Abstract
Background
Enhanced recovery after surgery (ERAS) pathways have been shown to improve pancreatic surgery outcomes, though feasibility in a community hospital remain unclear. We hypothesized that an ERAS protocol would reduce hospital length of stay (LOS) without increased morbidity.
Methods
An ERAS pathway was initiated for patients undergoing pancreatic surgery at a community cancer center and compared to a historical cohort. The primary outcome was hospital LOS. Secondary outcomes included 30-day readmission rates, comprehensive complication index (CCI®), textbook outcomes (TO), and mortality.
Results
A total of 144 patients were included, with 63 patients in the ERAS group and 81 in the control group. The mean LOS decreased significantly in the ERAS group (6.85 [± 4.8]) vs 9.96 [±6.8] days, P = .001), without an increase in 30-day admission rates or CCI.
Conclusions
Implementation of an ERAS protocol in a community setting reduced LOS without a corresponding increase in readmission rates or morbidity.
Introduction
Pancreatic surgery is one of the most sophisticated and challenging intra-abdominal procedures performed and coupled with oncologic therapy remains the standard of care for pancreatic malignancy. Centralization of pancreatic surgery has been accelerated in recent decades, driven not only by national trends in the centralization of cancer care1,2,3 but also by national quality metrics, such as those published by the Leapfrog Group, 4 that have set minimum standards for surgeon and hospital volume thought to be associated with better outcomes. 3 Although specialized centers are able to provide sophisticated care, centralization to a diminishing number of select regional tertiary or quaternary academic centers burdens many patients with further travel, heavier financial burdens, and separation from their support network.5,6 By providing local access to these services, community centers with experience in complex surgery and oncology care play an essential role in providing a high standard of care in their communities.
Even at specialized centers, and despite significant advances in postoperative care, pancreatic surgery remains associated with prolonged recovery time. The national median length of stay is 8.5 days with a morbidity rate of nearly 30% for all patients undergoing pancreaticoduodenectomy (PD). 7 Enhanced recovery after surgery (ERAS) pathways have been demonstrated to reduce hospital length of stay (LOS) and expedite patient recovery through evidence-based, multidisciplinary protocols for a variety of oncologic surgeries.8-10 Previous literature has validated the effectiveness of ERAS pathways implemented at high volume academic centers for patients undergoing PD,9,11,12 while implementation of ERAS pathways for patients undergoing colorectal surgery at a community hospital was effective in reducing LOS without increased morbidity. 13
We hypothesized that development and implementation of ERAS pathways for patients undergoing pancreatic surgery was feasible in a community hospital setting and would reduce LOS and accelerate patient recovery.
Methods
Study Design and Patient Selection
This retrospective cohort study was approved by the Western Michigan Cancer Center Institutional Review Board. All patients in this study underwent elective pancreatectomy by a single hepatobiliary surgeon at 2 community hospitals, beginning with the start of a community hepatobiliary program in 2015. Control patient were identified retrospectively and included patient who underwent surgery between October 2015 and implementation of the ERAS program in July 2019. Prior to implementation, no standardized protocol was utilized and all postoperative decisions were made by the surgical oncology team on a per patient basis. Patients were excluded from the control group if they received perioperative total parenteral nutrition (TPN) or were treated urgently or emergently. Postoperative follow-up in the control group included a postoperative visit in 7-10 days after discharge without any defined postoperative engagement from the surgical clinic.
The ERAS protocol was developed in 2019 and was based on published guidelines which include preadmission patient education, preoperative nutrition, locoregional anesthetic use, conservative perioperative fluid management, multimodal pain management with minimization of narcotic use, early resumption of oral intake, and early ambulation.14,15 Patients were enrolled prospectively beginning in 2019 and included all patients that underwent elective pancreatectomies at one of two community hospitals by the same hepatobiliary surgeon. To be included, experimental patients were Eastern Cooperative Oncology Group (ECOG) Performance Status 14 ECOG 0/1 score; they were also required to comply with 5 days of immunonutrition preoperatively and were managed postoperatively as part of the ERAS protocol (See supplemental file). Excluded patients included those who did not complete preoperative immunonutrition, who received perioperative TPN, or who were treated urgently or emergently. Enhanced Recovery Protocol (ERAS) measures for the experimental group included preoperative patient counseling with the surgeon, nursing staff, dietician, and oncology navigator. Preoperative immunonutrition (eg, IMPACT Advanced Recovery®; Nestlé Health Science, Switzerland) was prescribed for 5 days with consumption of three 178 mL cartons each day. The ERAS for each type of procedure outlines protocols for preoperative patient education, multimodal postoperative pain management, early ambulation, advancement of diet, goal-directed fluid management, anticoagulation as previously published by our program, 16 and early discharge.14,15 Close follow-up consisting of a telephone call one day after discharge and follow-up in clinic within 1 week of discharge was established for each patient. Decisions regarding number and timing of telephone visits and postoperative appointments were left at the discretion of the nursing staff, physician assistant, and physician following the initial phone consultation. Most patients were seen at 1, 4, and 8 weeks post discharge.
Development and implementation of the program required multidisciplinary collaboration among surgeons, surgery residents, nursing staff, anesthesia providers, pharmacists, physical and occupational therapists, operating room staff, and clinic staff, with one-on-one and group discussions among health care professionals. A standardized ERAS order set was developed. Keys to the success of the program are clear and consistent messages about expectations for patients regarding activity, diet, and pain management before, during, and after their hospital stay. Postoperative goals were communicated each day with the patient and multidisciplinary care team, aiming for early discharge. Discharge criteria required tolerating adequate oral intake (soft or general diet), adequate fluid intake, adequate pain control, and ability to ambulate safely and perform activities of daily living.
Compliance with the preoperative protocol was augmented by a nurse navigator employed by the cancer center communicated with patients before and after surgery to identify and barriers to compliance with the protocol and ensure follow-up. Discretionary funding has to provide the immunonutrition at no cost to the patient. The nurse navigator also contacted patients after discharge from the hospital on a scheduled basis to confirm adhere to prophylactic postoperative anticoagulation. Surgical residents familiar with the ERAS protocol worked with nursing staff and dieticians during the patient’s inpatient stay to maintain protocol implementation. Deviation from the ERAS protocol occurred with postoperative complications, which were managed on an individual basis.
The primary outcomes of interest included length of hospital stay (LOS) and 30-day readmission rate. Secondary outcomes included textbook outcomes (TO) and major morbidity, represented by a Clavien-Dindo classification between grades 3 and 5. Surgical site infections (SSIs), postoperative pancreatic fistula (POPF) of grade C17, post-pancreatectomy hemorrhage (PPH), and delayed gastric emptying (DGE) were recorded as additional secondary outcomes. Textbook outcomes (TOs) were defined by the absence of 6 complications: bile leak, grade B PPH, grade B/C POPF, Clavien-Dindo complication grade ≥3, in-hospital mortality, and re-admission within 30 days or prolonged LOS.17,18 The comprehensive complication index (CCI®) was calculated using the CCI® Calculator made available to the public domain by the not-for-profit organization AssessSurgery. 18
Statistical Analysis
Descriptive statistics were used for patient demographics, clinical characteristics, and outcome measures. Continuous variables were summarized using the mean (standard deviation; SD) for the approximately normally distributed data and the median (interquartile range; IQR) for skewed data. The differences between the 2 groups were tested using the independent t-test and the Mann-Whitney U-test. Categorical variables were summarized using counts and proportions. Pearson’s chi-squared (X2) and Fisher’s exact tests were used to test group differences of categorical variables. Statistical analysis was performed with R (version 3.1.1) and SPSS (IBM SPSS Statistics for Windows, Version 28.0. Armonk, NY: IBM Corp). Exact P-values were calculated for reliability with statistical significance set at a value of P < .05.
Results
Demographic Characteristics of Patients in the Control and ERAS Groups.
BMI, body mass index; SD, standard deviation; MCN, mucinous cystic neoplasm; SCN, serous cystic neoplasm; IPMN, intraductal papillary mucinous neoplasm.
Perioperative Characteristics.
IV, intravenous; EBL, estimated blood loss (in milliliterl); L, liter; CCI, Comprehensive Complication Index; ASA, American Society of Anesthesiologists; TAP, transversus abdominis plane.
Study Outcomes
Outcomes.
LOS, length of stay (days); CCI, Clinical Complication Index.
Subgroup Analysis by Type of Procedure.
The percentage of patients achieving textbook outcome (TO) in the ERAS group was higher than that in the control group, though this did not reach statistical significance (63.5% vs 48.1%, P = .066). When comparing TO between surgical approaches, TO for PD in the ERAS group was higher compared to the control group (54.3% vs 39.2% P = .191). Mean Comprehensive Complication Index (CCI®) was lower in the ERAS group than in the control group (10.6 vs 17.1, P = .035). This difference was statistically significance. The 2 deaths were due to postoperative multisystem organ failure and PPH. The ERAS group had lower rates of Grade C POPF (1.7% vs 8.6%; P = .067) and DGE (6.4% vs 16%, P = .074).
Other secondary outcomes were similar between the two groups with the exception of superficial surgical site infection (SSI): six patients in the ERAS cohort developed a superficial SSI compared to one in the control group (P = .043). Rates of deep organ space infection and postoperative hemorrhage were similar, as were collective Clavien-Dindo complication rates.
All patients who were enrolled in the ERAS study arm were compliant with preoperative immunonutrition and postoperative anticoagulation. This was achieved.
Discussion
This study demonstrates that with proper protocolization and institutional support good patient outcomes for pancreatic resection are achievable in the community setting. The push for centralization to high volume centers has raised the question of whether pancreatic resection should be performed outside of high volume centers; while there is literature on the successful implementation of an ERAS program at a community hospital 21 for colorectal patients, the reproducibility of these programs in the community setting for pancreatic resection has not been well established. Despite not meeting high volume criteria, our ERAS protocol was successfully implemented with patient outcomes comparable to those at high volume centers.19,21-24 We were able to decrease LOS, particularly in pancreaticoduodenectomy patients, from 10 days to 6.9 days and this decrease was not associated with increased readmissions. The LOS for patients in the distal pancreatectomy group was also reduced (6.36 vs 8.43 days), although due to the small sample size this did not meet statistical significance.
We elected to implement the ERAS program on all patients as opposed to selecting only low-risk patients. Compliance with preoperative immunonutrition was facilitated by providing this at no cost to the patient, and with close follow up with a nurse navigator we were able to achieve 100% compliance with this aspect of the protocol. Protocols for multimodal pain management and goal-directed fluid management were emphasized by working closely with the anesthesia providers, and we found an improvement in utilization of regional anesthesia and decrease in intraoperative fluid utilization in line with the goals of our ERAS program. We also observed a slight increase in rates of laparoscopic surgery from 25% to 30% of distal pancreatectomies similar to previous reports on implementation of ERAS programs.19-21
This study had the benefit of several key resources within the community. Both community hospitals serve as teaching hospitals affiliated with one medical school, and surgical residents were involved in development and implementation of the protocol. A nurse navigator employed by the cancer was center followed patients closely, allowing us to achieve full compliance with immunonutrition and postoperative anticoagulation. Collaboration with our Anesthesiology colleagues was also vital in the success of the ERAS protocol implementation. Although the individual hospital volumes did not meet Leapfrog criteria, 4 this was ameliorated by the fact that surgical residents at both hospitals were familiarized with the ERAS protocol and were able to provide continuity and consistency. We found that one advantage of a smaller community based hospital is that interdisciplinary teams of nurses, dieticians, surgical residents, and anesthesia providers are often small and cohesive. This allowed us to collaborate effectively and achieve good patient outcomes despite lacking the resources of larger academic center.
We used several variables to report on outcomes after implementation of the ERAS program. Textbook outcomes (TO), which are a composite measure of important outcome variables, are a well suited to measure implementation of ERAs programs, as they capture an array of critical outcomes as opposed to a single measure.25,26 Hepatobiliary-pancreatic studies have found that using TO improved the reporting on hospital performance and interpretation of inter-hospital variation.25-27 TO at our institution were found to be comparable to other reported rates by large centers,27,28 while implementation of the ERAS program resulted in an overall non-significant increase in TO at our institution. This improvement was more pronounced in the PD group, where the rate of TO improved from 39.2 to 54.3% (P = .191), which is similar to previously reported outcomes at a single institution. 28
The Clavien-Dindo classification and the CCI are both validated and effective methods of reporting complications. 29 However, the CCI has been reported as a more accurate metric for use in high-risk patients and in pancreatectomy patients has been shown to have strong correlations with LOS and cost. 29 In our study, there was a decrease in CCI when comparing the ERAS to the control group. Evaluation of procedure-specific complications after implementation of the ERAS program did not identify any reductions in morbidities despite the noted decrease in CCI associated with the ERAS protocol. Previous authors have shown that CCI tends to be a more accurate classification index which may explain this difference in our patient cohort. 30 There was no difference in the rates of PPH, POPF, or delayed gastric emptying in the combined groups or when comparing pancreaticoduodenectomy or distal pancreatectomy, which is consistent with multiple reports of ERAS in pancreas surgery. However, it should be noted that all postoperative pancreatic fistulas (6 cases) developed in patients who underwent PD.
There are several limitations to our study. First, this is a study based in two community hospitals. Our patient population and hospital setting (and therefore, the total sample size) may not reflect other clinical environments. Both community hospitals are affiliated with a medical school and a general surgery residency program; surgical residents were instrumental in protocol implementation, but not all community hospitals will have this benefit. Nonetheless, given the paucity of literature from community hospitals reporting data from ERAS programs, this is a useful study that clearly demonstrates reproducibility of this internationally pioneered ERAS program within a community center. A second limitation was the intrinsic retrospective design of the control group, data for which was collected prior to conceptualization of the study. This limited our ability to determine accurate rates of grade B pancreatic fistula as currently defined by the ISGPS to include >3 weeks of drainage. 17 Although the ERAS elements were collected prospectively, some perioperative parameters could have been underreported, particularly compliance to some ERAS elements. Compliance remains challenging when implementing ERAS pathways, and is therefore essential to audit the compliance and outcome of ERAS pathways frequently and alter management where necessary. Third, since the outcomes of the implementation of ERAS were compared to those of a historical cohort that might have influenced the outcomes, some unknown differences between both cohorts may not have been captured in the baseline characteristics. Additionally, some of the improvements may be related to the impact of accumulated institutional and physician experience with pancreatic surgery.
Conclusions
We have demonstrated that implementation of a pancreas-specific ERAS program at a community hospital is feasible and is associated with improvements in LOS without an increase in readmissions or morbidities. These effects were more pronounced in the pancreaticoduodenectomy patients where improvements in LOS and textbook outcomes were observed. Our results validate the use of pancreas ERAS programs to include community hospital care settings when they are implemented with appropriate planning and teamwork.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.