ing what is currently known about the natural history of drug use and the role of treatment in the lives of drug users: what proportion of users enter treatment, when, how, and why they do so. We next briefly explore current treatment data. We then turn to the potential for reducing demand through expansion of treatment, with particular attention to integration of drug treatment with the criminal justice system. Finally, we consider how the impact of policy changes designed to expand, improve, or better integrate drug abuse treatment services can be modeled and researched.1


Entry into substance abuse treatment is clearly not the only or even the most prevalent pathway to stopping harmful levels of psychoactive substance use. Natural history cohort studies of alcoholics, for example, have revealed that most people modulate or stop heavy use on their own, without formal treatment. One large survey of randomly selected adults in the general population (Sobell et al., 1996) found that 78 percent of individuals who had recovered from an alcohol problem for 1 year or more did so without help or treatment. Dawson and colleagues (2005) reported findings from another large population-based study, the National Epidemiologic Survey of Alcohol and Related Conditions. Overall, independent of current recovery status, only 26 percent of adults with prior alcohol dependence reported having received treatment. However, this percentage was considerably higher (49 percent) among those who were currently abstaining and much lower (12-19 percent) among those who were still drinking at some level. One factor clearly associated with independent change versus treatment entry is severity of alcohol use and associated problems: those who elect treatment have a more serious substance use history than those who change on their own (Carballo et al., 2008).

Long-term studies that contribute to understanding of the natural history of opiate and stimulant use have also been conducted with drug users. In contrast to the research with alcoholics, however, for which large samples of the general population have formed the basis for research, samples of drug users have generally been drawn from people in treatment programs. This difference may reflect, in part, the difficulties of conducting population-based research with the relatively small (in relation to the general population) and “hidden” population represented by drug users.

George Vaillant (1973) pioneered the research with opiate users by


We are indebted to Sindelar and Kilmer (2007) for their presentation at the committee’s workshop on many of the issues in this chapter.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement