Leveraging social networks for identification of people living with HIV who are virally unsuppressed

Published in AIDS, 2024

Recommended citation: Cummins B, Johnson K, Schneider JA, Del Vecchio N, Moshiri N, Wertheim JO, Goyal R, Skaathun B (2024). "Leveraging social networks for identification of people living with HIV who are virally unsuppressed." AIDS. 38(2):245–254. doi:10.1097/QAD.0000000000003767


This study investigates primary peer-referral engagement (PRE) strategies to assess which strategy results in engaging higher numbers of people living with HIV (PLWH) who are virally unsuppressed.


We develop a modeling study that simulates an HIV epidemic (transmission, disease progression, and viral evolution) over 6 years using an agent-based model followed by simulating PRE strategies. We investigate two PRE strategies where referrals are based on social network strategies (SNS) or sexual partner contact tracing (SPCT).


We parameterize, calibrate, and validate our study using data from Chicago on Black sexual minority men to assess these strategies for a population with high incidence and prevalence of HIV. For each strategy we calculate the number of PLWH recruited who are undiagnosed or out- of-care and the number of direct or indirect transmissions.


SNS and SPCT identify 256.5 (95% CI: [234, 279]) and 15 (95% CI: [7, 27]) PLWH, respectively. Of these, SNS identifies approximately 159 (95% CI: [142, 177]) PLWH out-of-care and 32 (95% CI: [21, 43]]) PLWH undiagnosed compared to 9 (95% CI: [3, 18]) and 2 (95% CI: [0, 5]) for SPCT. SNS identified 15.5 (95% CI: [6, 25]) and 7.5 (95% CI: [2, 11]]) indirect and direct transmission pairs, while SPCT identified 6 (95% CI: [0, 8]) and 5 (95% CI: [0, 8]), respectively.


With no testing constraints, SNS is the more effective strategy to identify undiagnosed and out- of-care PLWH. Neither strategy is successful at identifying sufficient indirect or direct transmission pairs to investigate transmission networks.