Hospital utilization 4 years after bariatric surgery: sleeve gastrectomy versus Roux-en-Y gastric bypass

Abstract

BackgroundWhile sleeve gastrectomy (SG) has lower perioperative risk compared with Roux-en-Y gastric bypass (RYGB), long-term data about their differential impact on overall health are unclear. Hospital use after bariatric surgery is an important parameter for improving peri- and postoperative care. ObjectiveThis present study was aimed to compare SG and RYGB in terms of their effect on long-term hospital-based healthcare utilization. SettingMulticenter, statewide database. MethodsA retrospective cohort study of adult patients who underwent SG and RYGB between 2009 and 2011, with follow-up until 2015 and 2-year presurgery information. Propensity score–matched SG and RYGB groups were created using preoperative demographic characteristics, co-morbidities, and presurgery hospital use, measured by cumulative length of stay (LOS) and frequency of emergency department visits. Postsurgery yearly LOS, incidence of hospital visits, and the reason for the visit were compared. Primary outcomes included postoperative hospital visits during years 1 to 4 after bariatric surgery and cumulative LOS. Secondary outcomes included specific reasons for hospital use. ResultsThere were 3540 SG and 13,587 RYGB patients, whose mean (95% confidence interval [CI]) LOS was 1.3 (1.3–1.4), .9 (.8–1), 1 (.9–1.1), and 1.2 (1–1.3) days at years 1 through 4, respectively. Postoperative yearly LOS was similar between the 2 propensity-matched groups. The risk of hospitalizations (odd ratio .73, 95% CI .64–.84, P < .0001) and emergency department visits (odds ratio .84, 95% CI .75–.95, P = .005) was significantly lower for SG, during the first postoperative year. The reverse was seen at the fourth postoperative year, with higher risk of emergency department use after SG (odds ratio 1.16, 95% CI 1.01–1.33, P = .035). ConclusionPostoperative 4-year hospital utilization remains low for both SG and RYGB. The previously established lower early perioperative risk of SG was not appreciated for longer-term hospital use compared with RYGB.