The heroin epidemic in San Francisco: estimates of incidence and prevalence.

Much attention has recently been focused on the question of the extent of heroin use in America. In a worthwhile effort to call public attention to the problem, many estimates of its size have evidenced a tendency toward exaggeration and aggrandizement. This paper presents methods of estimation of the extent of heroin addiction which, when carefully employed, should effectively correct such distorted estimates. Two general types of estimation are employed, incidence and prevalence. Incidence estimates are concerned with new cases of heroin addiction that occur in a specific population within a given amount of time. These estimates are based upon self-report data from addicts regarding date of first addiction. When corrected for the "lag phase" (that period of time between onset of first addiction and eventual visit to a treatment program), these data permit accurate retrospective charting of incidence trends. Prevalence estimates focus on all known cases of heroin addiction in a specific population within a given amount of time. Three separate types of prevalence estimates from three separate sources are outlined in this article: estimates based on overdose death data, estimates based on crime statistics, and estimates of "unknown" addicts. In outlining these methods, this article describes the fluctuations in heroin addiction in one major American city, San Francisco, California. After analyzing data gathered from a sample of 2,367 addicts contacted over a 3-year period, this study suggests that the incidence of heroin addiction seems to have declined after 1970. Possible factors underlying this apparent decline in heroin addiction are then discussed, including the post-1970 maturation of the "population at risk," the effectiveness of antidrug media messages, the changing drug fashions in the heroin subculture, and the gradual deterioration of the quality and potency of street heroin.