OR10 RISK FACTORS FOR DELAYED GASTRIC EMPTYING AFTER PANCREATICODUODENECTOMY AND NEED FOR ENTERAL FEEDING ACCESS IN THE ERAS® ERA William B. Lyman 1 ,* , Michael Passeri 2 , Russell C. Kirks 1 , Allyson Cochran 2 , John B. Martinie 2 , Dionisios Vrochides 2 , Erin H. Baker 2 , David A. Iannitti 2 . 1 Department of Surgery, Carolinas Medical Center, Charlotte, United States; 2 Division of HPB Surgery, Carolinas Medical Center, Charlotte, United States * Corresponding author. Division of Colorectal Surgery, University of Michigan, Ann Arbor, United States. Objectives: Current ERAS® guidelines for pancreaticoduodenectomy (PD) do not support prophylactic intraoperative feeding tube placement during PD. Much debate remains between hepatopancreaticobiliary (HPB) surgeons regarding optimal timing and utility of feeding tubes for PD. This study was designed to determine any identifiable risk factors in patients undergoing PD under ERAS® protocol that may increase the risk for developing delayed gastric emptying (DGE) – possibly identifying patients that may benefit from an intraoperative prophylactic feeding tube for nutritional supplementation. Methods: We retrospectively analyzed all patients collected prospectively in the EIAS database who underwent PD at our institution since the implementation of ERAS® in September 2015 (n¼126). Using a combina- tion of literature review (PubMed) and expert consensus within our divi- sion, we identified 11 variables which could possibly contribute to DGE and subsequent need for feeding tube placement. (Table 1) We used ISGPS Definition of Grade B or C DGE as a clinically relevant occurrence (n¼34, 27.0%). STATA® statistical software and Pearson’s chi-squared test were used for all statistical analyses. Results: Preoperative symptoms of gastric outlet obstruction as well as history of chronic pancreatitis significantly correlated with development of Grade B/C DGE and subsequent need for postoperative feeding tube (p¼<0.001 and 0.014 respectively). (Table 1) Of the 31 patients who received an intraoperative feeding tube, only 11 patients (35.5%) devel- oped grade B/C DGE requiring enteral supplementation. Conclusion: While we agree with current ERAS® Society guidelines against routine placement of intraoperative feeding tubes during PD, we propose that patients presenting with symptoms of gastric outlet obstruction or those with a history of chronic pancreatitis may benefit from intraoperative placement of an enteral feeding tube because of the high risk for developing postoperative Grade B or C DGE. Disclosure of Interest: None declared. OR11 EVALUATING THE IMPACT OF SPECIFIC COMPLIANCE ITEMS OF AN ENHANCED RECOVERY AFTER SURGERY (ERAS®) PROTOCOL IN PANCREATICODUODENECTOMY PATIENTS Allyson Cochran * , William B. Lyman, Michael Passeri, Kendra Tezber, Misty Eller, Erin Baker, John Martinie, David A. Iannitti, Dionisios Vrochides. HPB Surgery, Carolinas Medical Center, Charlotte, United States * Corresponding author. Objectives: Compliance adherence in an ERAS ® program has been shown to reduce complications. While most studies look at overall or aggregated compliance groups in colorectal surgery, studies have yet to demonstrate the relationship between which specific compliance elements contribute to better outcomes in PD (pancreaticoduodenectomy) surgery. Methods: Post-ERAS implementation PD data from a single institution was collected from the ERAS ® Interactive Audit System from September 2015- October 2017 (n¼128). In-hospital and 30-day post-discharge measures were combined to create overall 30-day complication measures, which included: delayed gastric emptying (DGE) A-C, Clavien grade of 3+, sur- vival, length of stay (LOS), readmission, and pancreatic fistula (PF) B-C. Itemized compliance elements were binary. Linear and logistic regression were performed as appropriate; reported as coefficient and odds ratio, respectively. Significance was set at p<.05. Results: Oral supplements on day of surgery reduced the odds of DGE (OR¼0.31, p¼.028), while duration of IV fluids increased the odds for DGE (OR¼1.2, p¼.001) and Clavien grade 3+ complication (OR¼1.13, p¼.013). A patient had 11 times the odds of survival if PONV prophylaxis was given (OR¼11.3, p¼.039). Preadmission education (-4.8, p¼.046), postoperative compliance adherence (-0.24, p¼.005), and total compliance adherence (-0.30, p¼.016) were inversely associated with LOS. There were no sig- nificant predictors for readmission and PF. Though not statistically sig- nificant, postoperative compliance (OR¼0.95, p¼.063) and preadmission patient education (OR¼0.29, p¼.066) were highly suggestive for reducing DGE, and postoperative epidural anesthesia (OR¼7.5, p¼.054) and post- operative compliance (OR¼1.2, p¼.056) for contributing to 30-day survival. Conclusion: These results show that not only do aggregated compliance measures contribute to better patient care in PD patients, but that the specific pathway items are meaningful as well. Further, this study high- lights that a variety of pathway items are significant contributors, which suggests that reducing or consolidating elements in an ERAS ® protocol could perhaps be detrimental to patient outcomes. Disclosure of Interest: None declared. OR12 POPULATION-BASED EVALUATION OF ERAS IMPLEMENTATION IN MICHIGAN, USA Scott Regenbogen 1, 2, * , Anne Cain-Nielsen 2 , Emily George 3 , Edward Norton 2, 4,5 . on behalf of Michigan Value Collaborative, Michigan Surgical Quality Collaborative. 1 Division of Colorectal Surgery, University of Michigan, Ann Arbor, United States; 2 Center for Healthcare Outcomes and Policy, University of Michigan, Ann Arbor, United States; 3 Department of Surgery, University of Michigan, Ann Arbor, United States; 4 Department of Health Management and Policy, University of Michigan, Ann Arbor, United States; 5 Department of Economics, University of Michigan, Ann Arbor, United States * Corresponding author. Division of Colorectal Surgery, University of Michigan, Ann Arbor, United States. Objectives: To compare clinical and economic outcomes after colectomy in ERAS hospitals against trends in matched hospitals without ERAS. ERAS outcomes in high-volume specialty centers are well described, but uptake beyond such institutions is limited in the US. Studies have rarely differ- entiated the effects of ERAS from unrelated trends in perioperative care. In a statewide surgical collaborative actively promoting ERAS adoption, we evaluated implementation across a diverse set of hospitals, using an econometric approach to control for trends unrelated to ERAS. Methods: Using statewide data from the Michigan Surgical Quality Collaborative (outcomes registry) and Michigan Value Collaborative (episode-based claims registry), we identified hospitals that adopted ERAS programs for colectomy. We matched them against non-ERAS hospitals, according to bed size, teaching status, and case-mix-index. Using differ- ence-in-differences (DID) analysis, we compared trends in clinical out- comes and payments after colectomy between ERAS and control hospitals. Results: Of 72 hospitals, 16 (22%) had full ERAS programs. Pre-imple- mentation outcomes and trends were not different in ERAS and non-ERAS hospitals (length of stay [LOS] p¼0.53, readmissions p¼0.15). Overall, average postoperative LOS declined by approximately 1 day, with no Risk Factors No DGE (n¼92) Grade B/C DGE (n¼34) p-value Gatric Outlet Obstruction 2 7 <0.001 Chronic Pancreatitis 7 8 0.014 Irreversible Electroporation 12 2 0.256 Neoadjuvant Chemotherapy 24 4 0.180 Neoadjuvant XRT 5 1 0.571 Adenocarcinoma 68 21 0.184 Preoperative Weight Loss 61 23 0.887 Concomitant Vascular Resection 12 3 0.516 Intraoperative IVF>3000cc 46 19 0.558 Preoperative Opiate Use 32 13 0.720 Diabetes 27 9 0.751 Abstracts / Clinical Nutrition ESPEN 25 (2018) 166e209 170