CLINICAL HANDBOOK OF SCHIZOPHRENIA

220 IV. PSYCHOSOCIAL TREATMENTother outcome measures has been variable; consequently, there are usually inadequatesystematic data subjected to meta-analytic review. The Pitschel-Walz and colleagues(2001) review is more optimistic than that of Pilling and colleagues (2002) in the conclusionsdrawn about wider patient and family outcomes. With regard to patient outcomes,both reviews agree that there is some evidence of better medication compliance. Pitschel-Walz and colleagues also assert that there are indications of improved quality of life andbetter patient social adjustment in patients whose families were treated. Several studieshave demonstrated that these improved outcomes are achieved with reduced costs to society.As noted by Pilling and colleagues (2002), the potential benefit of the interventionsfor family members themselves has received relatively little attention. We must rememberthat although the trials sought to reduce stress in families, improvement in patient outcomes,and not family outcomes, was the prime target. When family burden was assessedas a secondary outcome, the results appear to be inconsistent. Szmukler and colleagues(2003) identified three randomized controlled trials aimed specifically at caregivers, althoughwith very brief interventions. Although these studies showed some advantages, interms of outcomes the advantages were only indirectly related to distress and burden(e.g., knowledge and attitudes). The use of different measures makes comparisons betweenstudies assessing caregiver outcomes problematic. A recent trial with a longer durationof intervention, which did focus primarily on improving caregiver outcomes, did notproduce encouraging results (Szmuckler et al., 2003). A two-phase intervention with sixsingle-family sessions, followed by 12 group family sessions, was compared with standardcare. Engagement in the trial was poor, and the authors reported that the caregiverprogram did not offer any significant advantage on any of the outcome measures: psychologicalmorbidity, negative appraisal, coping, or support. Szmuckler and colleaguesconcluded that there is still uncertainty about the most effective interventions for caregiversof patients with psychotic disorders.DISSEMINATION OF FAMILY INTERVENTIONSIn recent years there have been attempts to disseminate the benefits of family interventionin schizophrenia into routine service delivery. This has been largely through training programsdesigned to provide clinicians, mainly community psychiatric nurses, with theknowledge and skills required to implement the family work (see Tarrier, Barrowclough,Haddock, & McGovern [1999] for a review of dissemination programs). Despite thesolid evidence base for the efficacy of family-based psychological treatment programs inschizophrenia, and the efforts of the training programs, the implementation of familywork in routine mental health services has at best been patchy. The consensus view in theliterature is that family intervention implementation faces complex organizational and attitudinaldifficulties (e.g., McFarlane, Dixon, Lukens, & Luckstead, 2003), and insufficientattention has been paid to these problems in dissemination programs. In discussingthe factors that might make the transference from research to practice difficult, Mari andStreiner (1994) suggested that the requirements of durable service-oriented interventionsmay differ from those based on time-limited research models. In an attempt to demonstratethe effectiveness of family interventions in standard psychiatric settings that accountedfor these differences, a randomized controlled pragmatic trial was carried out byBarrowclough and colleagues (1999). The family intervention was based on the formalassessment of caregiver needs, and the program was carried out by a clinical psychologistin conjunction with the patient’s social worker; thus, training was in situ. The fact that