 4 (4.2%) 0.0020
• Antidepressants 8 (6.5%) 5 (5.2%) 0.7788 (extended-release, e.g.,
• Benzodiazepines 19 (15.4%) 16 (16.7%) 0.8518 OxyContin, MSContin,
Methadone)
• Muscle relaxants 97 (78.9%) 83 (86.5%) 0.1603
• Anticonvulsants 29 (23.6%) 52 (54.2%) 0.0001
• Acetaminophen 62 (50.4%) 45 (46.9%) 0.6868
• Antidepressants 16 (13.0%) 6 (6.3%) 0.1114
Postoperative day #1 NPRS pain scores
• Benzodiazepines 24 (19.5%) 18 (18.8%) 1.0000
• Minimum pain score 10.8 (16.0) 15.3 (17.0) 0.0249
• Muscle relaxants 97 (78.9%) 66 (68.8%) 0.1208
• Maximum pain score 75.0 (19.4) 75.8 (18.4) 0.791
• Acetaminophen 63 (51.2%) 54 (56.3%) 0.4919
Postoperative day #1 medications given
Postoperative day patient 2.2 (0.4) 2.3 (0.6) 0.4102
• NSAIDs 6 (4.9%) 11 (11.5%) 0.0777 out of bed
• Opioids, short-acting (immediate- 119 92 (95.8%) 1.0000 Postoperative day patient 2.2 (0.4) 2.2 (0.5) 0.6943
release) (96.7%) seen by PT
• Opioids, long-acting (extended-re- 23 (18.7%) 5 (5.2%) 0.0028 Postoperative complications
lease, e.g., OxyContin, MSContin,
Methadone) • Reintubation 2 (1.6%) 1 (1.0%) 1.0000
• Anticonvulsants 30 (24.4%) 71 (74.0%) <.0001 • MI
• Antidepressants 20 (16.3%) 10 (10.4%) 0.2312 • Death
• Benzodiazepines 35 (28.5%) 33 (34.4%) 0.3779 • Rapid response 2 (2.1%) 0.1928
• Muscle relaxants 113 87 (90.6%) 0.8096 • Other 4 (4.2%) 0.0355
(91.9%) Length of stay (hours) 96.2 (32.0) 92.3 (36.9) 0.1372
• Acetaminophen 98 (79.7%) 86 (89.6%) 0.0620
Postoperative day #2 NPRS pain scores rehabilitation facilities (Gruskay et al. 2015). To help pa-
• Minimum pain score 6.9 (12.1) 15.3 (18.9) 0.0003 tients get out of bed on postoperative day 1, the protocol
• Maximum pain score 73.5 (18.9) 74.8 (19.7) 0.5559 required urinary catheters to be removed in the morning
Postoperative day #2 medications given (for patients not yet discharged) after surgery, but compliance was poor with no change
between the two groups. The protocol also included
• NSAIDs 7 (5.7%) 9 (9.4%) 0.4352
early involvement of social work to identify any potential
• Opioids, short-acting (immediate- 111 83 (86.5%) 0.4063
discharge needs and to have physical therapy get all pa-
release) (90.2%)
tients out of bed on postoperative day 1 (Kanaan et al.
Smith et al. Perioperative Medicine (2019) 8:4 Page 8 of 9

Table 5 Postoperative follow-up within 30 days of discharge neurosurgery clinic visits and preoperative services visit
Pre-ERAS Post-ERAS *p in addition to printed literature that patients received.
(n = 123) (n = 96) value Information regarding fasting guidelines and day of sur-
Patient contacted 1 day after 4 (3.3%) 9 (9.4%) 0.0803 gery medication use has also been an issue for some pa-
discharge tients in the past at our institution, so education on
Pain at postoperative wound visit 45.6 (25.3) 48.9 (26.0) 0.3575 these topics was provided to patients verbally and in
(NPRS)
printed handouts prior to surgery.
Medications used at postoperative Some of the limitations of this study include a single
wound visit
institution, and it was a non-randomized, non-blinded
• NSAIDs 32 (26.0%) 24 (25.0%) 0.8728 project with historical patients identified from a record
• Opioids, short-acting (immedi- 100 (81.3%) 80 (83.3%) 0.7247 search of the EMR. Compliance from nursing, surgical,
ate-release)
and anesthesia teams in following the protocol was also
• Opioids, long-acting (extended- 18 (14.6%) 5 (5.2%) 0.0253 suboptimal which is reflected in the intervention compli-
release, e.g., OxyContin, MSContin,
Methadone) ance. There are many barriers to implementing ERAS
protocols such as ineffective communication among
• Anticonvulsants 27 (22.0%) 64 (66.7%) < .0001
team members, patient non-compliance, staff turnover
• Antidepressants 25 (20.3%) 22 (22.9%) 0.7416
with the need for continued education, and physician
• Benzodiazepines 27 (22.0%) 23 (24.0%) 0.7449 and staff non-compliance. This is not unique to this
• Muscle relaxants 97 (78.9%) 63 (65.6%) 0.0348 ERAS project and has been shown in other studies as
• Acetaminophen 49 (39.8%) 48 (50.0%) 0.1675 well (Kahokehr et al. 2009; Pedziwiatr et al. 2015). An-
Signs of infection present 2 (1.6%) 2 (2.1%) 1.0000 other limitation of this study was that pain medication
use was measured qualitatively and not quantitatively.
The use of morphine equivalents would have provided a
better comparison of narcotic pain use. It is also worth
2015). Social work and physical therapy are both very mentioning that improving communication in the peri-
busy services in our institution, and involving them early operative period made all care providers aware of the
was difficult due to time constraints and limited cover- ongoing QA project and may lead to the Hawthorne ef-
age on the weekends. The average time for physical ther- fect. Providers involved in care of the ERAS patients
apy to see the pre-ERAS and post-ERAS patients was may have been more particular about charting and the
similar at 2 days postoperatively. We did not observe any care they provided, because they were aware that data
significant reduction in LOS, which may be related to was being collected. This change in behavior by itself
poor compliance of protocol adherence. Hence, we were may result in better outcomes.
unable to determine whether the ERAS protocol ele- Despite the limitations, we were able to successfully
ments have any significant effect on LOS. implement the spine ERAS protocol at our institution
New ERAS protocols not only focus on specific peri- with improvement in some aspects of patient out-
operative interventions by surgeons and anesthesiolo- comes. The care of the lumbar fusion patients intra-
gists, but they have expanded to focus on preoperative operatively has become more standardized, and
patient education, building more effective team care perioperative teams have become more familiar with
models, improving patient satisfaction, and improved the protocol and compliance has continued to im-
discharge planning. Our ERAS protocol focused on prove. This early data showing decreases in PONV
some of these expanded ideas to improve patient educa- and long-acting opioid use is also promising as we
tion and interdepartmental teamwork. Preoperative pa- continue to move forward with this project. This
tient education has become an important part of study also demonstrates the areas where implementa-
improving patient care perioperatively. Educating pa- tions are most challenging for ERAS QI projects. Fu-
tients about expectations postoperatively can improve ture studies can focus on these areas for further
postoperative patient satisfaction and decrease patient compliance improvement.
morbidities and pain scores after lumbar surgery (Archer
et al. 2011). Smoking cessation is also an important issue Conclusions
included in the preoperative patient education. Smokers In our study, the ERAS protocol was associated with a
experience a higher rate of postoperative pseudarthrosis decrease in the incidence of postoperative nausea, a
and infection after spinal fusion. Smoking cessation can shorter duration of opioid use and a decrease in
help decrease these complications depending on the tim- long-acting opioid use. It also improved communication
ing of smoking cessation (Jackson 2nd and Devine among the perioperative team and improvement in pa-
2016). Smoking cessation was reiterated at both the tient education preoperatively. However, it did not result
Smith et al. Perioperative Medicine (2019) 8:4 Page 9 of 9

in clinically significant reductions in LOS, decreased Apfel CC, Heidrich FM, Jukar-Rao S, Jalota L, Hornuss C, Whelan RP, Zhang K,
postoperative pain scores, or decreased short-acting opi- Cakmakkaya OS. Evidence-based analysis of risk factors for postoperative
nausea and vomiting. Br J Anaesth. 2012;109 (5):742–753.
oid use. Moving forward, we have implemented steps Archer KR, Wegener ST, Seebach C, Song Y, Skolasky RL, Thornton C, et al. The
and education to improve adherence to the protocol, in effect of fear of movement beliefs on pain and disability after surgery for
particular improving the timeliness of postoperative lumbar and cervical degenerative conditions. Spine (Phila Pa 1976). 2011;
36(19):1554–62.
physical therapy and social work assessment of patients. Blackburn JMP, Leung YL, Walburn M. An enhanced recovery program for
elective spinal surgery patients. JCOM. 2016;23(10):462–9.
Abbreviations Childs JD, Piva SR, Fritz JM. Responsiveness of the numeric pain rating scale in
ASA PS: American Society of Anesthesiologists physical status; patients with low back pain. Spine (Phila Pa 1976). 2005;30(11):1331–4.
CPAP: Continuous positive airway pressure; EBL: Estimated blood loss; Dagenais S, Caro J, Haldeman S. A systematic review of low back pain cost of illness
EMR: Electronic medical record; ERAS: Enhanced recovery after surgery; studies in the United States and internationally. Spine J. 2008;8(1):8–20.
ERP: Enhanced recovery protocol; IV: Intravenous; IVP: Intravenous push; Daneshmand S, Ahmadi H, Schuckman AK, Mitra AP, Cai J, Miranda G, et al. Enhanced
LBP: Low back pain; LOS: Length of stay; NPRS: Numerical pain rating scale; recovery protocol after radical cystectomy for bladder cancer. J Urol. 2014;192(1):50–5.
NS: Neurosurgery; NSAID: Non-steroidal anti-inflammatory drug; Deyo RA, Weinstein JN. Low back pain. N Engl J Med. 2001;344(5):363–70.
OR: Operating room; OSA: Obstructive sleep apnea; PACU: Post-anesthesia Doleman B, Heinink TP, Read DJ, Faleiro RJ, Lund JN, Williams JP. A systematic
care unit; PCA: Patient-controlled analgesia; PO: Per os; PONV: Postoperative review and meta-regression analysis of prophylactic gabapentin for
nausea and vomiting; POS: Preoperative services; PSA: Presurgical admission; postoperative pain. Anaesthesia. 2015;70(10):1186–204.
PT: Physical therapy; QI: Quality improvement; SW: Social work Gornitzky AL, Flynn JM, Muhly WT, Sankar WN. A rapid recovery pathway for
adolescent idiopathic scoliosis that improves pain control and reduces time to
Funding inpatient recovery after posterior spinal fusion. Spine Deform. 2016;4(4):288–95.
Funding for data analysis was provided by the Stony Brook University Gruskay JA, Fu M, Bohl DD, Webb ML, Grauer JN. Factors affecting length of stay
department of anesthesiology. after elective posterior lumbar spine surgery: a multivariate analysis. Spine J.
2015;15(6):1188–95.
Availability of data and materials Gustafsson UO, Scott MJ, Schwenk W, Demartines N, Roulin D, Francis N, et al.
All data generated or analyzed during this study are included in this Guidelines for perioperative care in elective colonic surgery: enhanced recovery
published article. after surgery (ERAS®) society recommendations. World J Surg. 2013;37(2):259–84.
Jackson KL 2nd, Devine JG. The effects of smoking and smoking cessation on spine
Authors’ contributions surgery: a systematic review of the literature. Global Spine J. 2016;6(7):695–701.
JS was involved in the project implementation, data collection, data entry, Kahokehr A, Sammour T, Zargar-Shoshtari K, Thompson L, Hill AG. Implementation
and data interpretation and a significant contributor in writing the of ERAS and how to overcome the barriers. Int J Surg. 2009;7(1):16–9.
manuscript. SP contributed to project planning, project implementation, Kanaan SF, Waitman LR, Yeh HW, Arnold PM, Burton DC, Sharma NK. Structural
ERAS protocol training of staff, data collection, and data interpretation and equation model analysis of the length-of-hospital stay after lumbar spine
authoring the manuscript. CC was a major contributor in communication surgery. Spine J. 2015;15(4):612–21.
between departments and various offices, and she also played a large role in Kehlet H. Multimodal approach to control postoperative pathophysiology and
project planning, data collection, and implementation of the protocol. RD rehabilitation. Br J Anaesth. 1997;78(5):606–17.
was involved in the project planning and implementation. LN performed the Lassen K, Coolsen MM, Slim K, Carli F, de Aguilar-Nascimento JE, Schafer M, et al.
data analysis. KS was a major contributor to the data entry. TG played a role Guidelines for perioperative care for pancreaticoduodenectomy: enhanced recovery
in guiding the project implementation and editing manuscript. EBG played a after surgery (ERAS®) society recommendations. World J Surg. 2013;37(2):240–58.
role in overseeing the project implementation, data interpretation, and Loftus RW, Yeager MP, Clark JA, Brown JR, Abdu WA, Sengupta DK, et al.
authoring manuscript. All authors read and approved the final manuscript. Intraoperative ketamine reduces perioperative opiate consumption in opiate-
dependent patients with chronic back pain undergoing back surgery.
Ethics approval and consent to participate Anesthesiology. 2010;113(3):639–46.
This project was approved by a Stony Brook Medical Center Quality Mathiesen O, Dahl B, Thomsen BA, Kitter B, Sonne N, Dahl JB, et al. A
Improvement board to proceed as outlined above. comprehensive multimodal pain treatment reduces opioid consumption
after multilevel spine surgery. Eur Spine J. 2013;22(9):2089–96.
Consent for publication Muhly WT, Sankar WN, Ryan K, Norton A, Maxwell LG, DiMaggio T, et al. Rapid
Not applicable recovery pathway after spinal fusion for idiopathic scoliosis. Pediatrics. 2016;
137(4):E20151568–e20151568.
Competing interests Nygren J, Thacker J, Carli F, Fearon KC, Norderval S, Lobo DN, et al. Guidelines for
The authors declare that they have no competing interests. perioperative care in elective rectal/pelvic surgery: enhanced recovery after
surgery (ERAS®) society recommendations. World J Surg. 2013;37(2):285–305.
Pedziwiatr M, Kisialeuski M, Wierdak M, Stanek M, Natkaniec M, Matlok M, et al. Early
Publisher’s Note implementation of enhanced recovery after surgery (ERAS(R)) protocol - compliance
Springer Nature remains neutral with regard to jurisdictional claims in improves outcomes: a prospective cohort study. Int J Surg. 2015;21:75–81.
published maps and institutional affiliations. Proietti L, Scaramuzzo L, Schiro GR, Sessa S, Logroscino CA. Complications in lumbar
spine surgery: a retrospective analysis. Indian J Orthop. 2013;47(4):340–5.
Author details Wainwright TW, Immins T, Middleton RG. Enhanced recovery after surgery (ERAS) and its
1
Department of Anesthesiology, Stony Brook University Medical Center, 101 applicability for major spine surgery. Best Pract Res Clin Anaesthesiol. 2016;30(1):91–102.
Nicolls Rd, Stony Brook, NY 11794, USA. 2Department of Neurosurgery, Stony Wang MY, Chang PY, Grossman J. Development of an enhanced recovery after surgery
Brook University Medical Center, 101 Nicolls Rd, Stony Brook, NY 11794, USA. (ERAS) approach for lumbar spinal fusion. J Neurosurg Spine. 2017;26(4):411–8.
3
Department of Applied Mathematics and Statistics, Stony Brook University,
100 Nicolls Rd, Stony Brook, NY 11794, USA.

Received: 11 February 2019 Accepted: 29 April 2019

References
Ali ZS, Ma TS, Ozturk AK, Malhotra NR, Schuster JM, Marcotte PJ, et al. Pre-optimization
of spinal surgery patients: development of a neurosurgical enhanced recovery
after surgery (ERAS) protocol. Clin Neurol Neurosurg. 2018;164:142–53.