eP710 - Factors Associated with Prolonged Length of Stay After Laparoscopic Colorectal Resection Within an Enhanced Recovery Protocol

Purpose/Background: Enhanced recovery protocols (ERPs) have contributed greatly to optimizing patients perioperatively and significantly shortened lengths of stay (LOS) after surgery. We sought to identify factors that may help to predict patients who may require longer admissions despite the use of ERPs.Methods/Interventions: All patients undergoing laparoscopic colorectal resection with primary anastomosis and no diversion were identified from a prospectively maintained database, beginning in 2018 when standardized ERP was adopted at our institution. Median LOS for all segmental colectomy from the American College of Surgeons National Surgical Quality Improvement Program Database (ACS-NSQIP) for the years 2018-2020 (the last year that data is available) was used to classify patients who underwent surgery at our institution into those with LOS shorter (short LOS or SLOS group) and longer (long LOS or LLOS group) than the NSQIP LOS median. Demographic data as well as pre- and intra-operative clinical characteristics and 30-day postoperative outcomes were obtained from the prospective database. Data were assessed using chi-squared test for univariate and linear regression for multivariate analysis; all pre- and intra-operative variables with P < 0.2 were included in the multivariate model.Results/Outcomes: Median LOS for NSQIP patients was 3 days (interquartile range IQR 2-5) and the same for our cohort, although with a lower IQR of 2-4 days (P < 0.0001). Based on the NSQIP LOS, there were 343 SLOS and 165 LLOS patients in our cohort. SLOS were younger with median age 63 (IQR 52-72) vs 71 years (IQR 58-79, P < 0.0001) while BMI (median 27.1 kg/m2 for SLOS vs 26.9 for LLOS, P=0.99) and gender (SLOS 55% female, LLOS 51% female, P=0.47) were comparable. Preoperatively, more SLOS patients used alcohol (12 vs 5%, P=0.03) but had lower rates of other comorbidities: pneumonia (0 vs 2.4%, P=0.02); cardiac comorbidity (6 vs 15%, P=0.003); prior stroke (2 vs 7%, P=0.008); metastatic cancer (1 vs 6%, P=0.01); and prior abdominal surgery (30 vs 40%, P=0.03). SLOS had lower American Society of Anestheiolosigsts (ASA) class, with only 31% being ASA class 3 or higher compared to 58% of LLOS (P < 0.0001). SLOS was more likely to undergo surgery for diverticulitis (35 vs 23%) whereas LLOS was more likely to undergo surgery for cancer (44 for SLOS vs 52% for LLOS) or inflammatory bowel disease (6 for SLOS vs 14% for LLOS, P=0.01). SLOS had lower conversion to open surgery (4 vs 19%, P < 0.0001), intraoperative complications (3 vs 10%, P=0.005), and estimated blood loss (median 35 vs 50 mL, P=0.005), with more clean-contaminated wounds (83 vs 77%, P=0.02). On multivariate analysis controlling for independent variables with P < 0.2, the factors predictive of longer LOS were age, preoperative transfusion, conversion to open surgery, and wound class. Age, hypertension, and preoperative transfusion were associated with higher likelihood of requiring reoperation, while cardiac comorbidity was associated with readmission.Conclusion/

Discussion: Perioperative factors such as the need for preoperative transfusion, conversion to open surgery, and wound class are associated with longer LOS for laparoscopic colectomy when controlling for other factors. Awareness of this information will help during inpatient stay and discharge planning as well as counseling of patients undergoing laparoscopic colectomy procedures.Multivariate Regression on Length of Stay for Laparoscopic Segmental ColectomyOnly significant variables shown; included in model but not shown on table: alcohol use, smoking, diabetes, preoperative pneumonia, cardiac comorbidity (including heart failure, angina, myocardial infarction, percutaneous cardiac intervention, or cardiac surgery), hypertension, prior stroke, prior transient ischemic attack, disseminated cancer, prior abdominal surgery, primary diagnosis, intraoperative complication, estimated blood loss, ASA class, and drain left at end of surgery