Bariatric Surgery and Inflammatory Bowel Disease: National Trends and Outcomes Associated with Procedural Sleeve Gastrectomy vs Historical Bariatric Surgery Among US Hospitalized Patients 2009-2020

Purpose: The association between bariatric surgery and IBD-related inpatient outcomes is not well characterized. We report, analyze, and compare inpatient trends and outcomes among encounters with a history of bariatric surgery (Hx-MBS) compared to those receiving bariatric surgery during index admission (PR-MBS) admitted from 2009 to 2020.

Methods: Retrospective cohort design: the 2009-2020 National Inpatient Sample (NIS) databases were used to identify hospital encounters with patients aged ≥ 18 years with a history of MBS (Hx-MBS) or with procedure coding indicating MBS procedure (PR-MBS) according to International Classification of Diseases, Ninth (ICD-9-CM/ ICD-9-PCS) or Tenth Revision (ICD-10-CM/ICD-10-PCS) Clinical Modification/Procedure Coding System during index admission (ICD-9-CM: V4586; ICD-10-CM: Z9884; ICD-9-PR: 4382, 4389; ICD-10-PR: 0DB64Z3, 0DB63ZZ). Pearson χ2 analysis, analysis of variance, multivariable regression analyses, and propensity matching on independent variables were conducted to analyze significant associations between variables and for primary outcome inflammatory bowel disease-related admission, and secondary outcomes: diagnosis of nonalcoholic steatohepatitis, nonalcoholic fatty liver disease, or chronic mesenteric ischemia during admission.

Results: We identified 3,365,784 (76.20%) Hx-MBS hospitalizations and 1,050,900 hospitalizations with PR-MBS (23.80%). Propensity score matching analysis demonstrated significantly higher odds of inflammatory bowel disease, and chronic mesenteric ischemia for Hx-MBS compared to PR-MBS, and significantly lower odds of nonalcoholic steatohepatitis and nonalcoholic fatty liver disease for Hx-MBS compared to PR-MBS.

Conclusion: In our study, Hx-MBS was associated with significantly increased odds of inflammatory bowel disease and other GI pathologies compared to matched controls. The mechanism by which this occurs is unclear. Additional studies are needed to examine these findings.