JAMA Netw Open. 2026 Jun 1;9(6):e2618686. doi: 10.1001/jamanetworkopen.2026.18686.
ABSTRACT
IMPORTANCE: Metabolic bariatric surgery (MBS) is a highly effective but underused treatment for obesity. Greater understanding of diagnosis and referral pathways are needed to help inform appropriate treatment options for obesity.
OBJECTIVE: To examine factors associated with uptake of MBS among individuals with incident obesity diagnosis.
DESIGN, SETTING, AND PARTICIPANTS: This cohort study used insurance claims data from the Merative MarketScan Commercial Claims and Encounters Database to identify a cohort of individuals aged 18 to 64 years with an incident obesity diagnosis (index) in inpatient and outpatient settings between January 1, 2018, and December 31, 2022, as well as MBS uptake within 12 months of index to December 31, 2023.
EXPOSURES: Diagnoses classified using Agency for Healthcare Research and Quality Clinical Classifications Software Refined and therapeutic classes (Red Book) of drugs taken within the 12-month period prior to index date, as well as sociodemographic and health care factors determined at the index date.
MAIN OUTCOMES AND MEASURES: Factors associated with MBS uptake were examined using machine learning methods and multivariable logistic regression. Exclusion criteria included pregnancy-related visits; missing International Statistical Classification of Diseases and Related Health Problems, Tenth Revision, body mass index (BMI) Z-code; or prior MBS. The top 25 factors identified by a random forest model were ranked by Shapley Additive Explanation (SHAP) values, and odds ratios (ORs) were estimated for each top factor in a multivariable logistic regression model.
RESULTS: In this study of 109 849 individuals (mean [SD] age, 41.4 [11.8] years; 63 992 women [58.3%]) with an initial obesity diagnosis, 3268 (3.0%) received MBS within 12 months of index. The area under the receiver operating characteristic curve score was 0.88 (95% CI, 0.87-0.90). The top factors identified by the random forest model included surgeon (not elsewhere classified [NEC]) as provider (which includes medical institutions and organizations as well as health care professionals) (mean SHAP value, 0.00918 [95% CI, 0.00893-0.00945]), BMI category (mean SHAP value, 0.00501 [95% CI, 0.00499-0.00503]), family practice as provider (mean SHAP value, 0.00372 [95% CI, 0.00370-0.00374]), and female sex (mean SHAP value, 0.00370 [95% CI, 0.00368-0.00371]). Surgeon (NEC) as provider had the largest OR (18.97 [95% CI, 17.11-21.02]), followed by other provider types (orthopedic surgeon: OR, 5.33 [95% CI, 4.54-6.24]; dietitian: OR, 2.51 [95% CI, 1.97-3.20]). Several medication (antidiabetic agents, miscellaneous: OR, 1.42 [95% CI, 1.25-1.62]; psychotherapeutics, antidepressants: OR, 1.13 [95% CI, 1.01-1.25]), diagnostic (depressive disorders: OR, 1.16 [95% CI, 1.02-1.33]), and sociodemographic variables (female sex: OR, 2.45 [95% CI, 2.22-2.71]) were also significantly associated with MBS. There was a significant decrease in MBS uptake for individuals with incident obesity diagnosis in more recent year-quarters from 2019 through 2022.
CONCLUSIONS AND RELEVANCE: In this cohort study of individuals with an initial obesity diagnosis, uptake of MBS was patterned by demographic, clinical, and health care factors. This finding suggests substantial opportunity to improve access to MBS through targeted policies and interventions.
PMID:42301711 | DOI:10.1001/jamanetworkopen.2026.18686
