lished reporting outcome data on various treatment methods. Some of the areas covered include pharmacotherapies, aversion therapies, psychotherapy and counseling, mutual-help groups such as Alcoholics Anonymous, behavioral self-control training, and relapse prevention procedures. Typically, alcohol treatment programs offer a combination of modalities, ranging from detoxification and health care to occupational therapy and after-care group meetings. Even though treatment can be effective, however, many alcohol-dependent individuals either do not seek help or resist treatment, and most patients experience at least one relapse to drinking following treatment.

To better understand the benefits of treatment and to improve the percentage of patients who experience those benefits, researchers are working to define the active ingredients of various treatment strategies and to determine which patient factors influence treatment outcome (NIAAA, 1991). Related efforts include refining diagnostic classifications; developing improved tools for screening, diagnosis, and assessment; and improving treatment outcome evaluation.

Researchers also are working to provide information that will help to ensure that treatment services reach the populations in need. It has been well established that general medical expenditures by alcoholics and their families are reduced substantially following treatment—and the benefits derived from treating alcoholics offset costs to the general health care system (IOM, 1990b; IOM, 1989a). In light of these findings, increasing efforts will focus on expanding information about the capacities, quality, availability, utilization, and costs of alcoholism treatment services in relation to the need and the demand for those services.

Alcohol abuse and dependence is present in about 3 percent of the adult population at any given time (point prevalence), and occurs in about 14 percent of the population over the life course (lifetime prevalence) (Heizer, Burnam, and McEvoy, 1991). It begins early in adolescence (Figure 5.2). The twentieth percentile for age of diagnosis occurred during the earliest age of respondents in the ECA sample (18 years old), and 50 percent of the cases had their onset before age 25. For those cases that began in adulthood, the prodromal period was short: the difference between the twentieth percentile for diagnosis versus the twentieth percentile for first problem was less than two years.

Schizophrenia in DSM-III-R is an illness defined by inclusion and exclusion criteria with regard to psychotic symptoms, deterioration in functioning, and duration (see Table 5.6). The criteria for diagnosis have