Drug Policy: Insights from Mathematical Analysis

Illicit drug use is clearly an important health problem. There are some 600,000 emergency department episodes in the US every year that are related to illicit drugs (SAMHSA, 2002a). National mortality estimates are not available, but there are probably on the order of 20,000 drug-induced deaths a year (SAMHSA, 2002b), with many more indirectly related to drug use. Some 5 million Americans are in need of drug treatment, and less than 40% get it (Epstein and Gfroerer, 1998; Woodward et al., 1997). Injection drug use is a leading cause of the spread of infectious diseases such as HIV/AIDS and Hepatitis C (CDCP, 2001). The social costs of illicit drug use approach those of alcohol and tobacco (Rice et al., 1990; Bartlett et al., 1994; Harwood et al., 1998). No one has estimated how many quality adjusted life years are lost due to illicit drug use, but the number is no doubt substantial, particularly since those who die from illicit drug use are younger than those who die from most other causes. Not surprisingly there is an energetic debate concerning how best to control drug use and related consequences, to which Operations Research/Management Science has made important contributions. Nevertheless, drug policy is unlike other health policy domains in important ways, and this article begins with a review of some important differences. The following sections then highlight key insights quantitative models have generated concerning the relative effectiveness of different interventions, including how that effectiveness varies over the course of a drug epidemic.