National Variation in Elective Colon Resection for Diverticular Disease.

Objective: This study aims to characterize the extent of geographic variation in elective sigmoid resection for diverticulitis and to identify factors associated with observed variation.
Introduction: National guidelines for treatment of recurrent diverticulitis fail to offer strong recommendations for or against surgical intervention. We hypothesize that healthcare market factors will be significantly associated with geographic variation in colon resection for diverticulitis, a discretionary surgical intervention.
Methods: We used Center for Medicare Services 100% inpatient Limited Data Set (LDS) files from January 2013 through September 2015 to calculate an observed to expected standardized colon resection ratio for each hospital referral region (HRR). We then analyzed patient, hospital-, and market-level factors associated with variation of colectomy. For each HRR, a Herfindahl-Hirschman index, a measure of market competition, was calculated.
Results: A total of 19,557 Medicare patients underwent an elective colon resection for diverticulitis at 2462 hospitals over the study period. Standardized colon resection ratios ranged from 0 in the Tuscaloosa HRR to 3.7 in the Royal Oak, MI HRR. Few patient factors were associated with variation, but a number of hospital factors (size, area, profit status, and critical access designation) all were associated with variation. In an analysis of market factors, increased surgeon density, and decreased market competition were associated with higher predicted rates of colon resection.
Conclusion: We observed pronounced variation (excess of 3-fold) in standardized colon resection ratios for recurrent diverticulitis. Surgeon density and hospital level factors were strongly associated with this variation and may be the main drivers of colonic resection for diverticular disease. Further investigation and stronger national guidelines are needed to optimize patient selection for colectomy.
Competing Interests: The authors declare no conflict of interests.
(Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.)