National Trends in Short-term Outcomes Following Non-emergent Surgery for Diverticular Disease

Abstract

Introduction: Elective surgery for diverticulitis has evolved over the last decade. We aimed to evaluate the impact of changing practice patterns on postoperative outcomes. We hypothesized that the increased use of laparoscopy, and other management changes, would correlate with a decrease in postoperative complications.

Methods: Patients undergoing non-emergent surgery for diverticulitis from 2005 to 2013 were selected from the National Surgical Quality Improvement Program (NSQIP) database. We compared patient demographics, comorbidities, and operative approach by year of operation using chi-square tests and investigated temporal trends in postoperative outcomes using univariate, trend, and multivariate analyses.

Results: The analytic cohort, which included 29,893 patients, had increasing rates of obesity, advanced age, and higher American Society of Anesthesiologists (ASA) class over the study period. The use of laparoscopy increased significantly from 48 % in 2005/2006 to 70 % in 2013 (p < 0.001), while the rate of stoma creation remained unchanged (10-12 %, p = 0.072). The absolute risk of any postoperative complication decreased by 5.8 % over the study period, driven primarily by a reduction in infectious complications. Year of operation was a significant independent predictor of fewer complications for 2011-2013.

Conclusion: Despite a trend towards increasing patient complexity, there has been a decline in postoperative morbidity following non-emergent surgery for diverticulitis. This trend coincides with the steadily increasing use of laparoscopy in this population.

Keywords: Colectomy; Diverticulitis; Laparoscopy; Outcomes.

Fig. 1

Flow diagram showing patient selection for inclusion in the study and creation of cohorts

Fig. 2

Baseline patient characteristics of each yearly cohort. a Percentage of patients who are ASA class 3 or 4 in each yearly cohort (chi-square, p < 0.001; Mantel-Haenszel, p < 0.001). b Percentage of patients aged 65+ years in each cohort (chi-square, p = 0.001; Mantel-Haenszel, p < 0.001). c Percentage of patients with BMI ≥ 30 in each cohort (chi-square, p < 0.001; Mantel-Haenszel, p < 0.001). d Percentage of patients with ICD-9 diagnosis of acute diverticulitis without hemorrhage (rather than diverticulosis without hemorrhage) in each yearly cohort (chi-square, p < 0.001; Mantel-Haenszel, p < 0.001)

Fig. 3

Operative approach characteristics across yearly cohorts. a Percentage of patients undergoing laparoscopic approach (chi-square, p < 0.001; Mantel-Haenszel, p < 0.001). b Percentage of patients receiving an ostomy (either ileostomy or colostomy, diverting or end) in each yearly cohort (chi-square, p = 0.072; Mantel-Haenszel, p = 0.058)

Fig. 4

Univariate analysis of 30-day outcomes following surgery for diverticular disease across yearly cohorts from 2005 to 2013

Fig. 5

Trends in composite endpoints from 2005 to 2013

Fig. 6

Results from multivariate analysis of outcomes controlling for gender, age, BMI, operative approach (laparoscopic vs open), diagnosis (diverticulitis vs diverticulosis), comorbidities, functional status, and ASA class. The 2005/2006 cohort is used as the reference group. Estimated odds ratios (indicated by the solid circles) and 95 % confidence intervals for the estimate are displayed

Fig. 7

Univariate analysis showing rates of any postoperative complication across time, stratified by the operative approach. While the rates of complications within each category of operative approach have remained relatively stable over time, the rates of complications in the overall study population (dashed line) have decreased significantly. The p values are for Mantel-Haenszel tests of trend