Using the Super Learner algorithm to predict risk of 30-day readmission after bariatric surgery in the United States.

BACKGROUND: Risk prediction models that estimate patient probabilities of adverse events are commonly deployed in bariatric surgery. The objective was to validate a machine learning (Super Learner) prediction model of 30-day readmission after bariatric surgery in comparison with a traditional logistic regression. METHODS: This prognostic study for validation of risk prediction models used data from the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program. Patients who underwent elective laparoscopic gastric bypass or laparoscopic sleeve gastrectomy between 2015 and 2018 were included. Models used 5-fold cross-validation and were evaluated using the area under the receiver operating characteristic curve, the net reclassification index, and the integrated discrimination improvement. RESULTS: The 30-day readmission rate among 393,833 patients was 3.9%. Super Learner area under the receiver operating characteristic curve was 0.674 (95% confidence interval 0.670-0.679), compared to 0.650 (95% confidence interval 0.645-0.654) for logistic regression. The net reclassification index was 0.239 (95% confidence interval 0.223-0.254), and 0.252 (95% confidence interval 0.249-0.255) for those who were and were not readmitted within 30 days. The integrated discrimination improvement was 0.0032 (95% confidence interval 0.0030-0.0033). CONCLUSION: The Super Learner outperformed traditional logistic regression in predicting risk of 30-day readmission after bariatric surgery. Machine learning models may help target high-risk patients more optimally and prevent unnecessary readmissions.

Changes in Health Insurance Coverage Over Time by Immigration Status Among US Older Adults, 1992-2016.

Importance: Disparities in health insurance coverage by immigration status are well documented; however, there are few data comparing long-term changes in insurance coverage between immigrant and nonimmigrant adults as they age into older adulthood. Objective: To compare longitudinal changes in insurance coverage over 24 years of follow-up between recent immigrant, early immigrant, and nonimmigrant adults in the US. Design, Setting, and Participants: This population-based cohort study used data from the nationally representative Health and Retirement Study. Data were collected biennially from 1992 to 2016. The population included community-dwelling US adults born between 1931 and 1941 and aged 51 to 61 years at baseline. Statistical analysis was performed from February 3, 2017, to January 10, 2020. Exposures: Participants were categorized as nonimmigrants (born in the US), early immigrants (immigrated to the US before the age of 18 years), and recent immigrants (immigrated to the US from the age of 18 years onward). Main Outcomes and Measures: Self-reported data on public, employer, long-term care, and other private insurance were used to define any insurance coverage. Longitudinal changes in insurance coverage were examined over time by immigration status using generalized estimating equations accounting for inverse probability of attrition weights. The association between immigration status and continuous insurance coverage was also evaluated. Results: A total of 9691 participants were included (mean [SD] age, 56.0 [3.2] years; 5111 [52.6%] female). Nonimmigrants composed 90% (n = 8649) of the cohort; early immigrants, 2% (n = 201); and recent immigrants, 8% (n = 841). Insurance coverage increased from 68%, 83%, and 86% of recent immigrant, early immigrant, and nonimmigrant older adults, respectively, in 1992 to 97%, 100%, and 99% in 2016. After accounting for selective attrition, recent immigrants were 15% less likely than nonimmigrants to have any insurance at baseline (risk ratio, 0.85; 95% CI, 0.82-0.88), driven by lower rates of private insurance. However, disparities in insurance decreased incrementally over time and were eliminated, such that insurance coverage rates were similar between groups as participants attained Medicare age eligibility. Furthermore, recent immigrants were less likely than nonimmigrants to be continuously insured (risk ratio, 0.89; 95% CI, 0.85-0.94). Conclusions and Relevance: Among community-dwelling adults who were not age eligible for Medicare, recent immigrants had lower rates of health insurance, but this disparity was eliminated over the 24-year follow-up period because of uptake of public insurance among all participants. Future studies should evaluate policies and health care reforms aimed at reducing disparities among vulnerable populations such as recent immigrants who are not age eligible for Medicare.