How population-level drug supply data are integrated into health outcomes research: a scoping review

Despite growing recognition that changes in the unregulated drug supply are central drivers of drug-related morbidity and mortality, there is little consensus on how population-level drug supply data are, or should be, incorporated into health outcomes research. Here, we describe the role unregulated drug supply data play in health outcomes research in the substance use field. We searched the MEDLINE, Embase, Scopus, and Europe PubMed Central databases and relevant published conference abstracts on May 28, 2025. Studies must have incorporated population-level or ecological measures of the unregulated drug supply (i.e., nonpharmaceutical) as an independent variable and reported either individual-level or population-level health outcomes, including but not limited to drug poisonings (i.e., overdose), deaths (drug-related or all-cause), hospitalizations, and emergency department visits. Of the 31 included studies, 26 were conducted in the United States and 5 in Canada. Drug supply data were used in multiple analytic roles, including as primary variables and as covariates (or predictors) of health outcomes. These data were derived from 4 distinct sources: law enforcement seizures (n = 25), drug-checking services (n = 2), hospital or clinic urinalysis results (n = 3), or postmortem toxicology results (n = 3). Most studies assessed the impact of drug supply metrics on drug toxicity (overdose) counts or rates within a given region (n = 27). We found that population-level drug supply data demonstrate predictive capacity for health outcomes and may represent an underutilized confounding factor in policy evaluation. However, their incorporation into health research remains inconsistent and conceptually fragmented.