American Journal of Psychiatry published long-term RCT results of the 1990s Fast Track mental health program for children. Quick take: Not reliable evidence; the study is inaccurately described & analyzed as if children (vs schools) were randomized, likely leading to false positive results.

​

Program:

​

• Fast Track "blended parent behavior-management training, child social-cognitive skills tutoring, home visiting, and classroom social-ecology change from grades 1 through 10 to prevent children with early-emerging conduct problems from experiencing such problems in adulthood."

  
  

Study Design & Findings:

​

• The study randomized 55 schools (with a sample of 891 kindergartners) to Fast Track vs control. An earlier, age-25 follow-up found moderately promising effects on crime & other outcomes. The new, age-34 report examines effects on the children of program participants. The goal is to see if program impacts extend to the next generation.

​

• For that next generation, the study reports (i) no discernible effects on child mental health, & (ii) statistically significant reductions in children's use of some health - including mental health - services.

  
  

• Importantly, however, the published paper doesn't mention that 55 schools were randomized & instead describes & analyzes the study as if 891 children were individually randomized. Below is the paper's inaccurate study flow diagram.

• And below is the 2022 study preregistration describing the school-level randomization. Analyzing the study as if individuals (vs schools) were randomized can easily lead to false-positive findings; thus the reported effects on children's health service use are not reliable.

Comment:

• I'd encourage the researchers to reanalyze the study correctly. It's possible some of the positive findings will hold up but, even if not, the results of such a large, long-term RCT are of great policy value.