Obesity remains an epidemic in the United States, with greater than 30% of the adult population living with obesity [1]. This national trend has also led to an increase in associated medical comorbidities, such as hypertension, diabetes, and congestive heart failure [2]. Bariatric and metabolic surgery has been established as an effective and durable treatment for obesity and associated medical comorbidities, and it also results in improvement in quality of life [3,4,5,6,7,8,9]. Additionally, the safety of bariatric surgery has continued to improve over the last several decades [10].

In the United States, the prevalence of obesity remains higher among Black patients, compared with their White and Asian counterparts [11]. Despite the high prevalence of obesity, bariatric surgery remains underutilized throughout the nation but particularly among Black patients [12]. Furthermore, racial disparities do exist, with racial minorities having worse perioperative outcomes following bariatric surgery compared with their White counterparts. In a national study from 2015, Sheka et al. demonstrated that Black patients have higher readmissions and longer hospital length of stay (LOS) following laparoscopic Roux-en-Y gastric bypass (RYGB) and sleeve gastrectomy (SG) [13]. Edwards et al. also found two-fold higher 30-day mortality rates for non-Hispanic Black patients undergoing bariatric surgery, compared with non-Hispanic White patients [14].

However, it remains unclear whether recognition of these disparities over the last decade has led to any improvement in outcomes. Therefore, the purpose of this study was to perform a contemporary assessment of perioperative outcomes following bariatric surgery at the national level. We hypothesized that racial disparities for bariatric surgical care persist in the United States and that further work is needed to understand and address these disparities.

Materials and methods

Study population

A retrospective study was performed of primary SG and RYGB patients reported to the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQIP) from January 1, 2015 through December 31, 2021. Data was obtained from the MBSAQIP participant use data file (PUF), based on all programs participating in the MBSAQIP accreditation. This study was approved by the Institutional Review Board.

The bariatric surgery patients identified in the MBSAQIP dataset were then stratified by race (White and Black). Exclusion criteria included: revisional bariatric surgery, bariatric surgery other than SG or RYGB, and patients with missing race/ethnicity or race listed as any other than White or Black race.

Statistical analyses

Within the overall cohort, descriptive statistics by race (White and Black) were performed over the study period. Patient demographics and characteristics were captured including age, body mass index (BMI) at time of operation, sex (male/female), operation performed (SG or RYGB), American Society of Anesthesia (ASA) classification as well as other patient comorbidities (Table 1). Categorical variables were compared using the chi-squared test, and continuous variables were compared using Wilcoxon rank sum, Student’s t-test, analysis of variance or Kruskal–Wallis rank sum tests, depending upon distribution. Statistical significance was defined as p < 0.05. All analyses for this study were completed in SAS 9.4 (Cary, NC).

Table 1 Baseline demographic and clinical characteristics of all patients undergoing bariatric surgery between 2015 and 2021 and stratified by Black and White race (2015–2021)
Full size table

Matched analyses

A matched case–control study was then performed. Black and White bariatric surgery patients were propensity-score matched one-to-one based on the following patient and surgical characteristics: age, BMI at time of operation, sex (male/female), operation performed (SG or RYGB), ASA classification, preoperative functional health status (independent/partially dependent), hypertension, hyperlipidemia, preoperative renal insufficiency, preoperative diabetes mellitus (DM), preoperative obstructive sleep apnea, preoperative history of chronic obstructive pulmonary disease (COPD), steroid/immunosuppressant use, therapeutic anticoagulation, and history of pulmonary thromboembolism (PTE). Perioperative outcomes were then compared by race (Black and White).

Primary and secondary outcomes

The primary outcome was 30-day mortality following primary bariatric surgery. Secondary outcomes included 30-day hospital readmissions, hospital LOS, reoperations, and postoperative complications. Other postoperative complications included acute renal failure, PTE, intraoperative or postoperative cardiac arrest, stroke/cerebrovascular accident (CVA), progressive renal insufficiency, transfusion requirements within 72 h of surgery start time, wound disruption, surgical site infections, and postoperative pneumonia.

Subgroup analyses

Among the matched patients, additional subgroup analyses were performed. The patients were stratified by procedure performed (SG or RYGB). Within these subgroups, primary and secondary outcomes were assessed by race (Black and White).

Results

Overall, there were 193,071 Black and 645,224 White primary bariatric patients included in the initial study (Table 1). Black patients were younger [median age of 42 years; interquartile range (IQR) 34.4–50.5] than White patients (median age 45.4 years; IQR 36.5–54.7; p < 0.001). Black patients also had slightly higher median BMI of 45 kg/m2 (IQR 41–51), compared with median BMI of 44 kg/m2 (IQR 40–49) among Whites (p < 0.001). Black patients were more commonly female (87.25% vs. 78.51%; p < 0.001). Sleeve gastrectomy (SG) accounted for 78.95% of bariatric surgery among Black patients, compared with 70.66% of White patients (p < 0.001). Preoperative diagnoses of hypertension and renal insufficiency were more prevalent among Black patients, while White patients had higher rates of preoperative hyperlipidemia (Table 1). Prior history of smoking and history of MI was more commonly reported among White patients compared with their Black counterparts (Table 1). However, Black patients had higher reported histories of preoperative PTE (1.55% vs. 1.44%; p < 0.001).

Matched case- control: overall

A total of 219,566 Black (n = 109,783) and White (n = 109,783) bariatric patients were matched and included in the case–control. The two groups were well matched (Table 2). Overall, Black patients were found to have higher rates of 30-day mortality compared with White patients (0.02% vs. 0.01%; p = 0.03) and readmission (3.68% vs. 2.65%; p < 0.001). The most commonly reported source for readmission was nausea/vomiting or abdominal pain for all bariatric patients. This was reported more often as the reason for readmission for Black patients (53.6% of readmissions vs. 43.4%; p < 0.0001). Additional causes of readmissions were similar between the groups, including risks of DVT/PE (about 8–9% of readmissions). There were also more Black patients readmitted for obstructions, internal hernias, or strictures (7.6% vs. 6.4%; p = 0.047), while more White patients had gastrointestinal leak or ulceration as the cause of readmission (8.3%, n = 242 vs. 3.2%, n = 129, p < 0.0001). Stroke readmissions accounted for 20 readmissions among Black patients and only 3 White patients. 368 Black patients were admitted for PTE/DVT, compared to only 237 White patients.

Table 2 Patient Characteristics with the Matched Case–Control Study, by race (comparing Black and White patients), (2015–2021)
Full size table

Median length of stay was 1.0 day (IQR 1.0–2.0) regardless of race, and there were no differences in reoperation rates (Table 3). However, there was a higher postoperative PTE rate for Blacks compared with Whites (0.16% vs. 0.08%; p < 0.001) (Table 3). There was also a higher rate of postoperative stroke/CVA among Black patients (0.02% vs. 0.01%; p = 0.01) (Table 3). Likewise, Black patients had higher rates of progressive renal insufficiency (0.05% vs. 0.02%; p < 0.001) (Table 3). The differences in perioperative outcomes stratified by race persisted over the study period (Fig. 1). White patients had higher rates of wound disruption and surgical site infections (Table 3).

Fig. 1
figure 1

a Mortality rates for bariatric patients by race. b Readmissions for bariatric patients by race. c Postoperative Pulmonary Thromboembolism for bariatric patients by race

Full size image
Table 3 Postoperative outcomes among SG and RYGB patients, by race among matched case–control
Full size table

Subgroup analyses

Matched SG

Within the matched case–control, a total of 181,234 patients underwent SG, including 90,617 Black and White patients. Among SG patients, Black patients were again found to have higher rates of 30-day mortality and 30-day readmission (Table 4). Median LOS remained at 1 day for both Black and White patients. Black patients were also found to have higher incidences of progressive renal insufficiency and PTE (Table 4). Black SG patients also had higher rates of perioperative or postoperative transfusion requirements (0.51% vs. 0.44%; p = 0.03). White patients continued to have higher surgical site infections in the SG cohort (Table 4).

Table 4 Postoperative outcomes among SG patients, by race among matched case–control
Full size table

Matched RYGB

Among the matched patients, 38,332 underwent RYGB (19,166 Black and White patients). There were no statistically significant differences in postoperative mortality among RYGB patients by race, but a higher readmission rate was found for Black patients (5.45% vs. 4.87%; p = 0.01). Compared with White RYGB patients, Black RYGB patients also had longer median LOS (2.0 vs. 1.0 day; p < 0.001) and higher rates of acute renal failure (0.11% vs. 0.03%; p = 0.002). Black patients who underwent RYGB were found to have higher rates of stroke/CVA, intraoperative or postoperative cardiac arrest requiring cardiopulmonary resuscitation and PTEs (Table 5). White patients continued to have higher rates of surgical site infections among RYGB patients (Table 5).

Table 5 Postoperative Outcomes among RYGB patients, by Race among matched case–control
Full size table

Discussion

In this study of MBSAQIP data from 2015 through 2021, Black patients accounted for < 25% of primary bariatric surgery performed in the United States, despite the known higher prevalence of obesity among Black patients. Within the overall propensity score matched case–control, Black patients were found to have higher rates of perioperative mortality and hospital readmissions following bariatric surgery. Black patients also had higher rates of other perioperative complications including PTE, CVA, and progressive renal insufficiency following bariatric surgery. Within the subgroup analyses by procedure performed, Black patients undergoing SG had higher rates of postoperative blood transfusion requirements than White patients. Additionally, among patients who underwent RYGB, Black patients had longer LOS and higher rates of acute renal failure, stroke, and cardiac arrest.

The results of this study are consistent with previously published studies from the last decade, with continued evidence of low rates of complications following bariatric surgery [15, 16]. However, this study also demonstrates that worse perioperative outcomes persist for Black bariatric patients, compared with White bariatric patients. Despite the previous identification of racial disparities in postoperative outcomes for Black patients [13, 14, 17, 18], it remains largely unknown which factor(s) are contributing to the observed differences in postoperative outcomes for racial minority patients.

Socioeconomic factors have been demonstrated to be barriers to bariatric surgery [19]. Nonetheless, even with access to bariatric surgical care, racial minority patients continue to have worse postoperative outcomes. While this national study was not able to account for socioecological factors that may also contribute to these disparities, the identification of persistent disparities highlights the need to promote equity in healthcare for bariatric patients. Interestingly, Liu et al. found that Medicaid status and other social determinants of health at the neighborhood level were not associated with differences in long term outcomes such as weight loss at one year following bariatric surgery [20]. This group also found that patient-centered factors were more closely associated with 1-year outcomes than neighborhood level factors [21]. However, the relationship between these patient and community-level factors at the national level and short-term outcomes following bariatric surgery remains unclear.

Social determinants of health, including poor health literacy, have also been shown to impact access to bariatric surgery and outcomes following surgery. In their single institution study, Mahoney et al. found that lower health literacy levels were associated with worse long-term outcomes, including less weight loss following bariatric surgery [22]. The authors also previously published that patients with educational levels less than high school graduation had three-fold increased risks of hospital visits following bariatric surgery [22]. Furthermore, lower health literacy levels have been historically found among racial minority patients, further exacerbating inequalities in access to quality surgical care [23]. While lower health literacy may not directly account for higher rates of mortality for Black patients in this study, health literacy may play a role in the observed increased readmission rates for Black patients following bariatric surgery. As many of the readmissions, particularly among Black patients, were due to nausea/vomiting and abdominal pain, there may be opportunities for improvement in both the inpatient and outpatient settings to enhance symptom control and management. Importantly, this also highlights an actionable opportunity for patient-centered interventions, such as patient navigation systems [24], in order to provide more resources and bridge the health literacy gap for vulnerable patient populations. Likewise, improved access to primary care for racial minority patients may lead to better preoperative screenings and diagnoses of cardiovascular disease, which may also improve disparities in postoperative outcomes. However, future prospective studies are also needed among racially diverse bariatric patient populations in order to further identify barriers to equity in perioperative and surgical care so that they may be ameliorated.

This study is not without limitations. Due to the nature of the national dataset provided by the MBSAQIP, there is a lack of granularity to help elucidate the factors that may be contributing to worse outcomes in Black patients. For example, socioecological factors that may contribute to the observed findings are not reported by MBSAQIP. There may also be inconsistencies in reporting and capturing perioperative information as presented by the MBSAQIP database. Additionally, the data is limited to programs participating in the MBSAQIP accreditation and therefore may not be representative of the national population as a whole. Despite attempting to mitigate confounding bias in the study through propensity score matching, it is possible that residual confounding variables exist. The data also does not include information regarding patients’ access to primary care, which may influence postoperative outcomes, particularly those related to perioperative cardiovascular risks. Similarly, the study is not able to account for underlying differences such as under-treatment or under-diagnosis of hypertension, or renal disease which may be more prevalent among Black patients. Although prior work has demonstrated a relationship between health literacy and perioperative outcomes, unfortunately this study was not able to account for patients’ health literacy. However, we have several on-going studies that are investigating this knowledge gap. Finally, the study was limited to primary SG and RYGB patients and therefore does not account for the outcomes following other metabolic and bariatric operations or revisional surgery.

In summary, this national study demonstrated that Black primary bariatric surgery patients in the United States continue to have higher mortality and readmission rates and worse perioperative outcomes than their White counterparts. Socioecological factors may be driving these observed disparities in perioperative outcomes. Further work must be done to in order to determine contributing factors and effect improvement in outcomes in metabolic and bariatric surgical care for racial minority patients.