Mapping the spread of methamphetamine abuse in California from 1995 to 2008.

OBJECTIVES From 1983 to 2008, the incidence of methamphetamine abuse and dependence (MA) presenting at hospitals in California increased 13-fold. We assessed whether this growth could be characterized as a drug epidemic. METHODS We geocoded MA discharges to residential zip codes from 1995 through 2008. We related discharges to population and environmental characteristics using Bayesian Poisson conditional autoregressive models, correcting for small area effects and spatial misalignment and enabling an assessment of contagion between areas. RESULTS MA incidence increased exponentially in 3 phases interrupted by implementation of laws limiting access to methamphetamine precursors. MA growth from 1999 through 2008 was 17% per year. MA was greatest in areas with larger White or Hispanic low-income populations, small household sizes, and good connections to highway systems. Spatial misalignment was a source of bias in estimated effects. Spatial autocorrelation was substantial, accounting for approximately 80% of error variance in the model. CONCLUSIONS From 1995 through 2008, MA exhibited signs of growth and spatial spread characteristic of drug epidemics, spreading most rapidly through low-income White and Hispanic populations living outside dense urban areas.

[1]  Lon-Mu Liu,et al.  Impacts of federal ephedrine and pseudoephedrine regulations on methamphetamine-related hospital admissions. , 2003, Addiction.

[2]  L. Waller,et al.  Applied Spatial Statistics for Public Health Data: Waller/Applied Spatial Statistics , 2004 .

[3]  Joel Schwartz,et al.  Measurement error caused by spatial misalignment in environmental epidemiology. , 2009, Biostatistics.

[4]  Lon-Mu Liu,et al.  Impacts of federal precursor chemical regulations on methamphetamine arrests. , 2005, Addiction.

[5]  J. Caulkins The need for dynamic drug policy. , 2007, Addiction.

[6]  Andrew Thomas,et al.  WinBUGS - A Bayesian modelling framework: Concepts, structure, and extensibility , 2000, Stat. Comput..

[7]  D. Clayton,et al.  Bayesian analysis of space-time variation in disease risk. , 1995, Statistics in medicine.

[8]  R. Ostfeld,et al.  Spatio-temporal patterns in county-level incidence and reporting of Lyme disease in the northeastern United States, 1990–2000 , 2007, Environmental and Ecological Statistics.

[9]  David Bright,et al.  A systematic review of methamphetamine precursor regulations. , 2011, Addiction.

[10]  J. Besag,et al.  Bayesian image restoration, with two applications in spatial statistics , 1991 .

[11]  J. Caulkins,et al.  Markov Chain Modeling of Initiation and Demand: The Case of the US Cocaine Epidemic , 2004, Health care management science.

[12]  C. Castillo-Chavez,et al.  Raves, clubs and ecstasy: the impact of peer pressure. , 2005, Mathematical biosciences and engineering : MBE.

[13]  J. Andrew Royle,et al.  Hierarchical Spatiotemporal Matrix Models for Characterizing Invasions , 2007, Biometrics.

[14]  T Déirdre Hollingsworth,et al.  Controlling infectious disease outbreaks: Lessons from mathematical modelling , 2009, Journal of public health policy.