In the decades following the first recognition of AIDS cases in the United States, primary prevention of sexually acquired HIV infection has centered on supporting condom use through education about sexual practices associated with HIV acquisition, behavioral interventions, and distribution programs for both male and female condoms. These are all designed to increase consistent and correct condom use by sexually active men and women. Until recently, “unprotected” sex has been used as a synonym for sex without condoms.
However, beginning in 2011, trials conducted with gay, bisexual, and other men who have sex with men (MSM),1 heterosexually active men and women,2,3 and with persons who inject drugs4 have proven that use of daily oral antiretroviral medications for preexposure prophylaxis (PrEP) is highly effective as primary prevention with HIV-uninfected persons engaging in anal or vaginal sex. Preexposure prophylaxis, with coformulated tenofovir disoproxil fumarate 300 mg and emtricitabine 200 mg (marketed in the United States as Truvada), is currently recommended as a prevention option for persons at substantial risk of acquiring HIV infection through sex (or drug injection), especially for persons who do not consistently use condoms.5
Method adherence is critical to achieving high prevention effectiveness for both condoms (i.e., correct use during each sexual encounter) and PrEP (i.e., taking daily medication for months to years). Among MSM, when measured by self-report, high levels of method adherence are associated with similar effectiveness for male condom use during anal sex (70%)6 and PrEP (73%)1 in preventing HIV acquisition.
Following Food and Drug Administration approval of a PrEP indication for Truvada and release of Public Health Service clinical practice guidelines recommending PrEP as one prevention option,5 concerns have been expressed that the increasing use of PrEP will displace condom use. This is usually framed as “risk compensation,” defined as increasing one or more risk behaviors because of perceived reduced susceptibility to an undesired outcome that results from the use of a preventive behavior or intervention.7 However, decreases in condom use were not reported in PrEP trials, including after trial completion.4,8,9 In the context of PrEP, concern about risk compensation arises from a conviction that condoms remain the “best” option, against which other methods are suboptimal because they are not 100% effective.10,11 It is important to acknowledge that all currently available effective HIV prevention methods are partially protective (including consistent condom use) and require high method adherence, and some noncondom options are as effective as or more effective than condoms. This suggests that achieving high rates of protection for persons at substantial risk of HIV acquisition will depend on accurate risk assessment, selection of one or more effective prevention methods that are acceptable and usable in specific situations, and achieving high rates of method adherence.
For more than a decade, the estimated number of new HIV infections in the United States has remained stable at approximately 50,000 each year.12 However, the population rates of new infections and the number of new diagnoses each year vary by subpopulation. The subpopulation with the highest estimated HIV incidence and rising rates of new HIV diagnoses are black/African American MSM,13,14 especially adolescents and young adults.11,15 At the same time, rates of condom use have been falling among MSM16 and youth,17 even before PrEP uptake was beginning.
Although improving access to effective HIV prevention methods is important for all persons at substantial risk of acquiring HIV infection, it is especially critical for the highest-risk populations. In this analysis, we use a static model to examine the impact of method adherence and rates of PrEP and male condom use, independently and in combination, on the number of HIV infections anticipated in hypothetical cohorts of African American MSM.
MATERIALS AND METHODS
Our deterministic model parameters were estimated as follows.
The annual HIV incidence estimate was obtained from a study of MSM attending sexually transmitted disease (STD) clinics in New York City matched against the citywide HIV/AIDS registry.18 We estimated HIV incidence, its standard error (SE), and 95% confidence interval (CI) among African American MSM using published data from a cohort of 276 HIV-negative MSM with rectal gonorrhea or chlamydia followed for 464.7 person-years (PY) with 31 incident HIV diagnoses documented, and a matched sample of 276 HIV-negative MSM without rectal infections followed for 473.8 PY with 12 incident HIV diagnoses.18 This study reported HIV incidence estimates for 55 non-Hispanic blacks with and 55 without (tested negative) rectal chlamydia or gonorrhea infection. For those without rectal infection, 77.20 PY were observed with 5 new HIV diagnoses, resulting in an estimated incidence of 6.48 per 100 PY (95% CI, 2.37–14.36). For those with rectal infection, 78.21 PY were observed with 12 new HIV diagnoses, resulting in an estimated incidence of 15.34 per 100 PY (95% CI, 8.31–26.08). We combined the estimates of those with and without a rectal infection, resulting in 17 new HIV diagnoses observed for 155.41 PY. Based on a Poisson model with a robust SE, the estimated HIV incidence among non-Hispanic black MSM is 10.94 per 100 PY (95% CI, 6.24–19.18).
PrEP Effectiveness by Self-Reported Method Adherence
Estimates of PrEP effectiveness were obtained from the iPrEx international clinical trial of PrEP use among MSM.1 In the modified intention-to-treat analysis, there was 44% effectiveness (95% CI, 15–63) in reducing the risk of HIV infection among men randomized to daily oral PrEP with tenofovir disoproxil fumarate/emtricitabine. In analyses of PrEP effectiveness by a combined measure that included self-reported medication adherence, pills dispensed, and pills returned over scheduled visits, iPrEx investigators reported that effectiveness was 32% (95% CI, −41 to 67) among men with adherence less than 50% (which we have used to estimate effectiveness for “sometimes” use) and was 73% (95% CI, 41–88) among men with adherence of 90% or greater (which we have used to estimate effectiveness for “always” use).
We estimated the parameter and SE for PrEP effectiveness using the following technique. The estimated PrEP hazard ratio
where we divide efficacy (E) by 100 because it is on the percent scale. To obtain the 95% CI on the natural log (ln) scale, we use the same technique using the upper and lower limits of the 95% CI. Once we had the 95% CI on the ln scale, we took the difference between the upper and lower bounds of the 95% CI on the ln scale and divided by 2 to obtain an estimate of the SE on the ln scale and consequently an estimate of the variance (Supplemental Table 1, http://links.lww.com/OLQ/A98).
Condom Effectiveness by Self-Reported Method Adherence
Condom effectiveness estimates were obtained from an analysis of data on self-reported condom use during anal sex among MSM reporting at least one HIV-positive partner in 2 prospective HIV prevention trials.19 In that study, “always” condom use (100%) compared with “never” use was associated with a 70.5% reduction in HIV infection (95% CI, 58–79), and “sometimes” use (1%–99%) compared with “never” use was associated with a nonsignificant 8% reduction in HIV infection (95% CI, −25 to 32). The corresponding model HR parameter estimates on the ln scale and SE of the parameter estimated were available and used in our calculations.
The incidence (I) for the ith condom effectiveness class (always, sometimes, never) and jth PrEP effectiveness class (always, sometimes, never) is estimated as
, where RR is the risk ratio or hazard ratio for condom use and PrEP use with “never” as the reference category and I is the estimated incidence in the absence of PrEP or condom use. Our estimation for the incidence ith, jth class assumes that condom and PrEP use RRs are independent, and each RR is independent of the baseline I. To estimate the variance, we take the natural log (ln) of the estimated incidence,
which results in
. It follows, assuming independence of condom and PrEP use, that the variance is estimated as
The 95% CI of the estimated incidence is estimated assuming normality on the log-scale. Once we have our estimated incidence matrix, we can multiply these estimates by our desired population to estimate the annual incidence and corresponding 95% CI (Supplemental Table 2, http://links.lww.com/OLQ/A98).
Estimating the Number of HIV Infections Expected by Method Adherence
We first estimated the number of annual HIV infections expected without use of any prevention method (i.e., no PrEP or condom use) by 10,000 African American MSM with an annual incidence of 10.94%. We then applied the expected efficacy within each combination of PrEP and condom use class to determine the number of infections that would be prevented among 10,000 African American MSM. The result is expressed as the number of new HIV infections that would occur in each scenario.
We estimated prevention effectiveness (Table 1), annual incidence (Table 2), and the number of HIV infections (Table 3) that would occur for a hypothetical sample of 10,000 African American MSM who were followed up for 1 year in each of 9 mutually exclusive categories, each with a specific rate of PrEP and condom use method adherence. As reported in Table 3, among MSM who never use PrEP, there would be an estimated 323 (95% CI, 167–625) infections among men reporting always using condoms for receptive or insertive anal sex. Furthermore, among MSM who do not use PrEP, higher numbers of infections would occur among those who sometimes use condoms (1007; 95% CI, 531–1907) and among those who never use condoms (1094; 95% CI, 680–1760). Among MSM who never use condoms, there would be an estimated 744 (95% CI, 297–1863) infections among those who take their PrEP medication with less than 50% adherence to daily dosing—a 32% reduction, and 295 (95% CI, 111–783) infections among those always taking daily PrEP medication—a 73% reduction.
Among MSM who take PrEP medication with less than 50% adherence and sometimes use condoms during anal sex, the estimated number of HIV infections is 684 (95% CI, 260–1801). Consistent daily use of PrEP and sometimes condom use resulted in fewer HIV infections, 272 (95% CI–98, 754). Similarly, consistent condom use and less than 50% PrEP adherence resulted in 220 HIV infections (95% CI, 82–586). Finally, consistent PrEP adherence combined with consistent condom use resulted in 87 new HIV infections (95% CI, 31–245)—a 92% reduction, the highest level of protection.
Currently, most MSM are not using PrEP.20,21 As indicated in this analysis, for MSM never using PrEP, the only change in HIV prevention method use that is likely to result in an increased number of HIV infections is movement from not using PrEP while always using condoms (323 infections; 95% CI, 167–625) to using PrEP inconsistently in combination with inconsistent (684 infections; 95% CI, 260–1801)—a 112% increase—or never (744 infections; 95% CI, 297–1863) condom use—a 130 % increase.
A similar pattern of results is obtained when comparing HIV prevention effectiveness (Table 1) and annual HIV incidence (Table 2). It is important to note that the estimated HIV prevention effectiveness of consistent PrEP use and consistent condom use is 92% (95% CI, 81%–96.7%), a rate exceeding the individual effectiveness of consistent PrEP use alone (73%; 95% CI, 41%–88%) and consistent condom use alone (70.5%; 95% CI, 58.2%–79.2%).
In this analysis of a hypothetical sample of 10,000 African American MSM and based on published parameters, MSM with inconsistent or no condom use, currently most MSM,6,22–24 can increase their HIV protection by adding PrEP at either modest or high levels of adherence. Men who have sex with men who continue consistent condom use can further increase their protection from HIV infection by any PrEP use. The only situation likely to lead to an increased risk of HIV acquisition involves movement from consistent condom use for all acts of anal sex to becoming inconsistent users of both condoms and PrEP.
Decreases in condom use, commonly referred to as “risk compensation,” were not observed in PrEP trials, including after conclusion of the trials.4,8,9 However, in these trials, participants did not know if they were taking PrEP medication or placebo, or whether PrEP medication would prove to be effective in reducing the risk of HIV acquisition. Neither were decreases in condom use (or increases in syphilis incidence) observed during PrEP use in the open-label extension of the iPrEx trial.25 In this trial, participants were informed about the effectiveness of PrEP shown in prior trials and its relationship to adherence before deciding whether or not to use PrEP. In addition, all trial participants (including the open-label extension trial) received frequent, high-quality risk reduction counseling that is uncommon outside HIV prevention trials. It is currently unknown whether, or in which direction, condom use may change among persons receiving PrEP from their health care providers as part of routine HIV prevention clinical care. Consistent with PrEP implementation guidelines, persons using PrEP will be informed that high levels of HIV prevention effectiveness are possible with high adherence to daily dosing and will receive frequent HIV and sexually transmitted infections testing that may reinforce the value of condom use in addition to PrEP.
Risk compensation resulting from condom promotion has been demonstrated in a single randomized trial where persons in communities randomized to condom promotion increased both condom use and the number of sexual partners compared with those in control communities.21 Some studies suggest that condom risk compensation is resulting from a belief in the effectiveness of highly active antiretroviral therapy (HAART) in reducing risk of transmission from a virally suppressed HIV-infected sex partner.26–28
Currently, there are 2 antiretroviral-based HIV prevention methods proven highly effective in randomized controlled trials: consistent use of daily oral PrEP medication and consistent HAART use by persons diagnosed as having HIV infection to lower the risk of transmitting to uninfected partners.29 In addition, consistent condom use has been demonstrated to be a highly effective HIV prevention method in multiple prospective observational studies of heterosexual HIV-discordant couples,30, in an analysis of 5 observational studies of condom effectiveness among MSM in developing countries,31 and in an analysis of data from 2 HIV prevention trials conducted with MSM.19 As demonstrated in the analyses presented in this article, consistent use of both PrEP medication and condoms during anal sex results in the lowest number of anticipated annual HIV infections and the highest level of HIV prevention effectiveness (92%). Additional high levels of HIV prevention effectiveness are observed among inconsistent PrEP users with adherence less than 50% who consistently use condoms (79.9%) and among consistent PrEP users who sometimes or never use condoms (75.2% and 73.0%, respectively).
As a consequence of these findings, the way people think about HIV risk management might become akin to how contraceptives options are considered. For example, women are commonly encouraged to use the method with the highest proven effectiveness for pregnancy prevention that is acceptable and safe for them at a given period of time, most commonly daily oral hormonal contraception.32 In addition, they are informed of the importance of consistent condom use to prevent acquisition of STDs and the need to use condoms for pregnancy prevention when several doses of oral contraceptives have been missed.33 Similarly, among PrEP users, consistent condom use should be encouraged to protect against bacterial STDs (e.g., gonorrhea, chlamydia, and syphilis) that can increase susceptibility to HIV infection, and condoms should be especially encouraged when doses of PrEP medication have been missed.
In HIV prevention, we now have an opportunity to reevaluate how to discuss the use or nonuse of HIV risk management methods either individually or in combination. Sexual activity by persons without HIV infection may be condom protected, PrEP protected, HAART protected (if their HIV-infected partner is virally suppressed), dually, or multiply protected. The use of “unprotected sex” should now be reserved for sexual events where neither partner takes any action known to effectively reduce the risk of HIV acquisition or transmission.
There are several limitations of these findings. First, parameter estimates for condom use and daily PrEP adherence used in the models are based on self-reported behaviors in the original clinical trials. The benefits of using self-reported adherence data include comparable assessments for both prevention methods, and ease of assessment by clinicians, other prevention providers, and members of at-risk populations. Another limitation involves deriving models for only one population, African American MSM. This subpopulation was selected based on having the highest estimated HIV incidence, rising rates of new HIV diagnoses, and a known HIV incidence estimate. However, the pattern of HIV prevention effectiveness reported here would be the same for other populations with either higher or lower HIV incidence because the calculations involved are dependent on the relative effectiveness of condom use and PrEP. We anticipate that these modeling estimates will be further evaluated using other statistical methods, and also in the course of public health research and clinical practice health services research.
Concerns about possible reductions in condom use mitigating the preventive effectiveness of PrEP do not justify withholding PrEP from MSM at substantial risk of HIV acquisition because of their current inconsistent or non-use of condoms during sex. This may be a special concern for African American MSM, a population with high HIV incidence who stand to benefit from PrEP use but who may be prejudged as at higher risk for diminished condom use because of risk compensation than white MSM.34
As additional effective methods become available (e.g., vaginal and rectal antiretroviral microbicides), understanding the effects of using multiple prevention methods on HIV risk management will require new ways of balancing overall prevention effectiveness, dependent on method adherence for the components, and their use in various combinations. This analysis offers a useful approach for calculating HIV prevention effectiveness based on combinations of prevention methods and self-reported method adherence.
1. Grant RM, Lama JR, Anderson PL, et al. Preexposure chemoprophylaxis for HIV prevention in men who have sex with men. N Engl J Med 2010; 363: 2587–2599.
2. Baeten JM, Donnell D, Ndase P, et al. Antiretroviral prophylaxis for HIV prevention in heterosexual men and women. N Engl J Med 2012; 367: 399–410.
3. Thigpen MC, Kebaabetswe PM, Paxton LA, et al. Antiretroviral preexposure prophylaxis for heterosexual hiv transmission in Botswana. N Engl J Med 2012; 367: 423–434.
4. Choopanya K, Martin M, Suntharasamai P, et al. Antiretroviral prophylaxis for HIV infection among people who inject drugs in Bangkok, Thailand: A randomized, double-blind, placebo-controlled trial. Lancet 2013; 381: 2083–2090.
5. Centers for Disease Control and Prevention; U.S. Public Health Service. Preexposure prophylaxis for the prevention of HIV infection in the United State—2014: A clinical practice guideline 2014: 1–67. Available at: http://www.cdc.gov/hiv/pdf/guidelines/PrEPguidelines2014.pdf
. Accessed May 28, 2014.
6. Smith DK, Herbst JH, Zhang X, et al. Condom effectiveness for HIV prevention by consistency of use among men who have sex with men (MSM) in the U.S.. J Acquir Immune Defic Syndr 2014. In press.
7. Hogben M, Liddon N. Disinhibition and risk compensation: scope, definitions, and perspective. Sex Transm Dis 2008; 35: 1009–1010.
8. Mugwanya K, Donnell D, Celum C, et al. Sexual behaviour of heterosexual men and women receiving antiretroviral pre-exposure prophylaxis for HIV prevention: Post-unblinding analysis of the partners PrEP study. Sex Transm Infect 2013; 89 (suppl 1): A46.
9. Marcus JL, Glidden DV, Mayer KH, et al. No evidence of sexual risk compensation in the iPrEx trial of daily oral hiv preexposure prophylaxis. PloS One 2013; 8: e81997.
10. Leibowitz AA, Parker KB, Rotheram-Borus MJ. A US policy perspective on oral preexposure prophylaxis for HIV. Am J Public Health 2011; 101: 982–985.
11. Golub S. Tensions between the epidemiology and psychology of HIV risk: Implications for pre-exposure prophylaxis. AIDS Behav 2014.
12. Prejean J, Song R, Hernandez A, et al. Estimated HIV incidence in the United States, 2006–2009. PloS One 2011; 6: e17502.
13. Centers for Disease Control and Prevention; US Public Health Service. Preexposure prophylaxis for the prevention of HIV infection in the United States—2014: A clinical practice guideline 2014: 1–67. Available at: http://www.cdc.gov/hiv/pdf/guidelines/PrEPguidelines2014.pdf
. Accessed May 28, 2014.
14. Centers for Disease Control and Prevention. Estimated HIV incidence in the United States, 2007–2010. HIV Surveill Suppl Rep 2012; 17.
15. Centers for Disease Control and Prevention. Increase in newly diagnosed HIV infections among young black men who have sex with men—Milwaukee County, Wisconsin, 1999–2008. MMWR Morb Mortal Wkly Rep 2011; 60: 99–102.
16. Reece M, Herbenick D, Schick V, et al. Condom use rates in a national probability sample of males and females ages 14 to 94 in the United States. J Sex Med 2010; 7: 266–276.
17. Kann L KS, Shanklin SL, et al. Youth Risk Behavior Surveillance—United States, 2013. MMWR Surveill Summ 2014; 63 (SS-4): 1–168.
18. Pathela P, Braunstein SL, Blank S, et al. HIV incidence among men with and those without sexually transmitted rectal infections: Estimates from matching against an HIV case registry. Clin Infect Dis 2013; 57: 1203–1209.
20. Kirby T, Thornber-Dunwell M. Uptake of PrEP for HIV slow among MSM. Lancet 2014; 383: 399–400.
21. Rucinski KB, Mensah NP, Sepkowitz KA, et al. Knowledge and use of pre-exposure prophylaxis among an online sample of young men who have sex with men in New York City. AIDS Behav 2013; 17: 2180–2184.
22. McFarland W, Chen Y-H, Nguyen B, et al. Behavior, intention or chance? A longitudinal study of HIV seroadaptive behaviors, abstinence and condom use. AIDS Behav 2012; 16: 121–131.
23. Halkitis PN, Brockwell S, Siconolfi DE, et al. Sexual behaviors of adolescent emerging and young adult men who have sex with men ages 13–29 in New York City. J Acquir Immune Defic Syndr 2011; 56: 285–291.
24. Peterman TA, Tian LH, Warner L, et al. Condom use in the year following a sexually transmitted disease clinic visit. Int J STD AIDS 2009; 20: 9–13.
25. Grant RM, Anderson PL, McMahan V, et al. Uptake of pre-exposure prophylaxis, sexual practices, and HIV incidence in men and transgender women who have sex with men: A cohort study. Lancet Infect Dis 2014; 14: 820–829
26. Gorbach PM, Weiss RE, Jeffries R, et al. Behaviors of recently HIV-infected men who have sex with men in the year post-diagnosis: effects of drug use and partner types. J Acquir immune Defic Syndr (1999) 2011; 56: 176.
27. Heijman T, Geskus RB, Davidovich U, et al. Less decrease in risk behaviour from pre-HIV to post-HIV seroconversion among MSM in the combination antiretroviral therapy era compared with the pre-combination antiretroviral therapy era. AIDS 2012; 26: 489–495.
28. MacKellar DA, Hou S-I, Whalen CC, et al. HIV/AIDS complacency and HIV infection among young men who have sex with men, and the race-specific influence of underlying HAART beliefs. Sex Transm Dis 2011; 38: 755–763.
29. Smith DK, Pals SL, Herbst JH, et al. Development of a clinical screening index predictive of incident HIV infection among men who have sex with men in the United States. J Acquir Immune Defic Syndr 2012; 60: 421–427.
30. Weller S, Davis K. Condom effectiveness in reducing heterosexual HIV transmission. Cochrane Database Syst Rev 2002; 1: CD003255.
32. Jones J MW, Daniels K. Current contraceptive use in the United States, 2006–2010, and changes in patterns of use since 1995. Natl Health Stat Rep 2012; 60. Available at: http://www.cdc.gov/nchs/data/nhsr/nhsr060.pdf
. Accessed May 30, 2014.
33. Centers for Disease Control and Prevention. U.S. Selected Practice Recommendations for Contraceptive Use, 2013: Adapted from the World Health Organization Selected Practice Recommendations for Contraceptive Use, 2nd Edition. MMWR Recomm Rep 2013; 62 (RR05): 1–46.
34. Calabrese SK, Earnshaw VA, Underhill K, et al. The impact of patient race on clinical decisions related to prescribing hiv pre-exposure prophylaxis (PrEP): Assumptions about sexual risk compensation and implications for access. AIDS Behav 2014: 1–15.