A Randomized Controlled Trial of the Illness Management and Recovery Program for Persons With Schizophrenia

Participants were recruited from rehabilitation centers serving individuals with psychotic disorders. Inclusion criteria were a DSM-IV diagnosis of schizophrenia or schizoaffective disorder, proficiency in Swedish, and willingness to provide informed consent after receipt of detailed study information. A computerized random-number generator was used to assign the 41 participants who completed the baseline assessment and gave informed consent to one of the six IMR program groups (N=21) or to treatment as usual (N=20).

All participants received extensive outpatient psychiatric services at the rehabilitation centers, including case management, psychiatric treatments (antipsychotic medication and psychotherapy), and access to recreational and therapeutic activities, such as social, leisure, work, and support activities. Participants in the IMR group received IMR in addition to these services.

Illness management, psychiatric symptoms, quality of life, coping, recovery, and insight were assessed at baseline, posttreatment (nine months), and follow-up (21 months). Hospitalization and suicidality were assessed at baseline and at follow-up. Participants were not reimbursed for completing assessments. Participants' case managers, who were not blind to treatment assignment, provided ratings on the clinician version of the IMR scale, and assessments of psychiatric symptoms were conducted by trained clinicians who were blind to treatment assignment (see below).

At posttreatment (nine months after baseline) all participants completed the assessments. One participant in the IMR group left the program but completed assessments at posttreatment and at follow-up and was included in the final analysis. Thirty-eight participants completed the follow-up assessments at 21 months (19 from the IMR program and 19 from the control group). Two participants (one from each group) did not respond to a request to schedule the follow-up assessment, and one participant in the IMR group had died.

Illness management was measured with the 15-item Illness Management and Recovery Scale (IMRS). Parallel client and clinician versions have been developed to capture the target domains of the program, including personal goals, knowledge of mental illness, adherence to medication, social support, relapse prevention, coping with symptoms and functioning, and substance abuse and dependence (11). Previous studies of the IMRS have shown good internal consistency, test-retest reliability, and convergent validity (12,13). For the study reported here, the scales were translated into Swedish and then independently back-translated into English and compared with the original version to identify and correct any discrepancies.

Psychiatric symptoms were assessed with a modified version of the Psychosis Evaluation Tool for Common Use by Caregivers (PECC). The PECC is a 26-item scale that is used by a trained interviewer to rate severity of psychiatric symptoms, including psychosis, negative symptoms, mania, depression, cognitive symptoms, insight, and suicidality. Interrater reliability and interscale validity for the PECC have been established (14). Five interviewers completed two days of PECC training before the assessments, including ratings of videotaped interviews. They were considered to be fully trained when satisfactory agreement on each item was established (interrater reliability r >.80). Statistical analyses were conducted on the total PECC score and on subscale scores for psychosis, negative symptoms, mania, depression, and cognitive symptoms. The PECC insight subscale assesses insight on two dimensions: ability to identify experienced phenomena as symptoms of mental illness (for example, hallucinations, delusions, and blunted affect) and ability to see a causal relationship between experienced symptoms and the illness. A modified version of the PECC suicidality subscale was used to assess suicidal ideation. A dichotomized variable (presence or absence) evaluated suicidal ideation during the week before the assessment.

Psychosocial functioning was assessed with the Manchester Short Assessment of Quality of Life (MANSA) (15) and the Ways of Coping Questionnaire (WCQ) (16). The MANSA is a 12-item scale that elicits information about satisfaction with life as a whole and with job, financial situation, number and quality of friends, leisure, living situation, personal safety, persons in the client's household, sex life, relationship with family, physical health, and mental health. Satisfaction on each item is rated on a 7-point scale. Research has shown good validity, reasonable reliability, and associations with psychopathology (15). Scores are reported as the mean across items. The WCQ is a 66-item scale that assesses eight aspects of the respondent's coping style—confrontive coping, distancing, self-control, seeking social support, accepting responsibility, escape-avoidance, “planful” problem solving, and positive reappraisal. The eight factors have been shown to have good internal reliability and convergent validity for persons with schizophrenia (17). Scores are reported as the mean for each of the eight coping factors.

The Recovery Assessment Scale (RAS) is a 41-item instrument assessing perception of recovery from mental illness (18). The scale includes five factors: personal confidence and hope, willingness to ask for help, goal and success orientation, positive reliance on others, and not being dominated by symptoms. Research indicates high internal reliability, test-retest reliability, and convergent validity (19). Scores are reported as the mean across all 41 items.

Medical records were reviewed to record hospitalizations during the year before the baseline assessment and between the posttreatment and follow-up assessments. Hospitalization was defined as admission to a psychiatric unit for at least 24 hours (dichotomized as hospitalization or no hospitalization).

Treatment with antipsychotics was monitored through the use of Defined Daily Dose (DDD), a technical unit of measurement developed by the World Health Organization for presentation and comparison of drug consumption in drug utilization research; DDD is used to investigate whether a change in outcome measures is the result of changes in antipsychotic dosage (20). Antipsychotic dosage for each participant was converted to DDD at baseline, posttreatment, and follow-up. Results are descriptive and reported as means across the groups.

Participants in the IMR program group were engaged in weekly sessions for nine months, during which the 40 sessions were delivered. At each psychiatric outpatient rehabilitation center, sessions were cofacilitated by two clinicians. The clinicians were recruited on the basis of expressed interest in the program, and they received five full days of training before the study. During the program, the clinicians attended weekly supervision to resolve challenges involving the participants and the group format and to ensure high program fidelity. Groups were small, with three or four participants in each. Sessions lasted approximately one hour and were conducted at the rehabilitation centers. During each session, the IMR curriculum was reviewed by use of a PowerPoint presentation that covered the following topics: recovery strategies, practical facts about schizophrenia, the stress-vulnerability model and treatment strategies, building social support, effective use of medication, relapse reduction, coping with stress, coping with problems and persistent symptoms, and getting needs met in the mental health system. At the beginning of the program, participants identified personal recovery goals that were broken down into smaller goals and worked on over the course of the program. As described in the IMR manual, the curriculum was taught with a combination of psychoeducational, motivational, and cognitive-behavioral teaching methods. Between-session home assignments were chosen by participants and reviewed at the next session. Significant others were encouraged to support participants during the program.

Baseline between-group differences in background characteristics were evaluated with chi square tests and t tests. The IMR attendance rate was calculated as the group mean of attended sessions. Relative change in outcome variables measuring illness management, psychiatric symptoms, quality of life, coping, recovery, and insight were evaluated at posttreatment and follow-up by using a linear repeated-measures mixed model with baseline data inserted as random covariates. Post hoc analyses of relative change between groups were conducted with independent t tests, with the Bonferroni correction to control for type I errors. Effect size estimates were calculated by using Cohen's d for group mean scores from baseline to posttreatment and from baseline to follow-up. Hierarchical multiple regression analyses were conducted to investigate whether relative change in coping skills predicted or mediated relative change in psychiatric symptoms at posttreatment and at follow-up. Linear regression was used to investigate whether background characteristics predicted IMR attendance and whether IMR attendance predicted change in PECC total scores. Dichotomous data on hospitalization and suicidality from baseline to follow-up were analyzed with McNemar tests. SPSS release 18.0.1 was used for all statistical analyses.