Secondary Logo

Journal Logo

EPIDEMIOLOGY AND SOCIAL: CONCISE COMMUNICATIONS

Increasing preexposure prophylaxis use and ‘net prevention coverage’ in behavioural surveillance of Australian gay and bisexual men

Holt, Martina; Broady, Timothy R.a; Mao, Limina; Chan, Curtisb; Rule, Johnc; Ellard, Jeanned; O’Donnell, Darrylb,d; Grulich, Andrew E.b; Prestage, Garrettb; Bavinton, Benjamin R.b

Author Information
doi: 10.1097/QAD.0000000000002797

Abstract

Introduction

The introduction of preexposure prophylaxis (PrEP) has transformed HIV prevention by gay and bisexual men (GBM), Australia's most HIV-affected population [1]. Following the introduction of PrEP, new HIV infections fell for the first time in a decade, with reductions concentrated among Australian-born GBM [1–3].

Up to 18–26% of HIV-negative GBM were using PrEP by 2018 in Australia [4–6], similar to the level (20%) found in 2017 in the United States [7]. Echoing US research [8], we found that the early increase in PrEP use in Australia was associated with an equally rapid decline in condom use [9,10]. This raised questions about whether the early benefits of PrEP could be built upon or sustained, whether reduced condom use might undermine reductions in HIV infections, and whether the promotion of combination prevention methods could increase overall levels of protection in the community.

In this analysis, we assess further changes in PrEP uptake and condom use, as well as the overall level of protection offered by combination prevention methods (what we are referring to as ‘net prevention coverage’, explained below). We also consider whether the characteristics and behaviour of ‘at risk’ GBM have changed as PrEP use has proceeded, that is whether they appear more or less at risk of HIV.

Methods

Data were collected in ongoing, repeated, behavioural surveillance of GBM [11]. Participants were approached by peer recruiters at venues and events, supplemented by online recruitment. Recruitment occurred in seven states and territories every 1–2 years. Eligible participants were aged 18 or over (face-to-face recruitment) or 16 or over (online), identified as gay or bisexual or had sex with a man in the previous 5 years. Participants completed an anonymous questionnaire about demographics, sexual behaviour, testing, prevention, HIV treatment, and drug use. The study was approved by the Human Research Ethics Committee of UNSW Sydney.

Our primary outcome was a categorical measure of sexual practices and the use of HIV prevention methods with casual male partners in the 6 months prior to survey, constructed from questions about sex with casual partners, self-reported HIV status, HIV treatment and PrEP use (see Supplemental Digital Content, https://links.lww.com/QAD/B963). Casual sex remains the primary context for HIV transmission between GBM in Australia [12]. The previously published measure was adapted to form six mutually exclusive categories [9,11]:

  • (1) No anal intercourse with casual partners (participants of any HIV status).
  • (2) Consistent condom use with casual partners (participants of any HIV status).
  • (3) Any condomless anal intercourse with casual partners (CAIC) by HIV-positive participants on HIV treatment with an undetectable viral load (UVL).
  • (4) Any CAIC by HIV-negative participants on PrEP.
  • (5) Any CAIC by HIV-positive participants not on HIV treatment or with a detectable viral load.
  • (6) Any CAIC by HIV-negative or untested participants not on PrEP.

Categories 1–4 were classified as ‘safe sex’ and 5 and 6 as presenting a risk of HIV transmission or infection. We define ‘net prevention coverage’ as the total proportion of participants who reported safe practices (the sum of categories 1–4). This is a conservative measure of prevention coverage, as it focuses on what participants did, and does not consider partners’ practices. For example, participants in categories 5 and 6 may be having CAIC with partners of the same perceived HIV status (serosorting), with partners on PrEP (PrEP sorting) or partners who have an UVL (viral load sorting).

We report trends in the above categories and net prevention coverage between 2014 and 2019. Trends over time are tested using binary logistic regression (e.g. no anal intercourse vs. any anal intercourse, any CAIC vs. no CAIC) with year as the independent variable. We also assessed the characteristics and behaviour of participants in category 6 (i.e. ‘at risk’ participants) over the same period (2014–2019). Dependent variables with a significant bivariate relationship (P < 0.05) with survey year (the independent variable) were included in the multivariate logistic regression model. Crude and adjusted odds ratios (ORs) and confidence intervals (CIs) are reported.

Results

Participant characteristics

A total of 52 394 surveys were completed between 2014 and 2019. Of these participants, 32 048 (61.2%) reported sex with casual male partners and are included here. The median age of these 32 048 participants was 34 years (interquartile range = 27–45). They resided in New South Wales (n = 10 822; 33.8%), Victoria (n = 10 378; 32.4%), Queensland (n = 6591; 20.6%), or other states and territories (n = 4257; 13.3%). Most identified as gay (n = 28 842; 90.3%) or bisexual (n = 2230; 7.0%), and were born in Australia (n = 22 465; 70.4%). A minority reported Aboriginal or Torres Strait Islander heritage (n = 1083; 3.4%). Participants self-reported HIV-negative (n = 26 235; 81.9%), HIV-positive (n = 3066; 9.6%) or untested/unknown status (n = 2747; 8.6%). Most HIV-positive participants were on treatment and had an UVL (n = 2641/3066; 86.1%). The proportion of HIV-negative participants on PrEP rose from 1.2% (57/4769) in 2014 to 33.7% (2303/6842) in 2019.

Trends in sexual practices and HIV prevention coverage during casual sex

Figure 1 and Table 1 show the trends in sexual practices that participants reported with casual male partners in the 6 months prior to survey. The proportions of participants who did not have anal intercourse (category 1) or who had consistent condom use (category 2) declined. The proportion who reported any CAIC increased. The group who were HIV-positive, had an UVL and reported CAIC (category 3) increased slightly. Conversely, the group who were HIV-positive, had a detectable viral load or were not on treatment and reported CAIC (category 5) decreased. The largest change was the increase in the proportion who were HIV-negative, on PrEP and had CAIC (category 4). The group of ‘at risk’ participants (HIV-negative and untested participants not using PrEP) decreased. Among all participants with casual male partners, net HIV prevention coverage in the sample (categories 1–4) increased.

F1
Fig. 1:
Sexual practices with casual male partners in the 6 months prior to survey, by participant HIV status and use of antiretroviral-based prevention.
Table 1 - Trends in sexual practices reported by participants with casual male partners in the 6 months prior to survey.
2014, n (%) 2015, n (%) 2016, n (%) 2017, n (%) 2018, n (%) 2019, n (%) Odds ratio (95% CI) Trend (P value)
No anal intercourse 806 (18.0) 925 (19.1) 1048 (17.9) 1008 (17.2) 933 (17.5) 840 (14.8) 0.96 (0.94–0.97) <0.001
Consistent condom use 1994 (44.6) 2042 (42.2) 2340 (39.9) 1771 (30.3) 1406 (26.3) 1315 (23.2) 0.81 (0.80–0.82) <0.001
Any CAI 1672 (37.4) 1875 (38.7) 2479 (42.3) 3074 (52.5) 3006 (56.2) 3514 (62.0) 1.25 (1.23–1.26) <0.001
Subcategories of participants who had CAI
 HIV-positive on treatment with undetectable viral load 215 (4.8) 281 (5.8) 331 (5.6) 393 (6.7) 365 (6.8) 331 (5.8) 1.05 (1.02–1.08) 0.002
 HIV-negative on PrEP 32 (0.7) 59 (1.2) 265 (4.5) 911 (15.6) 1125 (21.1) 1763 (31.1) 2.07 (2.01–2.13) <0.001
 HIV-positive not on treatment or detectable viral load 70 (1.6) 52 (1.1) 40 (0.7) 32 (0.6) 31 (0.6) 32 (0.6) 0.79 (0.74–0.86) <0.001
 HIV-negative/untested not on PrEP 1335 (30.3) 1483 (30.6) 1843 (31.4) 1738 (29.7) 1485 (27.8) 1388 (24.5) 0.94 (0.93–0.96) <0.001
Net prevention coverage 3047 (68.1) 3307 (68.3) 3984 (67.9) 4083 (69.8) 3829 (71.7) 4249 (74.9) 1.07 (1.05–1.08) <0.001
Total 4472 4842 5867 5853 5345 5669
Note: The odds ratios are from logistic regression tests for changes over time. An odds ratio less than one means that the relative odds of participants being in that category (e.g. consistent condom use) has decreased over time. An odds ratio over one means that the relative odds of participants being in that category (e.g. HIV-negative on PrEP) has increased over time. CAI, condomless anal intercourse; CI, confidence interval; PrEP, preexposure prophylaxis.

Focusing solely on participants who had CAIC (categories 3–6), the proportion reporting an UVL or PrEP use (i.e. CAIC protected by biomedical prevention) increased from 14.8% (247/1672) in 2014 to 59.6% in 2019 (2094/3514; OR = 1.59, 95% CI 1.55–1.62, P < 0.001).

Restricting to HIV-negative and untested participants who were not using PrEP (non-PrEP-users), the proportion reporting no anal intercourse with casual partners increased from 18.7% (733/3930) in 2014 to 20.2% in 2019 (616/3046; OR = 1.03, 95% CI 1.01–1.05, P = 0.005), the proportion who reported consistent condom use declined from 47.0% (1847/3930) to 35.1% (1069/3046; OR = 0.89, 95% CI 0.88–0.90, P < 0.001), and the proportion who reported any CAIC increased from 34.4% (1350/3930) to 44.7% (1361/3046; OR = 1.10, 95% CI 1.09–1.12, P < 0.001).

Analysis of characteristics of HIV-negative and untested men at risk of HIV

Table 2 shows trends in the characteristics of HIV-negative and untested participants who were not using PrEP but did report CAIC (‘at risk’ participants; category 6). Between 2014 and 2019, the proportions of ‘at risk’ participants who identified as bisexual (7.2–9.5%), were born overseas (24.2–30.2%) and had a university degree (41.5–50.7%) increased. ‘At risk’ participants became less likely to report more than 10 recent male partners (39.6–30.3%), but more likely to report condomless anal intercourse with regular male partners (49.8–54.7%) and frequent CAIC (23.7–30.1%). They reported more frequent viral load sorting (9.6–14.9%) and recent HIV testing (61.6–68.5%) but became less likely to report injecting drug use (7.5–3.7%) or the use of drugs for sex (31.0–22.3%). PrEP sorting was included in the questionnaire from 2017 onwards. Between 2017 and 2019, ‘at risk’ participants became more likely to report frequent PrEP sorting during CAIC, increasing from 21.4% in 2017 to 33.9% in 2019 (OR = 1.37, 95% CI 1.27–1.49, P < 0.001).

Table 2 - Trends in characteristics of HIV-negative and untested participants who were not using preexposure prophylaxis but reported condomless anal intercourse with casual male partners in the 6 months prior to survey.
2014, n (%) 2015, n (%) 2016, n (%) 2017, n (%) 2018, n (%) 2019, n (%) Odds ratio (95% CI) Trend (P value) Adjusted odds ratio (95% CI) Trend (P value)
Age, M (IQR) 34.2 (25–42) 34.3 (25–42) 34.6 (25–42) 34.0 (25–41) 35.0 (25–44) 35.2 (25–43) 1.21 (1.04–1.41) 0.014 1.17 (1.00–1.37) 0.056
Sexual identity
 Gay 1218 (90.2) 1353 (91.4) 1657 (90.2) 1561 (90.4) 1331 (90.2) 1208 (87.3) 0.95 (0.91–0.99) 0.009 0.95 (0.90–0.99) 0.018
 Bisexual 97 (7.2) 91 (6.1) 127 (6.9) 118 (6.8) 110 (7.5) 131 (9.5) 1.07 (1.02–1.12) 0.008 1.07 (1.01–1.12) 0.016
Born overseas 325 (24.2) 357 (24.1) 470 (25.6) 460 (26.6) 412 (28.0) 418 (30.2) 1.07 (1.04–1.10) <0.001 1.05 (1.01–1.08) 0.005
Aboriginal or Torres Strait Islander 56 (4.2) 57 (3.9) 50 (2.7) 59 (3.4) 44 (3.0) 64 (4.6) 1.00 (0.94–1.07) 0.892
Full time employment 850 (62.9) 943 (63.8) 1184 (64.5) 1112 (64.2) 919 (62.0) 884 (63.7) 1.00 (0.97–1.02) 0.843
University degree 560 (41.5) 669 (45.3) 857 (46.7) 843 (48.7) 698 (47.2) 703 (50.7) 1.06 (1.04–1.10) <0.001 1.05 (1.03–1.08) <0.001
No. of male partnersa
 0–1 61 (4.5) 93 (6.3) 125 (6.8) 117 (6.8) 102 (6.9) 106 (7.7) 1.08 (1.03–1.14) 0.002 1.08 (1.02–1.15) 0.010
 2–5 439 (32.7) 462 (31.3) 584 (31.9) 563 (32.6) 528 (35.7) 517 (37.4) 1.05 (1.02–1.08) <0.001 1.06 (1.03–1.09) <0.001
 6–10 312 (23.2) 361 (24.4) 423 (23.1) 415 (24.0) 347 (23.5) 339 (24.6) 1.01 (.98–1.04) 0.639 1.02 (0.98–1.05) 0.307
 >10 532 (39.6) 561 (38.0) 697 (38.1) 634 (36.7) 503 (34.0) 419 (30.3) 0.57 (0.54–0.59) <0.001 0.91 (0.88–0.94) <0.001
Group sexa 702 (52.5) 796 (53.8) 993 (54.2) 887 (51.3) 762 (51.5) 693 (50.3) 0.98 (0.95–1.00) 0.048 0.99 (0.96–1.02) 0.425
Any condomless anal intercourse with regular male partnersa 675 (49.8) 740 (49.9) 961 (52.1) 935 (53.8) 754 (50.8) 759 (54.7) 1.03 (1.01–1.06) 0.011 1.03 (1.00–1.06) 0.021
Frequent CAICa 321 (23.7) 342 (23.1) 457 (24.8) 499 (28.7) 435 (29.3) 418 (30.1) 1.09 (1.06–1.12) <0.001 1.10 (1.07–1.14) <0.001
Frequent serosorting during CAICa 676 (49.9) 782 (52.7) 942 (51.1) 862 (49.6) 731 (49.2) 717 (51.7) 1.00 (0.97–1.02) 0.696
Frequent viral load sorting during CAICa 130 (9.6) 182 (12.3) 236 (12.8) 243 (14.0) 218 (14.7) 207 (14.9) 1.09 (1.05–1.13) <0.001 1.08 (1.04–1.13) <0.001
HIV testa 805 (61.6) 973 (67.8) 1287 (71.8) 1184 (71.6) 996 (69.5) 929 (68.5) 1.05 (1.02–1.08) <0.001 1.06 (1.03–1.09) <0.001
Postexposure prophylaxis usea 63 (5.6) 99 (8.0) 146 (9.3) 120 (8.4) 96 (7.9) 61 (5.4) 0.99 (0.94–1.04) 0.745
Any injecting drug usea 92 (7.5) 71 (5.3) 93 (5.3) 62 (3.8) 48 (3.5) 48 (3.7) 0.85 (0.80–0.90) <0.001 0.87 (0.82–0.94) <0.001
Use of party drugs for sexa 384 (31.0) 353 (26.4) 439 (25.1) 409 (24.9) 353 (25.3) 292 (22.3) 0.93 (0.91–0.96) <0.001 0.96 (0.92–0.99) 0.008
Total 1355 1483 1843 1738 1485 1388
CAIC, condomless anal intercourse with casual male partners; CI, confidence interval; IQR, interquartile range; M, mean.
aIn the 6 months prior to survey.

Discussion

Our analysis of national behavioural surveillance data demonstrates a historic change in the sexual practices of Australian GBM. Between 2014 and 2019, the most commonly used HIV prevention strategy by GBM with casual partners changed from condom use to PrEP. Consistent condom use with casual partners fell, while PrEP use increased dramatically. In the same period, net prevention coverage, or the proportion of GBM with casual partners who reported safe practices with those partners, inclusive of condoms, PrEP and UVL, increased. Our previous finding of community-level risk compensation [9,10], decreasing condom use by non-PrEP-users after PrEP rollout, has continued but appears to have been outweighed by the increasing level of prevention coverage in the population. We believe these shifts in practice and increase in prevention coverage are important contributing factors to recent declines in HIV infections among Australian GBM [1–3], with annual HIV notifications among GBM falling from 811 in 2014 to 660 in 2017 (the most recent year national data are available) [1].

Assessing the characteristics of ‘at risk’ GBM (HIV-negative and untested men who had CAIC but didn’t use PrEP) suggests this group has become less at risk of HIV over time, as they reported fewer male partners, more recent HIV testing, less injecting drug use and drug use for sex. This group reported more frequent condomless sex with casual partners (consistent with declining levels of condom use in the sample) but also more PrEP and viral load sorting. This suggests that the proportion of sex acts that pose a risk of HIV transmission has fallen, even when participants are not personally protected by strategies like PrEP or condoms. Our analysis suggests that many GBM who were previously at risk of HIV have adopted PrEP use, consistent with other studies [2,5,6,13]. We also note the rising proportions of ‘at risk’ participants in our sample who identified as bisexual or who were born overseas. This aligns with national trends that show that recent gains in HIV prevention have been concentrated among Australian-born gay men [1–3].

Our analysis is limited by the use of self-reported, cross-sectional data, which may be subject to recall, desirability or other biases, and cannot track the practices of individuals over time. Our behavioural surveillance samples are skewed towards men who are socially and sexually involved with gay men in metropolitan areas (who may be at risk of HIV), and are unlikely to be representative of the broader population of GBM [14]. However, these are the largest ongoing surveys of GBM in Australia, and we believe the trends we have identified indicate a broad shift in HIV prevention in this population.

In conclusion, the introduction of PrEP has transformed the prevention practices of Australian GBM, rapidly becoming the most commonly used HIV prevention strategy during casual sex. Together, the use of PrEP, UVL and condoms has led to an increase in HIV prevention coverage and a reduction of risk among non-PrEP-users. These conditions are conducive to further reductions in HIV transmission in Australia. As noted by UNAIDS [15], our method can assess the changing mix of prevention methods in the combination prevention era as well as the overall level of prevention coverage in an HIV-affected population. We encourage other jurisdictions to consider this type of monitoring and reporting as they seek to harness the potential benefits of combination prevention.

Acknowledgements

The authors thank all the gay and bisexual men who participated. Funding for the research was received from state and territory health departments and the Australian Government Department of Health. M.H., T.R.B., L.M., C.C., G.P. and B.R.B. contributed to the study design and data collection. M.H. conceived of the idea for the article and wrote the initial article with input from T.R.B. and B.R.B. T.R.B. conducted the statistical analyses with input from M.H. and B.R.B. All authors read and commented on earlier drafts of the article and agreed with the final version.

Conflicts of interest

There are no conflicts of interest.

References

1. The Kirby Institute. HIV, viral hepatitis and sexually transmissible infections in Australia: annual surveillance report 2018. Sydney: The Kirby Institute, UNSW Sydney; 2018.
2. Grulich AE, Guy R, Amin J, Jin F, Selvey C, Holden J, et al. Population-level effectiveness of rapid, targeted, high-coverage roll-out of HIV preexposure prophylaxis in men who have sex with men: the EPIC-NSW prospective cohort study. Lancet HIV 2018; 5:E629–E637.
3. Gunaratnam P, Heywood AE, McGregor S, Jamil MS, McManus H, Mao L, et al. HIV diagnoses in migrant populations in Australia – a changing epidemiology. PLoS One 2019; 14:e0212268.
4. Holt M, Lee E, Lea T, Bavinton B, Broady T, Mao L, et al. HIV preexposure prophylaxis cascades to assess implementation in Australia: results from repeated, national behavioral surveillance of gay and bisexual men, 2014–2018. J Acquir Immune Defic Syndr 2020; 83:e16–e22.
5. Holt M, Lea T, Bear B, Halliday D, Ellard J, Murphy D, et al. Trends in attitudes to and the use of HIV preexposure prophylaxis by Australian gay and bisexual men, 2011–2017: implications for further implementation from a Diffusion of Innovations perspective. AIDS Behav 2019; 23:1939–1950.
6. Hammoud MA, Vaccher S, Jin F, Bourne A, Maher L, Holt M, et al. HIV preexposure prophylaxis (PrEP) uptake among gay and bisexual men in Australia and factors associated with the nonuse of PrEP among eligible men: results from a prospective cohort study. J Acquir Immune Defic Syndr 2019; 81:e73–e84.
7. Sullivan PS, Sanchez TH, Zlotorzynska M, Chandler CJ, Sineath RC, Kahle E, et al. National trends in HIV preexposure prophylaxis awareness, willingness and use among United States men who have sex with men recruited online, 2013 through 2017. J Int AIDS Soc 2020; 23:e25461.
8. Chen YH, Guigayoma J, McFarland W, Snowden JM, Raymond HF. Increases in preexposure prophylaxis use and decreases in condom use: behavioral patterns among HIV-negative San Francisco men who have sex with men, 2004–2017. AIDS Behav 2019; 23:1841–1845.
9. Holt M, Lea T, Mao L, Kolstee J, Zablotska I, Duck T, et al. Community-level changes in condom use and uptake of HIV preexposure prophylaxis by gay and bisexual men in Melbourne and Sydney, Australia: results of repeated behavioural surveillance in 2013–2017. Lancet HIV 2018; 5:E448–E456.
10. Holt M, Murphy DA. Individual versus community-level risk compensation following preexposure prophylaxis of HIV. Am J Pub Health 2017; 107:1568–1571.
11. Holt M, Lea T, Mao L, Zablotska I, Lee E, Hull P, et al. Adapting behavioural surveillance to antiretroviral-based HIV prevention: reviewing and anticipating trends in the Australian Gay Community Periodic Surveys. Sex Health 2017; 14:72–79.
12. Down I, Ellard J, Bavinton BR, Brown G, Prestage G. In Australia, most HIV infections among gay and bisexual men are attributable to sex with ‘new’ partners. AIDS Behav 2017; 21:2543–2550.
13. Ryan KE, Mak A, Stoove M, Price B, Fairley CK, Ruth S, et al. Protocol for an HIV preexposure prophylaxis (PrEP) population level intervention study in Victoria Australia: the PrEPX study. Front Public Health 2018; 6:151.
14. Grulich AE, de Visser RO, Badcock PB, Smith AMA, Heywood W, Richters J, et al. Homosexual experience and recent homosexual encounters: the 2nd Australian Study of Health and Relationships. Sex Health 2014; 11:439–450.
15. UNAIDS. Seizing the moment: tackling entrenched inequalities to end epidemics (Global AIDS Update 2020). Geneva: UNAIDS; 2020.
Keywords:

Australia; HIV; MSM; preexposure prophylaxis; risk factors; sexual behaviour

Supplemental Digital Content

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.