Birds of a Feather: Estimating the Population Impact of Assortative Mixing by HIV Diagnosis Status and Pre-Exposure Prophylaxis Use on HIV Incidence Among Men Who Have Sex With Men Open Access

Maloney, Kevin (Summer 2021)

Permanent URL:


The advent of pre-exposure prophylaxis (PrEP) was a paradigm shift for human immunodeficiency virus (HIV) prevention. Uptake among men who have sex with men (MSM) was slow, but approximately 20% of PrEP-eligible MSM used PrEP in 2017. Despite optimism, HIV incidence among MSM has not declined as fast as projected given coverage levels. We hypothesize this may be explained by HIV serosorting and assortative mixing among MSM who use PrEP, which creates clusters of PrEP use in sexual networks and decreases the population benefit of PrEP. In this dissertation, we conducted three studies to explore the impact of PrEP sorting on HIV transmission among MSM.

In the first study, we estimated HIV serosorting and PrEP sorting patterns, using an egocentric sexual network study. We found strong evidence of assortative mixing among MSM with diagnosed HIV (39.3%), MSM who used PrEP (31.9%), and MSM who did not use PrEP (82.6%). We showed that naïve estimation of HIV and PrEP mixing matrices is biased. We presented a reclassification analysis to correct information bias.

In the second study, we used network estimation and simulation methods to describe cross-sectional sexual networks of MSM. We estimated that 45% of persistent and 24% of one-time partnerships among MSM are concordant without diagnosed HIV and without PrEP use. Network models based on degree and demographic mixing statistics produced only 70–80% of these partnerships. Our models provide evidence for inefficient network coverage of PrEP.

In the third study, we used a network-based model of HIV transmission to estimate the impact of PrEP sorting on the population benefit of PrEP. Our model showed 2.4% more infections over 10 years in the scenario with PrEP sorting compared to without PrEP sorting. The effect was relatively small, but PrEP sorting may interact with other network-level effects to limit HIV prevention in the real world.

These findings highlight the potential role of network-level factors in mediating the causal relationship between PrEP coverage and HIV incidence. Future research should investigate PrEP sorting in combination with other network properties to inform interventions to increase network coverage and the population benefit of PrEP.

Table of Contents

Chapter 1. Background and Significance 1

Chapter 2. Sexual mixing by diagnosed HIV status and pre-exposure prophylaxis use among men who have sex with men: stochastic reclassification to address information bias in egocentric network data 22

Chapter 3. Assortative Mixing by HIV Status and PrEP Use in Sexual Network Models of Men Who Have Sex with Men 43

Chapter 4. Modeling the impact of network clustered HIV pre-exposure prophylaxis on HIV incidence 66

Chapter 5. Conclusions and Public Health Implications 89

References 99

Appendix A. Chapter 2 Technical Supplement 114

Appendix B. Chapter 3 Supplementary Results 130

Appendix C. Chapter 4 Technical Supplement 140

About this Dissertation

Rights statement
  • Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
  • English
Research Field
Committee Chair / Thesis Advisor
Committee Members
Last modified

Primary PDF

Supplemental Files