The Comparative Effectiveness of Study Recruitment and Task Completion between Social Media and Clinical Sites in SPARK, an Online Research Initiative

Background:

Research has shown that social media is an effective tool for recruiting individuals for health research. One systematic review found benefits such as reduced costs, shorter recruitment periods, and better representation including improved participant selection in young and hard to reach populations. While social media may be effective at recruiting individuals to participate in research, less is understood about whether individuals recruited via social media are more engaged or if completion rates of study tasks are higher compared to research participants recruited through traditional approaches (e.g. direct contact from a clinical site).

Objectives:

The aim of this study was to compare recruitment and task completion between individuals and families who joined SPARK from January 31 to November 15, 2018 via social media outreach and those who joined through contact with a clinical site.

Methods:

Data from all individuals who enrolled in the SPARK study from January 31 to November 15, 2018 were included in this study. The referral source was defined by the web URL from which an individual joined, specifically a social media platform URL (e.g., Facebook) or through a clinical site-specific URL (e.g., https://sparkforautism.org/UNC). For both individuals and families (biological mother, father and proband), enrollment was defined as consent to the genetic component of the SPARK study (Y/N), and tasks were defined as completion of initial surveys (Y/N) and saliva kit return (Y/N).

Results:

During the study period, a total of 20,312 individuals and 2,243 families initiated the enrollment process in SPARK via social media, compared to 8,772 individuals and 1,815 families from clinical sites. However, for both individuals and families, those recruited through a clinical site were significantly more likely to complete online enrollment in SPARK as compared to those recruited through social media (p<.05 for both individuals and families). Forty nine percent (n=9,957) of individuals and 42% (n=947) of families recruited via social media completed their first round of online surveys compared to 51% (n=4,501) of individuals and 34% (n=611) of families recruited via clinical sites (p<.05 for both individuals and families). Among those recruited via social media, 31% (n=6,352) of individuals and 43% (n=975) of families returned their saliva kit(s), compared to 60% (n=5,260) of individuals and 70% (n=1,275) of families recruited via clinical sites (p<.05 for both individuals and families).

Conclusions:

This study suggests that social media advertising may lead to enhanced study recruitment compared to traditional methods such clinical site recruitment. However, individuals and families that were recruited through a clinical site were more likely to complete online enrollment and return their saliva kits. We also found that families recruited via social media are more likely to complete online surveys. A multi-modal recruitment approach that combines social media advertising with clinical site staff to assist with task completion may improve recruitment and study participation.