viding additional professional treatment services to polyproblem substance abusers, even when these services are not "matched" to specific problems at the level of the individual patient.
McLellan and colleagues recently attempted a different type of "problems to services" matching research in two inpatient and two outpatient private treatment programs (McLellan et al., 1997b). Patients in the study (N = 130) were assessed with the ASI at intake and placed in a program that was acceptable to both the Employee Assistance Program referral source and the patient. At intake, patients were also randomized to either the standard or "matched" services conditions. In the standard condition, the treatment program received information from the intake ASI, and personnel were instructed to treat the patient in the "standard manner, as though there were no evaluation study ongoing." The programs were instructed to not withhold any services from patients in the standard condition. Patients who were randomly assigned to the matched services condition were also placed in one of the four treatment programs and ASI information was forwarded to that program. However, the programs agreed to provide at least three individual sessions in the areas of employment, family/social relations, or psychiatric health delivered by a professionally trained staff person to improve functioning in those areas when a patient evidenced a significant degree of impairment in one or more of these areas at intake. For example, a patient whose intake ASI revealed significant impairments in the areas of social and psychiatric functioning would receive at least six individual sessions, three by a psychiatrist and three by a social worker.
The standard and matched patients were compared on a number of measures, including number of services received while in treatment, treatment completion rates, intake to six-month improvements in the seven problem areas assessed by the ASI, and other key outcomes at six months. Matched patients received significantly more psychiatric and employment services than standard patients, but not more family/social services or alcohol and drug services. Second, matched patients were more likely to complete treatment (93% vs. 81%), and showed more improvement in the areas of employment and psychiatric functioning than the standard patients. Third, while matched and standard patients had sizable and equivalent improvements on most measures of alcohol and drug use, matched patients were less likely to be retreated for substance abuse problems during the six-month follow-up. These findings suggest that matching treatment services to adjunctive problems can improve outcomes in key areas and may also be cost-effective by reducing the need for subsequent treatment due to relapse.
Limitations of the Matching Services to Problems Approach—It is difficult to argue against the face validity of a treatment approach for poly-