Literature Collection

BACKGROUND: The prevalence of depression among older adults places a considerable strain on healthcare systems due to a shortage of psychiatrists for professional evaluations. This study reports a machine learning (ML) model to assist in screening for geriatric depression, enabling primary care practitioners to detect and respond to cases effectively and on time. METHODS: Data from the 2011-2018 National Health and Nutrition Examination Survey (NHANES) were used. To identify relevant variables for depression screening, features were selected using Least Absolute Shrinkage and Selection Operator (LASSO) regression. Synthetic Minority Over-sampling Technique (SMOTE) was applied to address class imbalance in the training set. Seven ML algorithms were employed to develop predictive screening models. Model performance was evaluated using standard ML metrics and clinically relevant impact measures, ensuring a comprehensive assessment. To improve interpretability, Shapley Additive exPlanations (SHAP) were used to visualize the contribution of each feature. RESULTS: The analysis included 3802 participants, with 933 (24.54 %) identified as having depression. Thirteen key variables were selected for model development. The Extreme Gradient Boosting (XGBoost) model demonstrated the best performance, with an accuracy of 0.82 in test set and a maximum Area Under the Curve (AUC) of 0.88 in Receiver Operating Characteristic (ROC) analysis. Sleep disorders, gender, poverty-income ratio (PIR), serum albumin levels, and segmented neutrophil count were identified as the influential predictors. CONCLUSION: The ML model developed for screening depression in older adults showed strong predictive performance and clinical applicability, supporting health workers in the early identification and management of depression.

IMPORTANCE: The rate of maternal mortality in the United States is 2-fold to 3-fold greater than that in other high-income countries. While many national initiatives have been developed to combat maternal mortality, these efforts often fail to include mental illness. OBJECTIVE: To highlight the underrecognized contribution of mental illness to maternal mortality, which is nearly double that of postpartum hemorrhage. EVIDENCE REVIEW: A topic outline was developed to include challenges in measuring perinatal mental conditions and mortality rates; contributions of social determinants of health to mental conditions and mortality; perinatal psychiatric disorder characterization; mechanisms by which maternal mental illness increases mortality, specifically, suicide and addictive disorders; access limitations and care "deserts"; prenatal stress and its impact on reproductive outcomes; increasing clinician expertise through cross-disciplinary education; intervention sites and models; and asserting that mental health is fundamental to maternal health. Publications in the last 3 years were prioritized, particularly those relating to policy. References were selected through consensus. Sources were PubMed, Ovid, direct data published on government websites, and health policy sources such as the Policy Center for Maternal Mental Health. FINDINGS: Priority was given to recent sources. Citations from 2022-2023 numbered 26; within the last 5 years, 14; and historical references, 15. Recommendations to address each topic area serve as concluding statements for each section. To mitigate the contributions of mental illness to the maternal mortality risk, a coordinated effort is required across professional and governmental organizations. CONCLUSIONS AND RELEVANCE: Concrete programmatic and policy changes are needed to reduce perinatal stress and address trauma, standardize the collection of social determinant of health data among perinatal patients, increase access to reproductive psychiatry curricula among prescribers, reduce perinatal mental health and obstetrical deserts, institute paid parental leave, and support seamless integration of perinatal and behavioral health care. Moreover, instead of focusing on a relatively minor portion of the contributors to health that current medical practice targets, fortifying the social foundation strengthens the prospects for the health of families for our current and future generations.

IMPORTANCE: Agonist medications for opioid use disorder (MOUD), buprenorphine and methadone, in carceral settings might reduce the risk of postrelease opioid overdose but are uncommonly offered. In April 2019, the Massachusetts Department of Correction (MADOC), the state prison system, provided buprenorphine for incarcerated individuals in addition to previously offered injectable naltrexone. OBJECTIVE: To evaluate postrelease outcomes after buprenorphine implementation. DESIGN, SETTING, AND PARTICIPANTS: This cohort study with interrupted time-series analysis used linked data across multiple statewide data sets in the Massachusetts Public Health Data Warehouse stratified by sex due to differences in carceral systems. Eligible participants were individuals sentenced and released from a MADOC facility to the community. The study period for the male sample was January 2014 to November 2020; for the female sample, January 2015 to October 2019. Data were analyzed between February 2022 and January 2024. EXPOSURE: April 2019 implementation of buprenorphine during incarceration. MAIN OUTCOMES AND MEASURES: Receipt of MOUD within 4 weeks after release, opioid overdose, and all-cause mortality within 8 weeks after release, each measured as a percentage of monthly releases who experienced the outcome. Segmented linear regression analyzed changes in outcome rates after implementation. RESULTS: A total of 15 225 individuals were included. In the male sample there were 14 582 releases among 12 688 individuals (mean [SD] age, 35.0 [10.8] years; 133 Asian and Pacific Islander [0.9%], 4079 Black [28.0%], 4208 Hispanic [28.9%], 6117 White [41.9%]), a rate of 175.7 releases per month; the female sample included 3269 releases among 2537 individuals (mean [SD] age, 34.9 [9.8] years; 328 Black [10.0%], 225 Hispanic [6.9%], 2545 White [77.9%]), a rate of 56.4 releases per month. Among male participants at 20 months postimplementation, the monthly rate of postrelease buprenorphine receipt was higher than would have been expected under baseline trends (21.2% vs 10.6% of monthly releases; 18.6 additional releases per month). Naltrexone receipt was lower than expected (1.0% vs 6.0%; 8.8 fewer releases per month). Monthly rates of methadone receipt (1.4%) and opioid overdose (1.8%) were not significantly different than expected. All-cause mortality was lower than expected (1.9% vs 2.8%; 1.5 fewer deaths per month). Among female participants at 7 months postimplementation, buprenorphine receipt was higher than expected (31.6% vs 9.5%; 12.4 additional releases per month). Naltrexone receipt was lower than expected (3.4% vs 7.2%) but not statistically significantly different. Monthly rates of methadone receipt (1.1%), opioid overdose (4.8%), and all-cause mortality (1.6%) were not significantly different than expected. CONCLUSIONS AND RELEVANCE: In this cohort study of state prison releases, postrelease buprenorphine receipt increased and naltrexone receipt decreased after buprenorphine became available during incarceration.