Trends and geographic patterns in drug-poisoning death rates in the U.S., 1999-2009

Am J Prev Med. 2013 Dec;45(6):e19-25. doi: 10.1016/j.amepre.2013.07.012.


Background: Drug poisoning mortality has increased substantially in the U.S. over the past 3 decades. Previous studies have described state-level variation and urban-rural differences in drug-poisoning deaths, but variation at the county level has largely not been explored in part because crude county-level death rates are often highly unstable.

Purpose: The goal of the study was to use small-area estimation techniques to produce stable county-level estimates of age-adjusted death rates (AADR) associated with drug poisoning for the U.S., 1999-2009, in order to examine geographic and temporal variation.

Methods: Population-based observational study using data on 304,087 drug-poisoning deaths in the U.S. from the 1999-2009 National Vital Statistics Multiple Cause of Death Files (analyzed in 2012). Because of the zero-inflated and right-skewed distribution of drug-poisoning death rates, a two-stage modeling procedure was used in which the first stage modeled the probability of observing a death for a given county and year, and the second stage modeled the log-transformed drug-poisoning death rate given that a death occurred. Empirical Bayes estimates of county-level drug-poisoning death rates were mapped to explore temporal and geographic variation.

Results: Only 3% of counties had drug-poisoning AADRs greater than ten per 100,000 per year in 1999-2000, compared to 54% in 2008-2009. Drug-poisoning AADRs grew by 394% in rural areas compared to 279% for large central metropolitan counties, but the highest drug-poisoning AADRs were observed in central metropolitan areas from 1999 to 2009.

Conclusions: There was substantial geographic variation in drug-poisoning mortality across the U.S.

Publication types

  • Observational Study

MeSH terms

  • Age Distribution
  • Bayes Theorem
  • Cause of Death
  • Drug Overdose / epidemiology
  • Drug Overdose / mortality*
  • Humans
  • Models, Statistical*
  • Poisoning / epidemiology
  • Poisoning / mortality*
  • Population Surveillance
  • Probability
  • Rural Population
  • Time Factors
  • United States / epidemiology
  • Urban Population