Clusters and outbreaks contribute disproportionately to future HIV transmission in the United States

OBJECTIVES: Quantify future transmission associated with clusters of rapid transmission and among all persons with HIV (PWH); estimate the role of undiagnosed infection and diagnosed but unsuppressed infection in future transmissions. DESIGN: Network simulation model of sexually transmitted HIV in the United States that includes a cluster generation algorithm to simulate clusters of rapid transmission and associated networks. METHODS: Clusters were defined as at least three persons with diagnosed HIV in 2017 and with sequences, connected using a 0.5% genetic distance threshold. The cluster generation algorithm was then applied to all PWH (including persons with undiagnosed infection) to identify full clusters. We calculated cumulative transmissions, defined as the number of new infections during 2018-2022 per 100 PWH in 2017, overall and stratified by cluster membership, diagnosis, and viral suppression status. We simulated 1300 trials (130 runs; 10 replicates/run); medians and ranges are presented. RESULTS: Cumulative transmissions per 100 PWH were 19 [16, 22] for all PWH and 49 [0, 238] for full clusters. In full clusters, cumulative transmissions for persons with undiagnosed infection or diagnosed but virally unsuppressed infection were 85 [0, 633] and 61 [0, 300], respectively. Among all PWH, cumulative transmissions for persons with undiagnosed infection or diagnosed but virally unsuppressed infection were 59 [49, 74] and 23 [16, 29], respectively. CONCLUSIONS: Given high HIV transmission in the 5 years following cluster detection, intervening in clusters could have an outsized impact on averting future infections. Focusing testing, prevention, and care activities on persons in clusters and their networks is essential to achieve this goal.