5 - (SYM 31) Effects of Recruitment Messaging on Ethnic/racial Minority Enrollment in an RCT for Prenatal Insomnia: A Quasi-experimental Approach

Background: Pregnant Latinx and Black people enroll in randomized controlled trials for psychological interventions less often than their non-Latinx white counterparts (Nillni et al., 2018). Differences in enrollment are explained in part by warranted mistrust of medical research by minoritized people (Scharff et al., 2010). However, initial evidence shows that recruitment materials emphasizing ethnic/racial disparities in health outcomes can increase enrollment of ethnic/racial minorities (Brown et al., 2012) and may improve the generalizability of perinatal psychotherapies.

Method: Using a quasi-experimental design, recruitment email campaigns for a randomized clinical trial (RCT) testing the effects of a digital cognitive behavioral therapy for prenatal insomnia (dCBT-I) were sent out in two waves. Email campaigns were identical except that one email identified Latinx and Black communities as having an elevated risk for poor sleep. Analyses compared screening rates, and ultimate enrollment across the disparities and non-disparities recruitment messaging. Chi squared tests were conducted to examine the association of recruitment message on enrollment related outcomes across potential participant race and ethnicity.  

Results: The first email campaign, sent in September 2023, reached 1,862 pregnant people; a second email campaign scheduled to send in April 2024 is expected to reach an additional 1,500 pregnant people. Analyses for the first campaign reveal that emails with a disparities message resulted in 461 visits to the study website and 175 completed screeners compared to 530 visits to the study website and 215 screeners for emails without a disparities message. Despite greater proportions of Latinx (χ2(1, 390)= .61, p=.25) and Black(χ2(1, 390)= 2.11, p=.09) pregnant people completing screeners following their receipt of a disparities (29% and 24%) versus non-disparities recruitment message (24% and 16%), differences were not statistically significant. Analyses will be re-run following the second wave of data collection.

Conclusion:  Recruitment messages that explicitly call out ethnic/racial disparities in insomnia were not associated with greater initial screening or enrollment of minoritized pregnant people in a RCT for dCBT-I, nor were they associated with smaller numbers of white pregnant people who screened or enrolled. Additional data-driven suggestions to enhance diversity in dCBT-I trials, including examining differences in trial exclusion by race/ethnicity will be discussed.