Skip to main content

Risk factors for abdominal reoperations in bariatric patients

Elnahas AI, Reid JN, Lam M, Doumouras AG, Anvari M, Schlachta CM, Alkhamesi NA, Hawel J, Urbach DR. Surg Obes Relat Dis. 2022; 18(2):233-40. Epub 2021 Oct 23. DOI:

Background — With a growing bariatric population, a better understanding of the patient and health provider-related factors associated with later reoperations could help providers enhance follow up and develop reliable benchmarking targets.

Objective — To investigate the patient and provider-related risk factors associated with abdominal reoperations in bariatric patients.

Setting — This is a cohort study using data from a large clinical registry of Ontario bariatric patients between 2010 and 2016.

Methods — A multi-level mixed effect logistic regression model using hospital and surgeon identifiers as random effects was performed to adjust for clustering of patients. The primary outcome was any abdominal operation performed within two years of primary bariatric surgery.

Results — Among a cohort of 10,946 bariatric patients (86.6% receiving gastric bypass surgery), 15.8% underwent an abdominal operation within two years and about a third of these were urgent. The multi-level analysis demonstrated that 98% of patient variation among reoperations was a result of patient characteristics rather than disparities between surgeons or center experience. Type of procedure was not a significant factor after adjustment for surgeon and hospital level experience (OR 0.85, 95% CI 0.70-1.03). Concurrent abdominal wall (OR 2.40, 95% CI 1.26 – 4.59), hiatal hernia repairs (OR 1.29, 95% CI 1.02-1.62) and previously higher health care users (OR 1.30, 95% CI 1.15-1.46) were most significantly associated with reoperations.

Conclusions — Reoperations are significantly more common among certain bariatric patients, especially those undergoing concurrent hernia procedures. Reoperations were not associated with provider-related factors and may not be a suitable target for health provider benchmarking.