Fentanyl purity and overdose decline: A reexamination of geographic trends.

Abstract

Drug overdose deaths in the United States reached record levels during the unregulated ("illicit") fentanyl era before recently declining. A plausible hypothesis is that a sudden drop in fentanyl purity ("supply shock") beginning in 2023 caused the downturn. We evaluated this hypothesis by replicating a published analysis with regional overdose data, using transfer function models accounting for time trends and autocorrelation; negative controls tested for spurious correlation. Replicating the original paper, we extracted fentanyl purity data from a Drug Enforcement Administration (DEA) figure; overdose mortality used standard vital statistics sources. We found rising national fentanyl purity did not track fatal fentanyl overdose rates in most United States regions, showing only modest association in the West. When both purity and mortality later declined, the observed associations were also seen with unrelated macroeconomic indicators that shared the same time pattern, raising concerns about spurious correlation in the original study. Canadian analyses also revealed variation in both strength and direction of association between fentanyl purity and overdose deaths at subprovince level. Differential selection bias in law enforcement purity assessment may have compromised longitudinal and geographic generalizability of DEA data, notably stimulated reporting for higher-purity wholesale samples in 2023-2024. National fentanyl purity alone does not provide sufficient explanation for the recent decline in overdose deaths. Reductions in localized purity may influence localized overdose decreases, but alternative causes must be considered nationally, including drug supply changes, public health interventions, demographic shifts, behavior changes, post-pandemic recovery, cartel dynamics, and fewer initiates to illicit opioid use.