Predictors and Trends of Hemorrhage After Bariatric Surgery Using Two Large National Datasets: A Retrospective Observational Study

Introduction: The National Inpatient Sample (NIS) is a comprehensive representative database for inpatient hospitalizations; the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQIP) database aggregates surgical outcomes from nationally accredited bariatric surgical programs. There has been no comparison of postoperative hemorrhage rates nor a comparison of predictors of hemorrhage between the two databases. The aim of this study was to compare trends and predictors of significant hemorrhage after bariatric surgery using two national databases. Methods: The core NIS files of the Healthcare Cost and Utilization Project and the MBSAQIP Participant Use Data Files from 2016 to 2021 were concatenated, and vertical sleeve gastrectomy and Roux-en-Y gastric bypass procedures were included. Trends in a 3-d postoperative hemorrhage requiring blood transfusion were derived from both datasets. Demographic, comorbidity, and other health factors were analyzed to construct multivariable logistic regression models. Results: Briefly, 204,866 and 1,029,979 patients were identified in the NIS and MBSAQIP, respectively, and 3-d inpatient blood transfusion rates were 0.48% and 0.64%. Anticoagulants, Black race, diabetes, and increased age were associated with higher risk of a blood transfusion in both datasets. Vertical sleeve gastrectomy had a lower transfusion risk than Roux-en-Y gastric bypass (NIS: odds ratio: 0.62; 95% confidence interval [0.53, 0.71], P < 0.05; MBSAQIP: odds ratio: 0.52 [0.49, 0.55], P < 0.05). The NIS reported consistently lower annual 3-d transfusion rates relative to the MBSAQIP. Conclusions: Health care professionals should be aware of the associated risk factors for blood transfusion after bariatric surgery. Most blood transfusions occurred during the initial hospitalization.