Family Talk versus usual services in improving child and family psychosocial functioning in families with parental mental illness: a randomised controlled trial and cost analysis

Abstract

Background: Parental mental illness (PMI) is common and places children at high risk of developing psychological disorders. Family Talk (FT) is a well-known, whole-family, 7-session intervention designed to reduce the risk of transgenerational psychopathology. However, very few larger-scale evaluations of FT (across only a limited number of settings) have been conducted to date while there have been no cost analyses. This study aimed to assess the effectiveness and costs of delivering FT in improving child and family psychosocial functioning in families with PMI within routine mental health settings.

Methods: A total of 83 families with PMI, with children aged 5-18 years, were randomly assigned on a 2:1 ratio to receive either the FT intervention (n=55 families) or usual services (n=28 families) across 10 adult, child and primary care mental health sites in Ireland. Parental disorders included anxiety/depression (57%), Bipolar Disorder (20%), Borderline Personality Disorder (12%), Post-Traumatic Stress Disorder (8%) and psychosis (2%). Detailed assessments with parents were conducted at baseline and 6-month follow up.

Results: FT led to significant improvements in family functioning and child behaviour at 6-month follow up when compared to usual services, with medium effect sizes indicated. Parent participants with lower mental health literacy at baseline also showed significant post-intervention improvements. Those parents with less severe mental illness at baseline, and families with more partner and economic supports, reported additional significant post-intervention improvements in child depression/anxiety and parental mental health symptoms. The cost of FT amounted to €761.50 per family, although this decreased to €415.31 when recurring costs only were included.

Conclusion: The findings from this study, which was conducted within the context of a national programme to introduce family-focused practice in Ireland, demonstrate that FT is a low-cost intervention that improved child and family psychosocial functioning across different mental health disorders within routine adult, child and primary care mental health services. The findings contribute to the growing evidence base for FT, and provide a robust basis to inform practice and policy development for families with parental mental illness both in Ireland and elsewhere.

Clinical trial registration: https://www.isrctn.com/ISRCTN13365858, identifier ISRCTN13365858.

Keywords: COPMI; Family Talk; children; family-focused practice; mental disorder; mental health; parents; randomised controlled trial.

Figure 1

Family Talk sessions.

Figure 2

Study flow diagram.

Figure 3

Intervention vs. control group for family functioning (SCORE-15), ITT analysis. The black dashed line represents the line of no change at follow up (FU) from baseline (BL). The solid coloured lines represent the model means for control (green) and intervention (orange) groups, with the corresponding dashed lines being the 95% confidence intervals. The faint blue histogram indicates the distribution of baseline data. The histogram is not to scale, and is intended to be purely suggestive. Bar height values must not be read from the chart.

Figure 4

Intervention vs. control group for child behaviour (SDQ), ITT analysis. The black dashed line represents the line of no change at follow up (FU) from baseline (BL). The solid coloured lines represent the model means for control (green) and intervention (orange) groups, with the corresponding dashed lines being the 95% confidence intervals. The faint blue histogram indicates the distribution of baseline data. The histogram is not to scale, and is intended to be purely suggestive. Bar height values must not be read from the chart.

Figure 5

Intervention vs. control group for parental understanding of mental illness (PUMI), ITT analysis. The black dashed line represents the line of no change at follow up (FU) from baseline (BL). The solid coloured lines represent the model means for control (green) and intervention (orange) groups, with the corresponding dashed lines being the 95% confidence intervals. The faint blue histogram indicates the distribution of baseline data. The histogram is not to scale, and is intended to be purely suggestive. Bar height values must not be read from the chart.