Effectiveness of Good Chemistry Groups with dually diagnosed consumers  (1998)


Project Categories

Principal Investigator:
Diana DiNitto, Ph.D.
Deborah K. Webb, Austin-Travis County Mental Health & Mental Retardation

Duration: 6/94 – 2/98

This study tested the effectiveness of Good Chemistry Groups with clients who have severe mental and substance dependence disorders. Good Chemistry is based on principles of operant conditioning and group therapy. It combines psychoeducation with group process in order to improve client functioning. It is a low cost treatment that can be added to existing mental health or chemical dependency treatment.

Subjects (n=97) were randomly assigned to a control group which received traditional inpatient chemical dependency treatment, or to an experimental group which received this treatment plus Good Chemistry. Approximately equal numbers of men and women participated. The sample was 59% White, 28% Black, and 13% Hispanic. Most (74%) had a major mood disorder, 10% had a thought disorder, 12% had both a mood and thought disorder, and 3% had post traumatic stress disorder. The majority had prior criminal convictions, and about one-half were on probation or parole. At baseline, the groups did not differ on demographic characteristics; diagnoses; prior psychiatric hospitalizations; number of criminal convictions; or medical, vocational, alcohol, drug, legal, family/social, or psychological/psychiatric status (as assessed by the Addiction Severity Index ([ASI]). Follow-ups were conducted at 30, 60, and 90 days with clients and collateral contacts. Sixty-eight clients (70% of the sample) participated in at least one follow-up with approximately equal numbers of experimental and control subjects participating. Collaterals for 78 clients (80% of the sample) participated. There was at least one client or collateral contact for 83 clients (86% of the sample).

Eight major hypotheses were tested. One hypothesis, (1) that experimental subjects would attend self-help groups more than control subjects, was supported to some extent. Seven hypotheses were not supported: (2) that experimental subjects would use alcohol and other drugs less; (3) that they would use inpatient psychiatric and substance abuse treatment less; (4) that they would have better psychotropic medication compliance; (5) that they would have less legal involvement; (6) that they would have better psychological functioning; (7) that they would have better functioning as rated by collaterals; (8) that they would have better medical, alcohol, drug, legal, family/social, and psychological functioning assessed using the ASI. There were few major differences in the functioning of the control and experimental groups over the 90 day follow-up period, but the total sample had significantly improved functioning as measured by five domains of the ASI.

This study differed from most others in the field which used control or comparison groups in that it tested an add-on treatment rather than a primary intervention, and it did so in the context of chemical dependency rather than mental illness treatment. In addition, clients were treated in an inpatient setting where their exposure to the treatment was assured rather than in an outpatient setting where they could more easily avoid the treatment. This study also contained more persons with major mood disorders than most other studies to date, and it contained a more equal proportion of female and male subjects. The study was like most other controlled studies in finding that clients generally improve following treatment, but that experimental subjects did not do significantly better than control subjects. It was also similar in demonstrating the difficulties of engaging clients in dual diagnoses treatments. Of particular interest in this study was the large number of clients who participated in Alcoholics Anonymous (AA) following discharge from inpatient treatment, despite the difficulties described when people with severe mental illness attempt to participate in AA.

Like similar studies, this one is hampered by small sample size, difficulties locating clients at follow-up, a short follow-up period, and measurement issues. It may, however, contribute to the field of integrated treatment for clients with mental and substance disorders.

Hogg Foundation for Mental Health
Lois and Samuel Silberman Fund