and the contingency based counseling (Higgins et al., 1995) as compared against groups who received comparable amounts of all components except the target ingredient. In each case, these systematic and controlled examinations indicated that these individual components made a significant contribution to the outcomes observed, thus proving their added value in the rehabilitation effort. Extending this work on the use of positive reinforcement and behavioral contracting, Silverman and colleagues (Silverman et al., 1996) used essentially the same reinforcement contingencies and contracting procedures that had been applied by Azrin and Higgins to improve the performance of methadone maintained patients.

"Matching" Patients and Treatments—The past two decades have witnessed a great number of research studies attempting to "match" patients with specific types, modalities or settings of treatment. The approach to patient-treatment "matching" that has received the greatest attention from substance abuse treatment researchers involves attempting to identify the characteristics of individual patients that predict the best response to different forms of addiction treatments (e.g., cognitive-behavioral vs. 12-Step, or inpatient vs. outpatient) (Mattson et al., 1994; Project MATCH Research Group, 1997). In general, the majority of these "patient-to-treatment" matching studies have not shown robust or generalizable findings (see Gastfriend and McLellan, 1997). Another approach to matching has been to assess patients' problem severity in a range of areas at intake and then ''match" the specific and necessary services to the particular problems presented at the assessment. This has been called "problem-to-service" matching (McLellan et al., 1997b). This approach may have more practical application as it is consonant with the "individually tailored treatment" philosophy that has been espoused by most practitioners.

Substance abusers with comorbid psychiatric problems may be particularly good candidates for the "problem-to-service" matching approach; especially the addition of specialized psychiatric services for those most severely affected by psychiatric problems. For example, recent studies suggest that tricyclic antidepressants and the selective serotonergic medication fluoxetine may reduce both drinking and depression levels in alcoholics with major depression (Cornelius et al., 1997; Mason et al., 1996; McGrath et al., 1996). Similarly, the anxiolytic buspirone may reduce drinking in alcoholics with a comorbid anxiety disorder (Kranzler et al., 1994). Highly structured relapse prevention interventions may also be more effective in decreasing cocaine use, as compared to less structured interventions, in cocaine abusers with comorbid depression (Carroll et al., 1995).

Woody and colleagues have evaluated the value of individual psychotherapy when added to paraprofessional counseling services in the course of methadone maintenance treatment (Woody et al., 1983). In that study