 |
|
|
| |
US States With More PrEP Coverage See Bigger Drops in HIV Diagnoses
|
| |
| |
CROI 2024 (Conference on Retroviruses and Opportunistic Infections), March 3-6, 2024, Denver
Mark Mascolini
More state preexposure prophylaxis (PrEP) coverage correlated with steeper drops in new HIV diagnoses, according to a nationwide study by Patrick Sullivan and colleagues at Emory University in Atlanta [1]. The researchers acknowledged that other variables may contribute to falling HIV incidence, but the analysis did account for perhaps the most important other incidence-curbing factor, falling viral loads.
Sullivan and coworkers noted that PrEP use in the United States has climbed steadily since its approval in 2012; the country had an estimated 364,000 PrEP users in 2022. Previous research through 2017 confirmed a significant dose-response relationship between PrEP prescriptions per state and new HIV diagnoses in those states.
To see if that association has changed in more recent years, the Emory investigators conducted this analysis. They calculated state-by-state PrEP prescriptions using insurance data on pharmacy refills. And they used public datasets from AIDSVu.org to track PrEP users, viral suppression, and new HIV diagnoses. A CDC databank supplied estimates of people with PrEP indications (that is, people eligible for PrEP according to national guidelines). .
Sullivan and colleagues defined PrEP coverage as number of PrEP users per 100 people with an indication for PrEP. They figured estimated annual percent change (EAPC) in PrEP coverage by using Joinpoint Trend Analysis methods developed by the National Cancer Institute. This analysis controlled for yearly state-specific viral suppression.
PrEP uptake in the US flared from below 10% in 2014 to more than 90% in 2021. Over the same span, new HIV diagnoses fell gradually but steadily by about -2.3% per year. A small but consistent climb in proportion of people with HIV suppression from 2016 to 2021 must account for some of the dip in HIV diagnoses, and this analysis adjusted for the viral suppression/new diagnosis interaction.
Dividing the 50 states into quintiles of 10 states each, the researchers found that in the highest quintile 15.7% of people eligible for PrEP used it, compared with 5.8% of people in the 10 lowest-quintile states. In other words, PrEP coverage was 2.7-fold higher in the 10 states with the best PrEP coverage than in the 10 states with the least coverage. Coverage averaged 7.9 in second-quintile states, 8.8 in third-quintile states, and 11.1 in fourth-quintile states.(These percentages include low PrEP use in the early years of the 2012-2021 study period.)
Estimates controlling for the relentlessly rising viral suppression rate in the US portrayed a clear dose-response relation between PrEP coverage quintile and estimated annual percent change (EAPC) in HIV diagnoses in each of the 5 quintiles:
- Highest (5th) quintile: EAPC -8.0%
- 4th quintile: EAPC -3.5%
- 3rd quintile: EAPC -1.7%
- 2nd quintile: EAPC -0.9%
- Lowest (1st) quintile: EAPC +1.7%
The P value for this trend (0.0077) indicates this dose-response relationship is statistically significant.
Sullivan and colleagues listed a handful of factors that could confound the association between PrEP coverage and new HIV diagnoses: viral suppression, local public health investment, other prevention programs, and HIV testing programs. Although they cautioned that the PrEP coverage/HIV diagnosis association they found does not prove causality, they believe the significant dose-response relationship they confirmed and statistical adjustment for viral suppression suggest the link is causal.
The Emory team observed that HIV testing and PrEP referral programs, Medicaid expansion, and PrEP drug assistance programs are all linked to equity in PrEP use, which is critical to deriving the greatest HIV prevention benefit from PrEP.
Reference
1. Sullivan PS, Juhasz M, Le G, Brisco K, Dubose S. Association of state-level PrEP coverage and state-level HIV diagnoses, US, 2012-2021. CROI 2024 (Conference o
|
| |
|
|
|
|
|
