High Levels of Prevention-effective Adherence to HIV PrEP

An Analysis of Substudy Data From the EPIC-NSW Trial

Benjamin R. Bavinton, PhD; Stefanie Vaccher, PhD; Fengyi Jin, PhD; Garrett P. Prestage, PhD; Martin Holt, PhD; Iryna B. Zablotska-Manos, PhD; Rebecca Guy, PhD; Janaki Amin, PhD; David J. Templeton, PhD; Barbara Yeung, MPH; Mohamed A. Hammoud, PhD; David Lewis, PhD; David Baker, MBBS; Nila Dharan, MD; Anna M. McNulty, MBBS; Andrew E. Grulich, PhD


J Acquir Immune Defic Syndr. 2021;87(4):1040-1047. 

In This Article


Between March 2016 and April 2018, 9709 participants were enrolled in EPIC-NSW and 9596 were dispensed PrEP. Of those dispensed PrEP, 6429 participants (67.0%) consented to the behavioral survey substudy and completed the baseline survey, and 4401 completed at least 1 follow-up survey and were included in the analysis (45.9% of participants dispensed PrEP, 68.5% of those completing the baseline survey). Overall, the 4401 participants included in this analysis were similar to the full EPIC-NSW cohort and those who completed the baseline survey (Table 1); there was a marginal difference in the proportion born in Australia (61.1% versus 52.7%). The included participants had a slightly higher mean MPR than all participants dispensed PrEP in the parent study (0.85 versus 0.81) and compared with the participants excluded from the analysis (0.77).

The included participants overwhelmingly identified as male (98.8%) and most identified as gay (93.0%); 334 participants (7.6%) were not cisgender gay men. Median age was 36 years [interquartile range (IQR) = 30 to 46 years). More than half (61.1%) were born in Australia (including 1.6% of the overall sample who were Indigenous Australians), 13.2% were born in Asia, and 14.2% in high-income English-speaking countries. More than half (60.2%) were university educated. Less than half (44.7%) lived in <5% gay postcodes, whereas 33.4% lived in ≥20% gay postcodes. In the 3 months before enrolling in EPIC-NSW, 14.0% had taken PrEP (90 of the 618 participants who had previously taken PrEP had been enrolled in the precursor demonstration trial, PRELUDE[18,19]). At baseline, 70.2% (n = 3090) preferred daily PrEP, 15.7% (n = 691) periodic PrEP, and 12.0% (n = 530) event-driven PrEP.

Among included participants, the median number of follow-up surveys completed was 2 (range = 1 to 6; IQR = 1 to 4). Most participants (81.9%, n = 3605) reported daily PrEP use in the 3 months before each follow-up survey, whereas 6.7% (n = 296) had at least one 3-month follow-up period of nondaily use, and 11.4% (n = 500) had at least one 3-month period during which they did not take PrEP. During follow-up, 6.9% (n = 302) of participants reported any sex work, 67.9% (n = 2989) any group sex, and 43.7% (n = 1921) any drug use for the purposes of sex. One-quarter of participants (n = 1140, 25.9%) ever used methamphetamines. Over follow-up, 53.9% (n = 2370) of participants consistently preferred daily PrEP as their ideal dosing regimen. Four participants (0.1%) included in this analysis seroconverted (of 30 seroconverters in the parent study[13]).

Participants provided information on 12,399 "last full weeks." When examining the quarterly MPR for each participant for the quarter in which they completed the survey (and thus in which the last full week occurred), there was a relationship between quarterly MPR and the number of pills taken in the last full week (Table 2). For example, those who took no pills in the last full week had a median MPR in that quarter of 0.19, compared with a median of 1.00 for those who took ≥7 pills in the last full week.

Participants reported any sexual activity with male partners in 74.8% (n = 9272) of the 12,399 observed weeks, and they reported CLS with male partners in 60.4% (n = 7485) of weeks (Figure 1 and Table 3). These CLS weeks were reported by 3417 participants (77.6%). Of the 7485 CLS weeks, ≥7 PrEP pills were taken in 85.3% (n = 6382) of weeks, 4–6 pills were taken in 8.1% (n = 607), 1–3 pills in 0.5% (n = 38), and no pills in 6.1% (n = 458) of weeks, respectively. Overall, 6.6% (n = 497) of the 7495 CLS weeks were inadequately protected by PrEP.

Figure 1.

Sexual behaviors reported in "last full week" and PrEP protection.

One-third (n = 2521, 33.7%) of the CLS weeks involved higher risk sex, that is, with HIV-positive men with detectable/unknown viral load (n = 281 weeks) or unknown HIV status men (n = 2431 weeks); these weeks were reported by 1558 participants. Of the 2521 weeks involving higher risk CLS, participants took ≥7 pills in 87.4% (n = 2204) of weeks, 4–6 pills in 8.5% (n = 214), 1–3 pills in 0.5% (n = 12), and no pills in 3.6% (n = 91). There were 103 weeks when participants did not have adequate PrEP protection and reported higher risk CLS, equating to 4.1% of the higher risk CLS weeks (n = 103/2521), 1.4% of all CLS weeks (n = 103/7485), or 0.8% of all observed weeks (n = 103/12,399). The 103 inadequately protected higher risk CLS weeks were reported by 98 individuals (2.2% of included participants). When examining the 91 weeks in which no pills were taken, 67 (73.6%) were when participants reported taking no PrEP in the previous 3 months.

Among the 2521 higher risk CLS weeks, we examined factors associated with the 103 weeks inadequately protected by PrEP (Table 4). In multivariate analysis, inadequately protected weeks were less likely to be in participants aged 45 years or older, and there was a significant trend for age. They were more likely to be in participants living in a <20% gay postcode, in participants who preferred nondaily PrEP dosing options, and in participants who completed <3 follow-up surveys. Inadequately protected higher risk CLS weeks were not independently associated with sexual and gender identity, university education, country of birth, sexual partner numbers, group sex, sex work, drug use for the purposes of sex, or methamphetamine use.