Putting the Pieces Together: Harmonizing Egocentric and Digital Data Networks from the N2 Cohort Study (Preprint)

BACKGROUND Social network data are essential and informative for public health research and implementation as they provide details on individuals and their social context. OBJECTIVE We aimed to generate a more complete sociocentric-like “fuzzy” network by harmonizing alternative sources of egocentric and digital network data to examine relationships between participants in the N2 cohort study. Further, we propose to examine network peer effects of status neutral HIV care continuum cascade. METHODS Data were collected from January 2018 to December 2019 in Chicago, Illinois, US from a community health center and via peer referral sampling as part of the Neighborhoods and Networks (N2) Cohort Study, comprised of Black sexually minoritized men and gender expansive populations. Participants provided sociodemographics, social networks, sexual networks, mobile phone contacts, and Facebook friend list data. Lab-based information about HIV care continuum cascade was also collected. We used an experimental approach to develop and test a fuzzy matching algorithm to construct a more complete network across confidant, sexual, Facebook, and phone networks using R and Excel. We calculated social network centrality measures for each of these networks and then described the HIV care continuum within the context of each network. We then used Spearman’s correlation and a network autocorrelation model to examine social network peer effects with HIV status and care engagement. RESULTS A total of 412 participants resulted in 2,054 network connection (ties) across all confidant, sexual, phone, and Facebook networks - reaching the entire study sample in one fully connected “fuzzy” network. Results from the network autocorrelation model suggest that participants who were proximate to network members who were engaged in care were significantly more likely to be engaged in care (significant at p<0.05). CONCLUSIONS Using alternative sources of network data allowed us to fuzzy match a more complete network: fuzzy matching may identify hidden ties among participants which were missed by examining alternative sources of network data separately. Although sociocentric studies may be complex, more complete sociocentric-like networks may be generated using a fuzzy match approach that leverages peer referral, egocentric networks, and digital networks. Enriching offline networks with digital network data may provide insights into characteristics and norms that traditional egocentric approaches may not be able to capture. INTERNATIONAL REGISTERED REPORT RR2-10.2196/48548