Cholecystectomy or non-operative management for cholecystitis in elderly patients: An analysis based on hospital practice patterns

Abstract

Background and Aims: Early cholecystectomy improves outcomes in acute cholecystitis, yet surgeons often hesitate to operate on elderly and multimorbid patients, and prospective data in this group are lacking. This cohort study compared outcomes across hospitals with differing operative policies to estimate the effect of surgery among the elderly. Methods: Retrospective data using the Swedish National Inpatient Register, including 12,481 patients aged ⩾ 70 years treated for cholecystitis between 2015 and 2019, were analyzed. Outcomes included all-cause and gallstone-related mortality, length of stay related to gallstone disease during the index admission, and within 1 and 3 years. Hospitals and counties were categorized into quartiles based on cholecystectomy rates (Q1 = lowest, Q4 = highest). Survival was assessed using Cox proportional hazards regression. Length of stay was analyzed with linear regression. Results: Patients treated at Q1 hospitals were slightly older but had similar comorbidity burdens compared to Q4 hospitals. Surgical rates ranged from 25% to 60%. Overall mortality was lower in Q4 counties, but not significantly different at hospital level (hazard ratio (HR) = 0.86 95% confidence interval (CI) = 0.77–0.96). Gallstone-related mortality did not differ across quartiles. Although the index admission length of stay was similar, cumulative length of stay at 1 and 3 years was shorter for patients treated at Q4 hospitals compared to Q1 (0.87 fewer days at 3 years; 95% CI = 0.39–1.34). Conclusions: Hospitals with higher cholecystectomy rates did not demonstrate lower mortality, but regions with higher surgical rates showed improved survival. Higher operative rates were associated with fewer recurrences and reduced cumulative length of stay, suggesting that increased use of cholecystectomy benefits both patients and healthcare systems. Clinical trial registration: Not applicable.