Examining Racial/Ethnic Inclusivity and Subgroup Effects: A Review of Blueprint "Model" Programs

Background/Purpose. The Grand Challenge to Ensure Healthy Development for all Youth calls upon social workers to reduce racial and socioeconomic disparities in behavioral health problems through effective prevention interventions. The Blueprints for Healthy Youth Development is a registry of “experimentally proven programs” intended to guide decision-makers toward effective practice. Distinction as a “model” program typically requires two randomized controlled trials (RCTs), with sustained positive, and no iatrogenic, effects. It is unclear, however, who is included in these studies and whether some subgroups benefit more than others. Further analysis is required to understand whether Blueprint programs reduce, sustain, or exacerbate known disparities in the behavioral health of young people. We review the studies buttressing Blueprints “model” programs to examine whether studies are (Q1) inclusive of racially diverse youth, and (Q2) if racial subgroup effects are analyzed.

Methods. The Blueprints registry lists 17 “model” programs, based on 44 studies (M=2.6 per program), described in 62 peer-reviewed papers published from 1973 - 2018 (median year = 2006). The programs embedded in these studies were either universal (n=5), selective (n=3), indicated (n=4), or multi-tiered (n=5). Studies explored diverse outcomes (e.g., problem behavior, education, emotional wellbeing, physical health, positive relationships). Studies were coded: race/ethnicity of sample reported (y/n), racial/ethnic representativeness asserted (y/n), the racial/ethnic distribution of the sample (“no racial predominance” = no single race comprised over 50% of the study sample, “modest racial predominance” = single race >60%, “extreme racial predominance” = single race >80%); subgroup effects analyzed (y/n), a qualitative description of the methods and results of any subgroup analyses.

Results. Of the forty U.S. studies listed in the Blueprints registry, thirty-nine (97.5%) described the race/ethnicity of their sample. Of these, 31% divided participants into two racial groups (e.g., White / non-White), 43.5% into three to five groups, and 25.5% into >5 groups. Race/ethnicity was determined using self-reports (46%), administrative records (44%), or unspecified (10%). Sixteen studies (40%) mentioned the representativeness of their sample, with 11 asserting representativeness and 5 noting intentional or unintentional discrepancies with the population. Of the 39 studies reporting race/ethnicity, 11 studies (29%) reported no racial predominance. Twenty-four (60%) had a modest White predominance, 13 of which had an extreme White predominance. Three studies (8%) had an extreme Black predominance and one study had a modest Hispanic predominance. Of the 17 Blueprints programs, only 5 (29%) listed studies with stated intentions to examine treatment effects by race. Because race did not directly predict outcomes independent of other covariates, one study did not pursue an interaction. One paper stated there was “no evidence that participant ethnicity moderated the intervention” without any explication of method or data. Three programs reported significant subgroup effects, only one of which is interpreted as reducing a disparity.

Conclusions/Implications. Rigorously tested prevention programs in the U.S. transparently describe their sample race/ethnic composition. Although the majority of samples are predominantly White, very few report subgroup effects by race. Future RCTs should increase sample diversity and test subgroup effects in order to promote healthy development for *all* youth.