Readmissions following bariatric surgery are decreasing over time, but racial and ethnic disparities persist, according to new research. The findings were presented at the 2021 annual meeting of the American Society for Metabolic and Bariatric Surgery by Sean O’Neill, MD, PhD, a fellow in the Advanced MIS/Bariatric Fellowship Program at The Ohio State University Wexner Medical Center, in Columbus.
According to Sabrena Noria, MD, PhD, the surgical director of the Comprehensive Weight Management, Metabolic/Bariatric Surgery Program at The Ohio State University Wexner Medical Center, and principal investigator of the study, while bariatric surgery is safe and effective, postoperative readmissions persist, ranging from 1% to 13%, and disproportionately affect patients from underrepresented backgrounds and/or a lower socioeconomic bracket.
“Our previous institutional-level analysis demonstrated several predictors of overall readmissions, including prolonged length of stay, public insurance and younger age. However, when we stratified readmissions by urgency, length of stay predicted urgent reasons for readmission, like a leak or bleed, while public insurance and younger age predicted nonurgent reasons such as dehydration,” Dr. Noria said. “However, given that this study was too small to detect differences in race/ethnicity, we sought to extend our previous work by examining the readmissions, stratified by urgency, in a national sample.”
To shed light on the issue, the researchers analyzed the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Project (MBSAQIP) database between 2015 and 2018.
“In determining our patient cohort, given medically and surgically complicated patients have issues that can drive readmissions and confound results, we wanted to examine patients that would be considered a low surgical risk, to better understand differences in readmissions by urgency. As such we created an ‘ideal’ cohort characterized by basic obesity-related comorbidities, who underwent the most common laparoscopic procedures, with a smooth perioperative course,” Dr. Noria said. “We even went so far as to also exclude any ideal patient who had a drain placed, despite that being common practice prior to 2017, to avoid the possibility of including patients that had a complicated course. While this ensured an ideal cohort, it decreased our examined population from 46% to 39% of all patients in the database.”
Specifically, the researchers only included patients with the most common weight-related comorbidities, a body mass index of 35 to 70 kg/m2, ages 18 to 79 years, who underwent primary laparoscopic sleeve gastrectomy or Roux-en-Y gastric bypass without having a drain placed, who had no perioperative complications, and a hospital length of stay of two days or less.
The researchers identified 28 specific urgent reasons for readmission, six nonurgent reasons and an “other” category for unrelated etiologies. The analyses included bivariate comparisons and multivariate logistic regression to identify predictors of overall, urgent and nonurgent readmissions.
The sample included 292,547 patients. Roughly 75% of the patients underwent sleeve gastrectomy and the majority were white (62%) and female (81%). Of the total, 23% had diabetes and 43% had hypertension. The overall readmission rate in this low-risk cohort was 2.75%, with 1.27% due to nonurgent reasons and 1.06% due to urgent reasons.
The two most common reasons for readmission overall were in the nonurgent category, with 46% of all readmissions caused by nausea; vomiting; fluid, electrolyte or nutritional depletion; or abdominal pain without a specific diagnosis.
Between 2015 and 2018, overall readmissions decreased from 3.0% to 2.63%, driven by a reduction in nonurgent readmissions (1.4%-1.2%) with no change in urgent readmissions. Bivariate analysis demonstrated that surgically ideal patients readmitted after bariatric surgery were more likely to be female; identify as Black; have a higher BMI, gastroesophageal reflux disease (GERD), hypertension or insulin-dependent diabetes; and undergo a Roux-en-Y gastric bypass (Table).
Using multivariate analysis to determine predictors of readmission, Roux-en-Y gastric bypass, identification as Black or Hispanic, female sex, GERD, hypertension and insulin-dependent diabetes independently predicted an increased likelihood of readmission. Stratifying by urgency, predictors of nonurgent readmissions paralleled those of overall readmissions, whereas urgent readmissions differed according to demographic/comorbidities variables. Specifically, while Roux-en-Y gastric bypass, identification as Black or Hispanic, and GERD all remained significant, for urgent readmissions a higher baseline BMI and obstructive sleep apnea were also predictive.
| Table. Characteristics of Readmitted Bariatric Patients BMI, body mass index; GERD, gastroesophageal reflux disease | |||
| Readmitted | Not Readmitted | P Value | |
|---|---|---|---|
| (n=8,046) | (n=284,501) | ||
| Procedure, n (%) | |||
| Sleeve gastrectomy | 5,071 (63) | 220,366 (77) | <0.001 |
| Roux-en-Y gastric bypass | 2,975 (37) | 64,135 (23) | <0.001 |
| Patient Characteristics | |||
| Age, years (SD) | 42.7 (12) | 43.3 (11.7) | <0.001 |
| BMI, kg/m2 (SD) | 45.2 (6.8) | 44.9 (6.6) | <0.001 |
| Female sex, n (%) | 6,763 (84) | 230,122 (81) | <0.001 |
| Obstructive sleep apnea, n (%) | 2,830 (35) | 97,458 (34) | 0.087 |
| GERD, n (%) | 2,680 (33) | 78,107 (27) | <0.001 |
| Hypertension, n (%) | 3,658 (45) | 123,270 (43) | <0.001 |
| Insulin-dependent diabetes, n (%) | 719 (9) | 16,903 (6) | <0.001 |
| Non–insulin-dependent diabetes, n (%) | 1,315 (16) | 47,271 (17) | 0.518 |
| Race/Ethnicity, n (%) | |||
| White | 4,578 (57) | 176,024 (62) | <0.001 |
| Black | 1,778 (22) | 49,122 (17) | <0.001 |
| Hispanic | 1,177 (15) | 40,127 (14) | 0.183 |
| Native American | 49 (1) | 1,643 (1) | 0.713 |
| Asian | 29 (>1)) | 1,377 (>1)) | 0.114 |
| Unknown | 435 (5) | 16,208 (6) | 0.267 |
| BMI, body mass index; GERD, gastroesophageal reflux disease | |||
“Overall, the study suggests that while there is a decline in readmissions, driven by nonurgent causes, there continues to be a disparity in terms of race, despite the fact that all patients examined were the ideal bariatric candidate,” Dr. Noria said. “Whether you look at overall readmissions or stratify them by urgency, patients who identify as Black or Hispanic were more likely to return to hospital within 30 days of surgery—patients who presumably were at low risk for complications. The question now is, why? Ultimately, this study underscores the need to move beyond large data sets, which by their very construction only collect de-identified clinical data, and examine these disparate outcomes in the context of social determinants of health.”
“I thought that this study was really interesting,” said Monique Hassan, MD, a bariatric surgeon at Baylor Scott and White Medical Center in Temple, Texas, who served as the discussant of the study. She said some of the results of the paper are limited by its retrospective design.
This article is from the November 2021 print issue.
Please log in to post a comment