Recruiting for a randomized clinical trial for late-life depression during COVID-19: outcomes of provider referrals vs. Facebook self-referrals

INTRODUCTION: The COVID-19 pandemic caused major impediments to the conduct of clinical research. Online recruitment, a growing area of interest even before COVID-19, may provide a strategy for facilitating access to clinical research during times of public health restriction. However, little is known about the effectiveness of online recruitment in older adults with psychiatric illness. The present study aimed to investigate the effectiveness of online recruitment via self-referrals from a Facebook advertising campaign, compared with traditional recruitment via provider referrals, during the COVID-19 pandemic. METHODS: The study period lasted from May 2020 through January 2022. During this period, we compared the performance of traditional recruitment from provider referrals versus self-referrals from Facebook during a clinical trial of pharmacotherapy for late-life depression. First, we calculated the rate of enrolment (i.e., randomization) from each referral source (yield). Next, we compared clinical and demographic characteristics of successfully enrolled participants from each source. Finally, we analyzed correlations between public health stringency and referral volume from each source over time, using the monthly average of the Bank of Canada Stringency Index for the province of Ontario. RESULTS: Twenty-four participants enrolled during the study period, 10 from provider referrals and 14 from Facebook self-referrals. Provider referrals had a significantly higher yield (30.3%) versus Facebook self-referrals (4.3%) (p<0.00001). Participants self-referred from Facebook had significantly more formal education (16.0 versus 12.6 years, p = 0.04); otherwise, participants from both sources were similar in terms of age, gender, and ethnicity. Public health stringency was negatively correlated with provider referral volume (r=–0.32) and positively correlated with Facebook self-referral volume (r=0.39) (Figure 1), though neither association reached statistical significance. CONCLUSIONS: Online recruitment is a promising avenue for improving access to clinical research in older adults with psychiatric illness, though the yield of online referrals may be lower than from traditional sources in this specialized population. We found a tendency for online recruitment to outperform traditional recruitment during times of increased public health stringency, and vice-versa, though this relationship did not reach statistical significance, possibly due to small sample sizes. When planning online recruitment campaigns, consideration must be given to the increased cost of this approach per enrolled participant; the need for a higher volume of referrals, relative to traditional methods to achieve recruitment targets; and socioeconomic barriers to computer literacy and access in this population. THIS RESEARCH WAS FUNDED BY: Parent study: PICORI, NIH.