Preventing preschool mental health problems: population-based cluster randomized controlled trial

OBJECTIVE: Prevention of child behavior problems may reduce later mental health problems. We compared the effectiveness, at the population level, of an efficacious targeted prevention program alone or following a universal parenting program. METHODS: Three-arm, cluster randomized controlled trial. One thousand three hundred fifty-three primary caregivers and healthy 8-month-old babies recruited from July 2010 to January 2011 from well-child centers (randomization unit). PRIMARY OUTCOME: Child Behavior Checklist (CBCL) externalizing and internalizing scales* at child ages 3 and 4.5 years. SECONDARY OUTCOMES: Parenting Behavior Checklist* and over-involved/protective parenting (primary caregiver report). Secondary caregivers completed starred measures at age 3. RESULTS: Retention was 76% and 77% at ages 3 and 4.5 years, respectively. At 3 years, intention-to-treat analyses found no statistically significant differences (adjusted mean difference [95% confidence interval (CI); p-value]) for externalizing (targeted vs usual care -0.2 [-1.7 to 1.2; p = .76]; combined vs usual care 0.4 [-1.1 to 1.9; p = .60]) or internalizing behavior problems (targeted vs usual care 0.2 [-1.2 to 1.6; p = .76]; combined vs usual care 0.4 [-1.1 to 2.0; p = .58]). Primary outcomes were similar at 4.5 years. At 3 years, primary and secondary caregivers reported less over-involved/protective parenting in both the combined and targeted versus usual care arm; secondary caregivers also reported less harsh discipline in the combined and targeted versus usual care arm. Mean program costs per family were A$218 (targeted arm) and A$682 (combined arm). CONCLUSION: When translated to the population level by existing staff, pre-existing programs seemed ineffective in improving child behavior, alone or in combination, but improved parenting.