icon-    folder.gif   Conference Reports for NATAP  
 
  CDC 2019 National
HIV Prevention Conference
March 18-21, 2019
Atlanta
Back grey_arrow_rt.gif
 
 
 
PrEP prescribers multiply as adherence
and persistence get high marks
 
 
  2019 National HIV Prevention Conference, March 18-21, 2019, Atlanta
 
Mark Mascolini
 
US preexposure prophylaxis (PrEP) prescribers multiplied more than 5-fold from 2014 through 2017, according to analysis of a huge pharmacy database [1]. A separate study of private insurance and Medicaid rolls figured that half of commercially insured US PrEP users had at least 90% adherence to this HIV-thwarting two-drug strategy [2]. Other work unveiled at the 2019 National HIV Prevention Conference found that chemsex is not a barrier to PrEP adherence by MSM [3,4] and that a simple HIV risk prediction score can help pinpoint men who would benefit from starting PrEP in an STI clinic [5].
 
More than 5-fold jump in US PrEP prescribers
 
In 2017 the United States had more than 5 times as many PrEP prescribers as it did in 2014, the year the Centers for Disease Control and Prevention (CDC) recommended PrEP for high-risk people [1]. The finding reflects numbers extracted from a pharmacy database, IQVIA, that includes more than 90% of retail prescriptions in the country. From 2014 through 2017, the number of tenofovir/emtricitabine (TDF/FTC) PrEP prescribers jumped from 6368 to 34,337, a 5.4-fold increase.
 
Physicians made up 80.6% of PrEP prescribers in 2014, and that proportion waned to 73.0% in 2017. Across that same span, the proportion of nurse practitioner prescribers rose from 10.4% to 16.1%, and the proportion of physician assistant prescribers climbed from 7.0% to 8.3%. Three quarters of all PrEP prescribers (74.4%) worked in urban areas, 29.9% in the West, 29.1% in the South, 24.2% in the Northeast, and 16.7% in the Midwest (Figure 1).
 

0322191

Figure 1. Proportions of PrEP prescribers in four US regions in 2014-2017. (Source: Zhu W, et al, CDC [1].)
 
Average number of PrEP patients rose most from 2014 to 2017 among nurse practitioners (2.9 to 6.5), followed by physician assistants (3.3 to 5.6) and trailed by physicians (2.2 to 3.5). Average PrEP patient numbers climbed from 2.3 to 3.9 in urban areas, from 1.8 to 3.0 in partially rural areas, and from 1.1 to 1.5 in rural regions.
 
The CDC investigators suggested that "a continuing increasing trend in the number of nurse practitioners and physician assistants who prescribe PrEP can enhance the capacity for HIV prevention with PrEP for key populations in both urban and rural settings."
 
Better PrEP adherence, persistence with private insurance versus Medicaid
 
Adherence to TDF/FTC PrEP in 2012-2016 proved better among people using private commercial insurance than in Medicaid users, according to a CDC analysis of Truven Health MarketScan databases [2]. PrEP persistence--time from initial PrEP prescription fill to a fill gap of more than 30 days--proved longer in the commercially insured group.
 
CDC researchers used the Truven databases to create a cohort of 18-to-64-year-olds who began PrEP between January 2012 and December 2016. No one got diagnosed with HIV during a 12-month observation period. The CDC team measured adherence as proportion of days covered (PDC), which figures the percentage of days a person has prescribed PrEP drugs during the 12-month period. They considered a PDC of 90% or higher sufficient adherence.
 
The study groups included 4172 commercially insured PrEP users and 177 Medicaid PrEP users. Median PDC-measured 12-month adherence stood at 89.0% in commercially insured people and 71.0% in Medicaid users (Table 1). Respective proportions with a PDC at or above 90% were 49.0% and 24.3%. Median gap between PrEP refills stood at 7.7 days in the commercially insured group and 9.5 days in the Medicaid group.

0322192

Median PrEP persistence came to 13.7 months in people with commercial insurance and about half that, 7.2 months, in PrEP users on Medicaid. In the Medicaid group 55.3% had 6-month persistence; that proportion dwindled to 32.1% for 12-month persistence. PrEP persistence proved significantly shorter in younger than older people in both insurance groups. Compared with white Medicaid users, black Medicaid users has significantly shorter median PrEP persistence (4.0 versus 8.1 months, P = 0.0127).
 
Because younger people get infected with HIV at higher rates than older people in the United States, the CDC investigators proposed that "interventions designed to improve younger PrEP users' adherence and persistence might increase the impact of PrEP on HIV incidence."
 
from study authors: PrEP for HIV prevention is a key strategy in "Ending the HIV Epidemic". For PrEP to be effective, it must be taken as directed by a physician and during periods of time when a person is at-risk of acquiring HIV. It's important to point out that we were unable to correlate our persistence estimates with persons' risk behaviors. Hence, it is possible that some of those who stopped using PrEP may no longer need PrEP. Other studies also showed that it is common for young users to start, stop, and re-start PrEP. Our study only examined the first continuous episode of PrEP use and may miss persons who used PrEP on demand. We are planning a study to investigate HIV infection among those who had lower PrEP adherence or discontinued PrEP to better understand the impact of low persistence.
 
Chemsex not tied to poor PrEP adherence in English MSM
 
Having sex while using psychoactive drugs, called chemsex, did not boost chances of poor PrEP adherence in a study of 388 English MSM who took part in the PROUD PrEP trial [3]. The finding is important because chemsex does raise chances of HIV transmission.
 
PROUD randomized 544 MSM who had condomless anal sex to immediate daily TDF/FTC PrEP or to defer PrEP for 1 year [4]. During follow-up 3 men in the immediate group and 20 in the deferred group picked up HIV infection, numbers that translated into an 86% lower HIV risk in the immediate-PrEP group. The chemsex analysis involved 388 PROUD participants who self-reported PrEP adherence and chemsex use, defined as sex under the influence of mephedrone, crystal methamphetamine, or gamma-hydroxybutyrate. The researchers used univariable and multivariable regression analyses to identify associations between chemsex and suboptimal (less than 100%) adherence.
 
On 1479 trial visit forms, men reported recent chemsex on 570 forms (38.5%) and suboptimal PrEP adherence on 442 forms (29.9%). Univariable analysis linked suboptimal adherence to age 18 to 25, current full-time student status, and perceiving poor adherence when entering the study. An analysis that adjusted for age and initial perception of poor adherence saw no association between chemsex and poor PrEP adherence (adjusted odds ratio 1.24, 95% confidence interval 0.88 to 1.75, P = 0.227).
 
PROUD investigators believe these new findings "suggest PrEP remains a feasible and effective HIV prevention method for high-risk MSM engaging in chemsex." They endorsed regular PrEP monitoring to help chemsex users manage their risk.
 
HIV risk prediction score for MSM could inform PrEP use
 
Three behavioral and clinical variables predicted new HIV infection in an analysis of almost 12,000 high-risk MSM attending a Seattle STI clinic [5]. University of Washington researchers proposed that the tool could help clinicians and public health agencies pinpoint MSM who would benefit from PrEP.
 
The University of Washington team began their work by creating a retrospective longitudinal cohort of men attending a Seattle STI clinic who had (1) at least two negative HIV tests between January 2001 and December 2015 or (2) one negative HIV test and a subsequent positive test. They randomly split these men in a 2-to-1 ratio to create a derivation dataset (to identify HIV diagnosis predictors) and a validation set (to test the accuracy of identified predictors). Sexual and clinical prediction variables came from data gathered at each man's initial clinic visit. The researchers used a standard statistical technique to figure out which variables were the strongest predictors and so should remain in the prediction model. Then they repeated this exercise after limiting the dataset to January 2011 to December 2015 to reflect the current antiretroviral era.
 
The study group included 11,883 MSM who made 28,621 visits. Two thirds of the group were white, 9% black, 6% Asian, and 5% Hispanic. Median age stood at 32 years. Through 14.3 years of follow-up, 693 MSM had a first positive HIV test to yield a 1-year cumulative HIV diagnosis incidence of 1.04% (95% confidence interval [CI] 0.97% to 1.12%). The full prediction model incorporated 12 variables, including mostly clinical and behavioral factors (Table 2).
 

0322193

A receiver operating characteristic (ROC) curve of 0.69 (95% CI 0.65 to 0.72) indicated that this 12-item model does a fairly good job predicting HIV risk. The researchers determined that a risk score cut point of 8+ had 67% sensitivity and 61% specificity in predicting new HIV infection. (Sensitivity is the ability of the test to identify people who will get HIV infection, while specificity is the ability of the test to identify people who will not get HIV.) Cumulative 1-year HIV diagnosis incidence was 0.5% with a risk score of 0 to 7 and 2.2% with a risk score of 8 or higher.
 
When the researchers limited the analysis to 2011-2015, they came up with an HIV risk prediction tool including only 5 factors (Table 2). The ROC curve indicating HIV-prediction accuracy with this 4-item model was 0.60 (95% CI 0.53 to 0.67). A risk score cut point of 6+ had 55% sensitivity and 60% specificity in predicting HIV. One-year cumulative HIV incidence was 1.0% for scores of 0 to 5 and 2.8% for scores of 6 or higher.
 
Finally the researchers derived a simplified model for clinical use including only 3 variables (Table 2). The ROC curve came close to the value for the 2011-2015 model, 0.59 (95% CI 0.52 to 0.65). A risk score cut point of 1+ had an HIV-predicting sensitivity of 58% and specificity of 58%.
 
The University of Seattle team called for future work to validate the model in other US populations, including a more racially diverse group, a Southeastern US population, and a group not gathered at an STI clinic.
 
References
 
1. Zhu W, Huang YL, Hoover K. HIV pre-exposure prophylaxis (PrEP) prescribers increased five-fold in the United States from 2014-2017. 2019 National HIV Prevention Conference, March 18-21, 2019, Atlanta. Abstract 6193.
 
2. Huang YL, Tao G, Hoover K. Adherence to and persistence with HIV pre-exposure prophylaxis in the United States, 2012-2016. 2019 National HIV Prevention Conference, March 18-21, 2019, Atlanta. Abstract 5214.
 
3. O'Halloran C, Rice B, Desai M, Gafos M. Chemsex is not a barrier to PrEP adherence among men who have sex with men in England. 2019 National HIV Prevention Conference, March 18-21, 2019, Atlanta. Abstract 6134.
 
4. McCormack S, Dunn DT, Desai M, et al. Pre-exposure prophylaxis to prevent the acquisition of HIV-1 infection (PROUD): effectiveness results from the pilot phase of a pragmatic open-label randomised trial. Lancet. 2016;387:53-60. https://www.ncbi.nlm.nih.gov/pubmed/26364263
 
5. Tordoff D, Barbee L, Khosropour C, Hughes J, Golden M. Derivation and validation of an HIV risk prediction score among men who have sex with men to inform PrEP initiation in an STD clinic setting. 2019 National HIV Prevention Conference, March 18-21, 2019, Atlanta. Abstract 6008.