Importance: The outcomes of bariatric surgery vary considerably across patients, but the association of race with these measures remains unclear.
Objective: To examine the association of race on perioperative and 1-year outcomes of bariatric surgery.
Design, setting, and participants: Propensity score matching was used to assemble cohorts of black and white patients from the Michigan Bariatric Surgery Collaborative who underwent a primary bariatric operation (Roux-en-Y gastric bypass, sleeve gastrectomy, or adjustable gastric banding) between June 2006 and January 2017. Cohorts were balanced on baseline characteristics and procedure. Conditional fixed-effects models were used to evaluate the association of race on outcomes within hospitals and surgeons. Data analysis occurred from June 2006 through August 2018.
Main outcomes and measures: Thirty-day complications and health care resource utilization measures, as well as 1-year weight loss, comorbidity remission, quality of life, and satisfaction.
Results: In each group, 7105 patients were included. Black patients had a higher rate of any complication (628 [8.8%] vs 481 [6.8%]; adjusted odds ratio, 1.33 [95% CI, 1.17-1.51]; P = .02), but there were no significant differences in the rates of serious complications (178 [2.5%] vs 135 [1.9%]; adjusted odds ratio, 1.32 [95% CI, 1.05-1.66]; P = .29) or mortality (5 [0.10%] vs 7 [0.10%]; adjusted odds ratio, 0.73 [95% CI, 0.23-2.31]; P = .54). Black patients had a greater length of stay (mean [SD], 2.2 [3.0] days vs 1.9 [1.7] days; adjusted odds ratio, 0.30 [95% CI, 0.20-0.40]; P < .001), as well as a higher rate of emergency department visits (541 [11.6%] vs 826 [7.6%]; adjusted odds ratio, 1.60 [95% CI, 1.43-1.79]; P < .001) and readmissions (414 [5.8%] vs 245 [3.5%]; adjusted odds ratio, 1.73 [95% CI, 1.47-2.03]; P < .001). At 1 year, black patients had lower mean total body weight loss and as a percentage of weight (32.0 kg [26%]; vs 38.3 kg [29%]; P < .001) and this held true across procedures. Remission of hypertension was lower for black patients (564 [40.0%] vs 1096 [56.0%]; P < .001), but the rate of sleep apnea remission (467 [62.6%] vs 615 [56.1%]; P = .005) and gastroesophageal reflux disease (309 [78.6%] vs 453 [75.4%]; P = .049) were higher. There were no significant differences in remission of diabetes with insulin dependence, diabetes without insulin dependence,or hyperlipidemia hyperlipidemia. Fewer black patients than white patients reported a good or very good quality of life (1379 [87.2%] vs 2133 [90.4%]; P = .002) and being very satisfied with surgery (1908 [78.4%] vs 2895 [84.2%]; P < .001) at 1 year.
Conclusions and relevance: Black patients undergoing bariatric surgery in Michigan had significantly higher rates of 30-day complications and resource utilization and experienced lower weight loss at 1 year than a matched cohort of white patients. While sleep apnea and gastroesophageal reflux disease remission were higher and hypertension remission lower in black patients, comorbidity remission was otherwise similar between matched cohorts. Racial and cultural differences among patients should be considered when designing strategies to optimize outcomes with bariatric surgery.