Implications for practice and research

• While there is an urgent need for interventions for improving the quality of life for children whose parents have a mental illness, at present there is a paucity of rigorous evidence to guide practice.

• Research is required that incorporates user-centred values, short-term and long-term child and parenting outcomes, high-quality cost data, and in-depth qualitative studies of the views of parents and children.

• Intervention designs and outcome measures need to be child centred.

Context

Given that one in five young people have a parent with a mental illness1 and that between 44% and 74% of these children are at a higher risk than their peers of developing mental health problems,2 it is imperative that effective programmes are delivered to break the generational cycle of mental illness. Bee and colleagues present a synthesis of the clinical effectiveness, cost-effectiveness and acceptability of interventions for improving quality of life in children in families where a parent has a mental illness.

Findings

Twenty-nine randomised control trial (RCT) or quasi-RCT studies were identified. Three trials targeted parents and/or children where the parent had an SMI, while 26 trials targeted parents and/or children where the parent had depression. Overall, trials were of poor or unclear quality. Meaningful analysis across studies was difficult given the heterogeneity of trials, outcome measures and intervention formats. Evidence of effect was found in 18 programmes designed for severely depressed mothers of infants, which predominately targeted maternal depression, rather than parenting skills or outcomes for children. One economic evaluation in relation to postnatal depression was found.

Acceptability data (intervention uptake, adherence, client satisfaction and views) were inconsistently reported, and in-depth qualitative studies of the views of parents and in particular children were sparse. Tentative conclusions suggest that children and parents value peer support, with the relationship between parents and staff being important in determining programme acceptability.

Gaps were identified in relation to the provision of programmes designed for older children, fathers and parents with diagnoses other than depression.

Commentary

Major methodological problems precluded Bee and colleagues from making conclusive statements about the evidence base of interventions to support quality of life for children living with a parent with a mental illness, a finding which resonates with previous, less structured reviews.2 ,3 Few studies collect medium and long-term follow-up effects or investigate the relationship between specific intervention components and effects.

Nicholson argues that the paucity of programme evidence may be attributable to the complexity of families’ lives and a corresponding lack of appropriate methodological approaches.4 Bee and colleagues suggest that stakeholder views need to be elicited to identify the optimal content format and delivery of interventions. They also suggest that existing parenting interventions with proven efficacy in multirisk families may be potential candidates for modification and piloting with families affected by parental mental illness. The authors urge for the routine reporting of a client's family circumstances and an integration of child-centred outcomes in intervention trials. The economic cost of intervening, as well as the cost of not intervening, warrants further investigation.

The review is clear that we do not know enough about appropriate intervention targets (eg, the child, the parent, the parent–child relationship and the family) or content (such as psychotherapy, psychoeducation) of programmes supporting families where a parent has a mental illness. Nonetheless, clinicians need to recognise the importance of the family when treating a parent's mental illness; one study found that the treatment of a parent's depression, even when successful, was insufficient to change the mother–child relationship.5 This finding suggests that parenting and child-related issues need to be incorporated into treatment protocols. Given the heterogeneous nature of these families, it is likely that a multipronged approach will be required for different family types and members.