Predictors of 30-day Readmission After Colectomy

Bindhu Oommen, MD, MPH, Samuel W Ross, MD, MPH, Joel F Bradley, MD, Kristopher B Williams, MD, Amanda L Walters, MS, Amy E Lincourt, PhD, MBA, Brant T Heniford, MD, Vedra A Augenstein, MD. Carolinas Medical Center, Department of Surgery, Division of Gastrointestinal & Minimally Invasive Surgery, Charlotte, NC.

INTRODUCTION: Hospital readmission after surgery is a strong indicator of surgical performance for payers and contributes to rising healthcare costs. We sought to identify risk factors for readmission after elective colectomy.

METHODS AND PROCEDURES: The American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database was queried for colectomy by CPT code. Inclusion criteria were elective colon resection with 30-day readmission data (available since 2011). Bivariate analysis was performed to determine variables that were potentially associated with readmission. Multivariate regression was then performed using those factors that correlated with readmission.

RESULTS: Of the 16,792 colectomies that met inclusion criteria, 1,989 colectomies (11.8%) were readmitted within 30 days; 52.8% were performed laparoscopically and 47.2% open. Of these 52% were female and 87% were Caucasian. While mean age (59.0±15.8 vs 60.6 ±14.7 years;p<0.001) and BMI (28.6 ±7.1 vs 28.2 ±6.7 kg/m2;p=0.046) were statistically different in the readmission group, they were not clinically different. Patients readmitted were more likely to have the following pre-operative conditions: diabetes (13.3% vs 11.6%;p=0.014), tobacco abuse (13.7% vs 11.8%;p<0.001), dyspnea (14.1% vs 11.6%;p=0.010), dependent functional status (16.3% vs 11.5%;p=0.015), COPD (14.9% vs 11.7%; p=0.006), ascites (17.9% vs 11.8%;p=0.037), CHF (21.0% vs 11.8%;p=0.011), dialysis–dependent renal failure (21.8% vs 11.8%;p=0.006), disseminated cancer (14.7% vs 11.7%;p=0.006), open wound/wound infection (16.1% vs 11.8%;p =0.035), chronic steroid use (19.1% vs 11.3%;p<0.001), >10% loss of body weight (14.9% vs 11.7%;p=0.011), bleeding disorders (17.5% vs 11.7%;p<0.001), ASA class IV or IV (17.0% vs 11.7%;p<0.001).
Postoperative complications were also associated with readmission (p<0.001 except where indicated): superficial surgical site infection (SSI) (24.3% vs 10.9%), deep SSI (47.5% vs 11.3%), intra-abdominal abscess (54.4% vs 9.8%), wound disruption (38.0% vs 11.5%), pneumonia (20.9% vs 11.7%), pulmonary embolism (46.2% vs 11.6%), failure to wean from ventilator (16.9% vs 11.8%;p=0.012), progressive renal insufficiency (45.8% vs 11.63%), acute renal failure (26.7% vs 11.8%), urinary tract infection (29.0% vs 11.3%), stroke (31.4% vs 11.8%), myocardial infarction (31.0% vs 11.7%), perioperative transfusion (18.3% vs 11.1%), deep vein thrombosis (44.7% vs 11.5%), sepsis (46.5% vs 10.3%), septic shock (23.6% vs 11.7%), reoperation (44.6% vs 10.0%). Comparing the readmission to non-readmission groups, the overall minor complication rate was 27.9% vs 6.8%, and major complication rate was 36.3% vs 9.7%. Mean operative time (207.9 ±113.2 min vs 180.2 ±99.5 min), mean length of stay (7.6 ±12.6 days vs 7.0 ±8.6 days), and mean time to discharge after surgery (6.9 ±4.3 vs. 6.6 ±6.3 days) were associated with readmission (p<0.001 for all). Readmission was higher among patients with open colectomies compared to laparoscopic (14.2% vs 9.7%;p<0.001). Multivariate regression (Table) demonstrated that age, ASA class, diabetes, smoking, CHF, chronic steroid use, bleeding disorders, open approach, OR time, and time to discharge are independently associated with 30-day readmission.

CONCLUSIONS: Quality improvement efforts aimed at targeting preoperative, operative, and postoperative risk factors have the potential to significantly improve patient outcomes, readmission rates, and decrease healthcare costs.

View Poster

« Return to SAGES 2014 abstract archive