Timing is everything: outcomes of 30,259 delayed cholecystectomies in New York State

Background: The timing of cholecystectomy in relation to outcomes has been debated. To our knowledge, there are no large population-based studies looking at outcomes and complications of delayed cholecystectomy [DC] (> 72 h after presentation). This study utilizes a statewide database to determine whether there are differences in patient outcomes for DC performed at 3–4 days, 5–6 days, and ≥ 7 days after presentation. Methods: The New York SPARCS database was used to identify adult patients presenting with a diagnosis of acute cholecystitis from 2005 to 2017. Patients aged < 18, those with missing identifier or procedure-date information, those who underwent early cholecystectomy < 72 h or upon readmission, were excluded. Patients undergoing DC at 3–4 days, 5–6 days, and ≥ 7 days were compared in terms of overall complications, hospital length of stay (LOS), 30-day readmissions/emergency department (ED) visits, and 30-day mortality. Results: 30,259 patients were identified. DCs were performed within 3–4 days (n = 19,845, 65.6%), 5–6 days (n = 6432, 21.3%), and ≥ 7 days (n = 3982, 13.2%). There was a stepwise deterioration in outcomes with increased delay to surgery (Fig. 1). When comparing 3–4 and ≥ 7 days, overall complications (OR = 0.418, 95% CI: 0.387–0.452), 30-day readmissions (OR = 0.609, 95% CI: 0.549–0.674), 30-day ED visits (OR = 0.697, 95% CI: 0.637–0.763), 30-day mortality (OR = 0.601, 95% CI: 0.400–0.904), and LOS (OR = 0.729, 95% CI: 0.710–0.748) were lower in the 3–4 day cohort. Conclusions: DC within 3–4 days is associated with fewer complications, readmissions and ED visits, and reduced LOS compared to DC at 5–6 or ≥ 7 days after presentation. In addition, 30-day mortality was also significantly different comparing 3–4 with ≥ 7-day cohorts. These data are important for guiding patients in the consent process and may point to choosing an earlier interval cholecystectomy for high-risk patients. Graphical abstract: [Figure not available: see fulltext.].