Substance Use
Quantifying the five American drug crises Meekang Sung* Meekang Sung Lucia Pacca Anusha M. Vable Monica J. Alexander Keith Humphreys Mathew V. Kiang
Background: The United States suffers from a decades-long drug addiction and overdose crisis, resulting in over a million deaths since 1999. In the common narrative, the drug epidemic is often described as four “waves” of drug mortality. Briefly, the first wave began with prescription opioids, shifting to the second wave of heroin around 2010, followed by the ongoing third wave of synthetic opioid (i.e., fentanyl) in 2013, and the recent fourth wave of methamphetamine-related deaths. We evaluate if this national mortality pattern holds true across states.
Methods: We used 1999-2022 data from the CDC to calculate annual state-level age-standardized mortality rates for deaths involving natural/semi-synthetic (i.e., prescription) opioids, heroin, synthetic opioids, and methamphetamine. We grouped states with similar drug mortality trajectories using multichannel sequence analysis with hierarchical clustering. We fit joinpoint regression models to quantify drug mortality trajectories within each cluster.
Results: We identified five distinct clusters. Clusters A and B comprise states mostly east of the Mississippi and generally follow the common four-wave narrative, with higher methamphetamine mortality in Cluster A. The other 3 clusters, comprising 27 states, diverge from the four-wave narrative. For example, Cluster E never experienced high prescription or heroin mortality but is experiencing high and rapidly increasing synthetic opioid and methamphetamine mortality.
Discussion: The “drug epidemic” is not a monolith but varies widely across states. These findings have significant implications for public policy. National drug control strategies may only address some states and potentially harm others. For example, policies to reduce opioid prescribing may prevent prescription opioid mortality in some states while reducing effective pain treatment in others. Likewise, variation across states suggests national metrics may be too crude to adequately evaluate impact.
