The Decline of Elective Colectomy Following Diverticulitis

BACKGROUND: The indications for interval elective colectomy following diverticulitis are unclear; evidence lends increasing support for nonoperative management. OBJECTIVE: This study aims to evaluate the temporal trends in the use of elective colectomy following diverticulitis. DESIGN: This is a population-based retrospective cohort study using administrative discharge data. SETTING: This study was conducted in Ontario, Canada. PATIENTS: Patients who had had an episode of diverticulitis managed nonoperatively and were eligible for elective colectomy, from 2002 to 2012, were selected. MAIN OUTCOME MEASURES: Changes in the proportion of patients who undergo elective colectomy following an episode of diverticulitis treated nonoperatively were evaluated. Cochran-Armitage was used to test for trends; adjusted analysis was performed by using multivariable logistic regression with generalized estimating equations. RESULTS: A total of 14,124 patients were admitted with an episode of diverticulitis and treated nonoperatively, making them eligible for interval elective colectomy. Median follow-up was 3.9 years (maximum, 10; interquartile range, 1.7-6.4). Overall, 1342 (9.5%) patients underwent elective colectomy; 33% of these colectomies were performed laparoscopically, and 7.5% patients received an ostomy. In-hospital mortality was 0.2%. The majority (76%) of elective operations were performed within 1 year of discharge (median, 160 days; interquartile range, 88-346). The proportion of patients undergoing elective colectomy within 1 year of discharge declined from 9.6% of patients in 2002 to 3.9% by 2011 (p < 0.001). The decline was most pronounced in patients <50 years of age (from 17% to 5%), and those with complicated disease (from 28% to 8%) (all p < 0.001). In multivariable regression, younger age, lower medical comorbidity, complicated disease, and early readmission were associated with elective colectomy. After adjusting for changes in patient characteristics, the odds of elective surgery decreased by 0.93 per annum (adjusted OR; 95% CI, 0.90-0.95). LIMITATIONS: Administrative health databases contain limited clinical detail; the rationale for elective surgery was not available. CONCLUSIONS: Consistent with evolving practice guidelines, there has been a decrease in the use of elective colectomy following an episode of diverticulitis.