Multifamily Group Treatment for Schizophrenia

Multifamily group treatment (MFGT) is an evidence-based intervention for treating persons with severe mental illness, particularly schizophrenia, and their families that integrates psychoeducation and behavioral family therapy in a multiple-family group format. Although multifamily group treatment was first pioneered more than half a century ago by Laqueur (Laqueur et al., 1964), only in the past 15 years has a coherent theoretical model and empirical evidence for its effectiveness been available to clinicians.

The work by William McFarlane and his colleagues in the United States has greatly contributed to the model's subsequent refinement and empirical support (McFarlane, 1983, 2002). There are four major stages in his MFGT program: joining (i.e., building rapport/alliance) among individual patients and families, conducting an educational workshop about schizophrenia for families, relapse prevention through the use of problem-solving format groups attended by both patients and families, and vocational and social skills rehabilitation in the same multifamily group. MFGT is delivered by two clinicians to groups of 5-8 families over a 2-year period. A main theoretical foundation of MFGT is that by increasing social network size and support by enabling families to benefit from each other's experiences in solving problems, better illness course and improved outcome occur. Across a number of clinical trials, MFGT based on the McFarlane model has been shown to decrease relapse and rehospitalization among patients with schizophrenia and to improve family well-being over the 2-year treatment period (see review in McFarlane et al., 2003). Consequently, the American Psychological Association (2004) recommends MFGT as a best practice for serious mental illness.

MFGT has indeed come of age, yet several understudied questions remain, two of which we focus on in the present review. First, are treatment gains sustained past the intervention period? In other words, given the chronic nature of schizophrenia, should we expect that the well-documented effects of MFGT in these patients last? Second, is MFGT effective across ethnic groups? Despite the benefits demonstrated in clinical trials, treatment studies have generally focused on European-American populations. The effectiveness of MFGT for other ethnic samples is virtually unknown. The two studies that we review below are among the most recent evaluations of outcome at 1-year post-treatment, with one study also evaluating the effects of MFGT on an Asian sample.

McDonnell and colleagues (2006) examined service utilization, both outpatient and inpatient, in 97 outpatients with schizophrenia (n = 64), schizoaffective disorder (n = 32), or other psychotic disorder (n = 1) at a large urban community mental health center during the 2-year MFGT period and at 1-year post-treatment follow-up. Patients were randomized into standard psychiatric care and MFGT (n = 53; female = 12) or standard care alone (n = 44; female = 9). Most patients received atypical antipsychotic medications and had low levels of psychiatric symptoms. Mean age of MFGT and SC patients was 31.9 years (SD = 8.7) and 33.8 years (SD = 10.2), respectively. Approximately 92% of participants (combined MFGT and SC study samples) were European-American. All MFGT sessions (24 sessions in year 1 and 12 sessions in year 2) followed McFarlane's model as defined by a treatment manual (McFarlane, 2002), and clinicians obtained scores greater than 70% on a measure of treatment adherence. Service utilization data were obtained from the management information system of the community mental health agency where all treatment was provided. Days of inpatient hospitalization at a community hospital's psychiatric ward and at a state psychiatric hospital were calculated separately. Outpatient service use was calculated based on hours of staff time summed across 1-year intervals.

In other papers from this study, compared to patients receiving standard psychiatric care (SC), participants who also received MFGT showed a reduction in negative symptoms and inpatient service use at the end of the first-year treatment period. These reductions were observed without significantly higher outpatient service use at 1 year among the MFGT participants compared to the SC group (Dyck et al., 2000, 2002). Moreover, relative to caregivers of the SC group, family caregivers who received MFGT showed reduced psychological distress across the 2-year treatment period (Hazel et al., 2004).

The results from the present paper were consistent with these previous results regarding reduction of inpatient use in the MFGT study sample relative to the SC sample; however, the present study examined these results more closely. During the first year of treatment, group membership predicted admission to the psychiatric ward of a local community hospital but not to a state psychiatric hospital, so that the SC patients were significantly more likely than MFGT patients to require hospitalization at a community hospital. Compared to SC participants, the MFGT participants were equally likely to be admitted to a state psychiatric hospital during the first year of the intervention period. However, at 1 year post-treatment, MFGT participants, relative to the SC participants, had statistically lower odds of being admitted to the state psychiatric hospital, which provided evidence for the positive effects of treatment on inpatient hospitalizations beyond the treatment period.

Analyses further showed that MFGT participants utilized significantly more outpatient services than those in the SC condition during the 2-year intervention period. However, no differences were found at 1-year post-treatment. When outpatient service utilization hours were summed across the entire 3-year period (2-year intervention period plus 1-year post-treatment), no group differences were found.

Taken together, the study findings regarding outpatient services utilization and hospitalizations suggest that MFGT has the most impact on inpatient hospitalizations. A major limitation of the McDonnell study is that the results may not generalize to persons who are ethnically and culturally different from the study sample. It is important to note that the study sample was mostly male and European-American.

Chien and Wong (2007) designed a randomized control trial that addressed the cultural gap found in the MFGT literature. The study's aim was to evaluate the effects at 1-week and 1-year post-treatment of a culturally-adapted MFGT on families of Chinese patients with schizophrenia recruited from two outpatient clinics in Hong Kong. A family needs assessment was first conducted to evaluate the family relationships and values, which was heralded by the study authors as an integral component of establishing a culturally-sensitive family intervention. This information was then integrated into the treatment, and focus was given to supporting the interdependence of family members in care-giving. Specific cultural themes incorporated in the treatment included valuing collectivism over individualism and emphasizing family and kinship ties. The MFGT was based on McFarlane's four-stage model and consisted of 18 sessions; each session met every other week and lasted about two hours.

Forty-two participants consisting of patients and their family members were randomly assigned to the psychoeducation group condition, and 42 consisting of patients and their family members were assigned to the SC alone condition. The mean age of the family members was 40.6 years (SD = 7.2). The mean age of the patients was 28.8 (SD = 4.8), with average illness duration of 3.6 years (SD = 1.8) at recruitment. The outcome measures included family functioning, burden of care, patient functioning and symptom ratings, and number and length of rehospitalizations. One researcher who was blind to study group membership administered the measures at baseline before randomization to condition and at two follow-up periods: one week and 1 year after completion of treatment.

Analyses revealed that at both 1-week and 1-year post-treatment, participants in the MFGT group compared to the SC participants reported significantly greater improvement in family functioning as well as significant reductions in family burden of care and the number of patient rehospitalizations. Compared to SC participants, participants in the MFGT group showed significant improvement in functioning and a significant reduction in length of rehospitalization only at the second follow-up (at 1-year post-treatment). Due to the relatively small sample size, the study authors were unable to test whether the culturally-sensitive adaptations provided an effect above and beyond the other components of the multifamily group structure. This could be dealt with in future studies by analyzing culturally-sensitive factors separately to see if they are essential to outcome.

The two main studies reviewed above, by McDonnell and colleagues (2006) and by Chien and Wong (2007), provide support for MFGT's positive impact well beyond the duration of treatment. Both studies suggest that follow-up at one year or more would be required to identify the significant longer-term effects of MFGT. Chien and Wong's study adds to the literature by investigating the efficacy of a culturally-adapted MFGT with Chinese participants. Their study suggests that incorporating specific cultural and family needs of group members may help improve family and patient outcomes. Future methodologies for tailoring MFGT to specific ethnic populations would be useful. In addition, studies that identify treatment factors that are adaptable, while maintaining the core ingredients of the treatment approach, would ensure that MFGT is sensitive to the needs and concerns of any group.