Unintended consequences of disseminating behavioral health evidence to policymakers: Results from a survey-based experiment

• Published in: Implementation research and practice 2023-01, Vol.4, p.26334895231172807-26334895231172807
• Main Authors: Purtle, Jonathan, Nelson, Katherine L., Lê-Scherban, Félice, Gollust, Sarah E.
• Format: Article
• Language: English
• Subjects: Short Report
• ISSN: 2633-4895; 2633-4895
• DOI: 10.1177/26334895231172807

Background Communication research demonstrates that messages often have unintended consequences, but this work has received limited attention in implementation science. This dissemination experiment sought to determine whether state-tailored policy briefs about the behavioral health consequences of adverse childhood experiences (ACEs), compared to national policy briefs on the topic, increased state legislators’/staffers’ perceptions of the policy brief relevance and parental blame for the consequences of ACEs, and whether effects differed between Democrats and Republicans. Method A preregistered, web-based survey experiment with U.S. state legislators/staffers was conducted in 2021 (n  =  133). Respondents were randomized to view a policy brief about the behavioral health consequences of ACEs that included state-tailored data (intervention condition) or national data (control condition) and then answered survey questions. Dependent variables were perceived policy brief relevance and parental blame for the consequences of ACEs. Results The mean policy brief relevance score was 4.1% higher in the intervention than in the control condition (p  =  .24), but the mean parental blame score was 16.5% higher (p  =  .02). When outcomes were dichotomized, 61.2% of respondents in the intervention condition rated parents as “very much to blame” for the consequences of ACEs compared to 37.1% in the control condition (p  =  .01). When the sample was stratified by political affiliation, the effect of the state-tailored policy brief on parental blame was larger in magnitude among Democrats and not significant among Republicans. The intervention policy brief increased the mean parental blame score by 22.8% among Democrats relative to the control policy brief (p  =  .007) and doubled the proportion rating parents as “very much to blame” (52.2% vs. 26.1%, p  =  .03). Conclusions Despite limited statistical power, state-tailored policy briefs significantly increased state legislators’/staffers’ perceptions of parental blame for the behavioral health consequences of ACEs, relative to a policy brief with national data. Unintended messaging effects warrant greater attention in dissemination research and practice. Plain Language Summary Prior studies have tested the effects of dissemination strategies on policymakers’ engagement with research evidence. However, little research has assessed the potential unintended consequences of disseminating evidence to policymakers. This knowl