Federal Grant Award Summary

The National Institute on Drug Abuse (NIDA) awarded Massachusetts General Hospital (MGH) a Project Grant totaling $476,808 under the Drug Use and Addiction Research Programs (CFDA 93.279) to develop precision risk modeling tools for reducing opioid overdose deaths among individuals experiencing homelessness. The award, dated May 15, 2026, with completion targeted for April 30, 2028, funds research that will derive and validate prediction models for opioid overdose mortality and all-cause mortality using 20 years of clinical and mortality data from a large Boston-based cohort. MGH will employ an ensemble of five supervised machine learning models—regularized logistic regression, extreme gradient boosting, support vector machines, random forest, and neural networks—to identify predictors of mortality among individuals with clinically diagnosed opioid use disorder (OUD).

The project will also characterize six-month OUD treatment trajectories and examine their association with subsequent mortality outcomes using group-based trajectory modeling to identify latent classes of treatment patterns. This research directly addresses a critical public health gap, as individuals experiencing homelessness experience opioid overdose death rates up to 30 times higher than the general population, yet existing overdose risk models have not been tailored to or validated in homeless populations. By incorporating dynamic, time-varying treatment patterns and social determinants specific to this vulnerable population, the project aims to generate evidence-based risk assessment tools that can improve treatment adherence and reduce mortality in this high-risk group.