There is widespread acceptance that substance misuse is a prominent problem in the lives of many patients with schizophrenia (Reference Chambers, Krystal and SelfChambers et al, 2001; Reference Graham, Maslin and CopelloGraham et al, 2001), and clinical guidelines for the integrated treatment of these ‘dual disorders’ are beginning to emerge (Reference Mueser, Norrdsy and DrakeMueser et al, 2003). Furthermore, substance misuse can take a heavy toll on the carers of persons with schizophrenia (Reference Dixon, McNary and LehmanDixon et al, 1995), yet patients with dual disorders whose family members continue to be involved in their lives may have a better course of illness than people whose carers provide no support (Reference ClarkClark, 2001). Although some research suggests that integrated mental health and substance misuse services are more effective than traditional segregated services (Reference Drake, Essock and ShanerDrake et al, 2001), controlled research leaves many unanswered questions as to the impact of integrated treatment (Reference Ley, Graham, Copello and BirchwoodLey, 2003), and more research is clearly needed, especially regarding the effects of family intervention for dual disorders.

EFFECTIVENESS OF CBT AND FAMILY INTERVENTION

Haddock et al (Reference Haddock, Barrowclough and Tarrier2003, this issue) have presented a well-designed and carefully executed analysis of economic and clinical outcomes associated with a standardised programme that included individual-based cognitive–behavioural therapy (CBT) combined with a family intervention programme. Despite the small sample, which limits generalisability, the paper offers useful information for decision-makers and raises some interesting questions. For example, the finding that improvements in Global Assessment of Functioning (GAF) and Positive and Negative Syndrome Scale (PANSS) scores were significantly greater for participants in the CBT and family intervention programme than for controls, but that change in substance use was not significantly different could be due to insensitivity of the substance use measures or it might suggest that even more potent effects on substance misuse would result in even greater gains in symptoms and functioning. The results might also suggest that patients with dual disorders can benefit from the CBT and family intervention programme even if they continue to use alcohol or other drugs. From a harm reduction point of view, this is an outcome worth pursuing in subsequent studies.

COST-EFFECTIVENESS

Few cost-effectiveness analyses have been conducted evaluating the effects of different approaches to integrating mental health and substance misuse services for patients with dual disorders (Reference Clark, Teague and RickettsClark et al, 1998). Haddock et al's use of bootstrapping to model cost-effectiveness outcomes offers some important guidance for administrators who are considering CBT and family intervention as a new treatment option. In old-style analyses relying solely on parametric statistics, we would simply conclude that the cost differences were not statistically significant, without gaining further insight into the relative efficiency of CBT and family intervention compared with the usual care. However, Figs 2 and 3 in the paper by Haddock et al (Reference Haddock, Barrowclough and Tarrier2003, this issue) clearly show that CBT and family intervention are more likely to produce outcomes more efficiently than is the usual care in the majority of cases and under a range of cost-effectiveness standards. Even so, readers are cautioned to note that CBT and family intervention were not cost-effective for all patients. Unfortunately, the small sample does not allow the types of subgroup analyses that might identify the patient or provider factors associated with greater cost-effectiveness.

A FAMILY PERSPECTIVE

Haddock et al note that inclusion only of persons who receive a significant amount of assistance from family carers is a research design feature that might reduce the generalisability of findings. This is also a significant strength in the analysis. Too many studies of interventions for people with severe mental illness fail to consider how carers are affected by the intervention. By including carer outcomes a more complete description of the impact of CBT and family intervention is given. It would have been even more enlightening, though more complicated, to extend the cost-effectiveness analysis to include measures of carer opportunity costs. Several studies show that family carers for individuals with severe mental illness face significant costs in time and out-of-pocket expenditures (Reference Clark and DrakeClark & Drake, 1994; Reference ClarkClark, 2001), yet economic measures of carers’ contributions remain the exception rather than the rule in cost-effectiveness analyses. One could imagine a second cost-effectiveness analysis from the carer's perspective.

Larger studies of CBT and family intervention for patients with co-occurring mental illness and substance use disorders are certainly needed to increase our confidence that it is sufficiently cost-effective to be adopted more widely. In addition, research should be conducted to disaggregate the effects of individual CBT v. family intervention for dual disorders. However, Haddock et al demonstrate that economic evaluations of clinical interventions do not have to be large to provide useful information.

DECLARATION OF INTEREST

None.