JMIR Publications

ABSTRACT

Background:

Facebook, a popular social media site, allows users to communicate and exchange information. Social media sites can also be used as databases to search for individuals, including cohort participants. Retaining and tracking cohort participants is essential for the validity and generalizability of data in longitudinal research. Despite numerous strategies to minimize loss to follow-up, maintaining contact with participants is time consuming and resource intensive. Social media may provide alternative methods of contacting participants who consented to follow-up, but could not be reached, thus are potentially ‘lost to follow-up’.

Objective:

The aim of this study was to determine if Facebook was a feasible method for identifying and contacting participants of a longitudinal pregnancy cohort who were lost to follow-up, and re-engaging them without selection bias.

Methods:

This study used data from the All Our Families cohort. Of the 2827 mother-child dyads within the cohort, 237 participants were lost to follow-up. Participants were considered lost to follow-up if they had agreed to participate in additional research, completed at least one of the perinatal questionnaires, did not complete the 5-year postpartum questionnaire, and could not be contacted after numerous attempts via phone, email or mail. Participants were considered to be matched to a Facebook profile if two or more characteristics matched information previously collected. Participants were sent both a friend request and a personal message through the study’s Facebook page, and were invited to verify their enrollment in the study. The authors deemed a friend request was necessary due to the reduced functionality of non-friend direct messaging at the time. If the participant accepted the study’s friend request, then a personalized message was sent. Participants were considered re-connected if they accepted the friend request or responded to any messages. Participants were considered re-engaged if they provided up-to-date contact information.

Results:

Compared to the overall cohort, participants who were lost-to-follow-up (n=237) were younger (P = .003), non-married (P = .023), had lower household income (P < .001), less education (P < .001) and self-identified as being part of an ethnic minority (P = .015). Of the 237 participants considered lost to follow-up, 113 (47.7%) participants were identified using Facebook. Among the 113 identified participants, 87 were contacted (77.0%), 37 were re-connected (32.7%) and 20 were re-engaged (17.7%). No significant differences were found between those identified on Facebook (n=113) and those who were not able to be identified (n=124).

Conclusions:

Facebook identified 48% (113/237) of participants who were considered lost to follow-up and may be a practical tool for reconnecting with participants. The results from this study demonstrate that social networking sites, such as Facebook could be included in the development of retention practices, and can be implemented at any point in cohort follow-up.