Addressing Gaps in HIV Preexposure Prophylaxis Care to Reduce Racial Disparities in HIV Incidence in the United States
Jenness S,Maloney K,Smith D,Hoover K,Goodreau S,Weiss K,Liu A,Sullivan P,Rao D,Rosenberg E

Background: HIV preexposure prophylaxis (PrEP) could reduce the racial disparities in HIV incidence among black men who have sex with men (BMSM),particularly in the Southeast US where the disparities are greatest. Achieving this goal will depend on race-specific rates of movement through the PrEP continuum of care from awareness to adherence and retention.

Methods: We expanded our mathematical model of HIV transmission for MSM, which simulates PrEP based on the bio-behavioral indications of CDC’s clinical practice guidelines, to include race-stratified transitions through the PrEP continuum from awareness to access to prescription to adherence to retention.Continuum parameters - consistently equal to or poorer for BMSM compared to white MSM (WMSM) - were estimated based on our Atlanta-based HIV incidence cohorts and published PrEP open-label studies. Models were calibrated to race-specific prevalence in these cohorts. We simulated four scenarios over a ten-year period: 1) no-PrEP (reference); 2) PrEP with the observed race-specific continuum parameters; 3) PrEP with BMSM parameters set to WMSM values; and 4) PrEP with BMSM parameters set to 20% higher than WMSM values.

Results: In the reference scenario, the disparity ratio of BMSM to WMSM incidence rates was 4.71. In the second, “as-observed” scenario, 8.4% of BMSM and 23.4% of WMSM were predicted to be on PrEP. Compared to no PrEP, incidence at year 10 was lower for both BMSM (HR = 0.77) and WMSM (HR = 0.56), with 14.1% and 33.1% of cumulative infections averted respectively. This stronger benefit for WMSM increased the disparity ratio to 6.31. In the third, equal parameters scenario, the disparity ratio (4.74) returned towards the reference scenario value, with slightly higher infections averted for WMSM (35.6% vs 29.6%) as a function of their higher levels of PrEP indications. With BMSM continuum parameters set to 120% of WMSM values, the hazard ratio for BMSM was stronger than for WMSM (0.40 vs 0.50), with the disparity ratio below the reference scenario(2.91).

Conclusion: Poorer levels of PrEP awareness, access, prescription, and adherence could limit the population-level prevention effects of HIV PrEP for BMSM, leading to higher than current disparities albeit at lower incidence rates for both races. Reducing HIV disparities with PrEP will require addressing race-specific gaps along the PrEP continuum to improve rates of PrEP initiation,adherence, and retention for BMSM in the United States.