Background/Purpose: Standard methods of recruitment for clinical research, such as traditional media advertisements, can be inefficient and expensive, especially for underserved communities and asymptomatic individuals. To address this need, we explored the use of online recruitment tools that might improve access to these populations. We evaluated the effectiveness of two online approaches to recruit subjects for an observational study of asymptomatic individuals at risk for developing rheumatoid arthritis (RA): 1) Web-based methods; and 2) Facebook-based advertisements. This strategy focused on asymptomatic first-degree relatives of individuals with RA who self-identify for an observational longitudinal study on transitions to clinical RA.

Methods: A county-wide web-based recruitment campaign was conducted from October through December 2020 consisting of image and text-based ads displayed on various websites and mobile applications. The goal was to identify individuals to participate in screening visit for a study evaluating the evolution of RA with the inclusion of individuals who were CCP3 >2x ULN and no inflammatory arthritis. Ads were targeted to potential subjects using a combination of geofencing (eg, visits to medical centers or rheumatologists), keyword retargeting (eg, “rheumatoid arthritis”), and website remarketing techniques. A Facebook-based recruitment campaign was then conducted from February through April 2021. Facebook users were targeted based on demographics, group membership, and user association. An individual that clicked on an ad was directed to a website containing information about the study, where they could fill out a contact form to provide their information to study staff. Both advertising campaigns were administered by a third-party vendor.

Results: A total of 413,289 ads were displayed during the web campaign and 392,408 ads were displayed during the Facebook campaign. Throughout the duration of the web campaign, there were 428 individual clicks on an ad (0.10% of ads). Only one person contacted the study staff and was screened but did not enroll. Throughout the duration of the Facebook campaign, there were 3649 individual clicks on an ad (0.93% of ads) and 43 people contacted the study staff for screening (p< 0.0001 for web-based vs Facebook for clicks). 34 individuals were successfully screened and two were enrolled, with mean age of 52.3±11.3 yrs. In terms of characteristics, 94% of screened individuals were female and 24% were Hispanic. The cost per click for the web campaign was $2.80 and for the Facebook campaign it was $0.71, which corresponded to a cost per self-identifying subject of $1600 for the Web campaign and $60 for the Facebook campaign.

Conclusion: The Facebook recruitment campaign resulted in more screened subjects and was more cost-effective than the web recruitment campaign. A preponderance of women responded to the campaign, and there was excellent representation among the Hispanic population. Facebook ad targeting improved access to individuals who were closely linked to rheumatoid arthritis patients in their social network and more likely to respond to the ad. The data suggest that social media campaigns can improve outreach to hard-to-reach populations.

« Back to ACR Convergence 2021

ACR Meeting Abstracts - https://acrabstracts.org/abstract/optimizing-social-media-as-a-recruitment-tool-for-hard-to-reach-populations-in-rheumatology-clinical-research/