Article Text
Abstract
Background Many public policies have potentially important but poorly understood long-run consequences for health, income, public cost and inequality over the lifecourse. We aim to improve understanding by (i) developing a novel discrete time lifecourse microsimulation model of an English birth cohort, and (ii) using it to extrapolate the lifecourse consequences of a training programme (‘Incredible Years’) for parents of young children exhibiting antisocial behaviour.
Methods Model: We simulate a cohort of 100,000 English children born in 2000–1, using data from the Millennium Cohort Study (MCS) to describe their characteristics and family circumstances. We model the year-by-year evolution of lifecourse outcomes using difference equations parameterised using quasi-experimental evidence and calibrated against longitudinal survey data. Difference equations vary across four key life stages (0–4, 5–24, 25–69 and 70+) to represent the causal pathways linking family circumstances, cognitive and socio-behavioural skills formation, conduct disorder and educational attainment in childhood to diverse later life outcomes including poverty, imprisonment, social security benefits, residential care, unhealthy behaviour, physical illness, mental illness and mortality.
Intervention: We assume training is offered to parents of all children at high-risk of conduct disorder at age 5. Effects on the parent-reported Strengths and Difficulties Questionnaire Conduct Subscale are drawn from a participant-level meta-analysis, with effect size differentiated by child gender, baseline behaviour problems and parental mental health.
Primary outcomes: We summarise lifetime benefits using ‘good life-years’ which go beyond conventional quality-adjusted life-years to adjust for income as well as illness. We examine differential benefits by parental income, mental health and baseline behavioural problems.
Results We estimate that parent training increases the lifetime wellbeing of a child at risk of conduct disorder by an average of 2.43 [95% CI 1.03 to 3.83] good years. On average, children with parents suffering mental health problems gain 8.04 [95% CI 6.75 to 9.34] good years; children with severe baseline behavioural problems gain 5.37 [95% CI 3.97 to 6.77] good years; children that are poorest 20% at birth gain on average 3.58 [95% CI 2.63 to 4.54] years and children that are richest 20% at birth gain on average 1.73 [95% CI 0.77 to 2.69] years.
Conclusion Parent training can yield substantial long-run benefits in years of good life gained, especially for children from disadvantaged families, though uncertainty remains about effectiveness for parents who do not seek help.