Preexposure prophylaxis (PrEP) has emerged as a powerful new tool in the fight against HIV [1–3]. However, the expanding use of PrEP has been accompanied by increases in other sexually transmitted infections (STIs) previously under control. Between 2013 and 2014, there was a 15% increase in the number of reported primary and secondary syphilis cases nationwide [4]. Among the 2014 cases, 83% were among MSM, an important target group for PrEP [4]. One concern about PrEP is the idea of risk compensation, namely that individuals taking PrEP may increase their risky sex behaviors and thereby increase the spread of STIs other than HIV [5]. Risk compensation has often been ignored on the basis of a few high-profile clinical trials that failed to find evidence for increased risk behaviors [1,2,6–8]. However, new evidence from real-world applications of PrEP suggests that increased risk behavior may indeed be a problem [3,9]. PrEP is definitely still valuable as a tool for lowering the likelihood of HIV transmission, but it must be accompanied by clear messages that vulnerable populations should continue practicing safe sex. Given that PrEP is becoming more widely available, public health advocates need to be even more active in ensuring that condom promotion is not neglected. A landmark study demonstrating the effectiveness of PrEP for reducing HIV incidence was the iniciativa profilaxis pre-exposicion (iPrEX) double-blind clinical trial [1]. One of the major arguments for lack of risk compensation comes from data from this study [6,7]. The researchers collected data on sexual risk behaviors and STIs, including gonorrhea, chlamydia, and syphilis in addition to HIV. The researchers concluded that risk compensation did not occur because there was either no change or a decrease in risk behaviors [7]. However, the data on risk behaviors relied entirely on self-report, which may have been skewed by desire to give the socially acceptable response. The researchers also noted that syphilis incidence did not differ as a function of whether people believed they were in the treatment or control group [7]. Surprisingly though, the researchers did not analyze or discuss incidence for gonorrhea or chlamydia, even though data were apparently available for these STIs in addition to syphilis [1]. The original studies on the impact of PrEP on HIV incidence used the standard practice of double-blind procedures to look at the biological effect on HIV incidence, so participants were not aware of whether they were actually taking PrEP or not [1,6,7]. This research design, however, is not appropriate for drawing conclusions about the psychological impact of PrEP on risk behaviors, as participants may not change their behavior unless they are confident that they are protected from HIV. Other double-blind studies face the same issue, which could explain their failure to find changes in condom usage [8,10,11]. If participants had been continuing to engage in safe sex only because they were not certain that they were actually taking PrEP, then some might have increased their risk behavior if unblinding had taken place partway through the study. In fact, a study in Kenya and Uganda demonstrated exactly this effect [12]. The study collected data on self-reported unprotected sex over the course of a placebo-controlled trial and continued to collect data on this variable for up to 12 months after unblinding. The researchers noted a statistically significant increase in unprotected sex with outside partners after unblinding, thereby providing evidence for increased risk behaviors, though there was no increase in the STIs measured [12]. A recent study that has been cited as evidence for lack of risk compensation is the PROUD study [2]. This was an open-label study in which people either began receiving PrEP immediately or after a deferral period of 1 year. The authors argued that risk compensation did not occur because the immediate and deferred groups did not differ in number of STI cases for syphilis, gonorrhea, or chlamydia [adjusted odds ratio (aOR) = 1.07; 95% confidence interval (CI): 0.78–1.46]. However, the authors had noted that visits to the clinic were higher for the immediate group than the deferred group [2]. They attempted to correct for this in their analysis of the odds ratio (OR). However, if one of the reasons for the higher number of clinic visits was that the people in the immediate group had a higher rate of STIs, then this correction would lead to a type II error. In fact, the unadjusted OR was 1.33 for STI cases in the immediate vs. deferred group. An additional concern with the PROUD study is that the deferred and immediate groups were living in the same community [2]. This may have led the members of the different study conditions to have sex with each other, thereby leading to cross-contamination and distorting the results. The authors also mention but do not discuss the finding that participants in the immediate group were reliably more likely to report condomless receptive anal sex with at least 10 different partners (P = 0.03). This is remarkable as it provides strong evidence that PrEP increases risk behavior regardless of the STI findings. The argument that vulnerable populations will not increase risk behaviors if they gain access to PrEP seems to be inconsistent with a number of focus group studies [13–15]. In one of these studies, 60% of gay and bisexual men living in Los Angeles admitted that they would decrease their condom usage if they gained access to PrEP [13]. In another study, 23% of gay and bisexual men stated that access to PrEP would reduce their condom usage, whereas only 14% reported it would increase their condom usage [14]. It is important to note that the people most likely to say that PrEP would reduce their condom usage were those engaging in the most risky behaviors, making them more vulnerable to STIs. Among people who had had condomless anal sex in the last 6 weeks, 33% said that PrEP would further decrease their condom usage, compared with only 10% among those who had not had condomless anal sex [14]. Another study also found that people were much more likely to see PrEP as making condoms unnecessary if they had not used a condom the last time they had sex (aOR = 7.45; 95% CI: 1.57–35.45) [15]. Interestingly, the people who said that they would reduce their condom usage were more likely to be white, college-educated, and have higher income, possibly because these people felt they could ‘afford’ to practice unsafe sex [14,15]. In addition to clinical trials and focus groups, it is worth looking at the broader impact of PrEP promotion on public health as a whole. In the last 2 years, studies on the large-scale use of PrEP in ‘real-world’ communities have begun to come out. The results are encouraging regarding the prevention of HIV transmission, but some of the secondary results provide cause for concern regarding the trends for some other STIs beside HIV. In a study of 657 individuals in San Francisco (99% MSM) who had recently begun taking PrEP through Kaiser Permanente, 41% said that they had decreased their condom usage, whereas only 3% said they had increased it [9]. The authors of this study also noted an unusually high rate of STIs (specifically chlamydia, gonorrhea, and syphilis), with 30% reporting at least one STI over the course of a 6-month follow-up [9]. A second study of 437 PrEP users (mostly MSM) analyzed numbers of new STI cases over time following PrEP initiation [3]. The researchers noted that over the first 24 weeks after PrEP initiation there was a decrease in STIs, but that between 24 and 48 weeks there was a sharp increase in STIs (P < 0.05). This suggests that people on PrEP may initially start out practicing safer sex due to greater accessibility to condoms through the PrEP program. However, over time they may become more careless and return to a pattern of unsafe sex. Patterns of unprotected sex among MSM in general have been rising in the United States in recent years at alarming rates. Looking at national survey data from the National HIV Behavioral Surveillance (NHBS) System, conducted by CDC, the percentage of HIV-negative MSM engaging in condomless anal sex increased from 34.7 to 40.5% between 2011 and 2014 [16]. The year 2011 is important to be considered as a baseline, as this was just prior to the approval of PrEP by the food and drug administration [17]. According to the NHBS data of San Francisco, which has one of the highest percentages of MSM PrEP users in the country [18], the percentage of MSM reporting consistent condom use dropped from 30.5% in 2011 to just 18.3% in 2014 [19]. A recent study of young MSM found that PrEP users had twice as many anal sex partners with whom they did not use condoms in comparison with non-PrEP users [20]. Surveys of MSM at the Seattle Gay Pride Parade indicate that PrEP usage increased more than six-fold (from 5 to 31%) among high-risk respondents between 2012 and 2015 [21]. Recent evidence from the Seattle area shows that STI rates are increasing sharply. The data on King County STI trends from 2013 to early 2016 are shown in Fig. 1. In this area, the number of early syphilis cases increased from 55 to 100 between the first quarter of 2013 and the first quarter of 2016 [22]. The number of cases of gonorrhea increased 42% (from 296 to 419), and cases of chlamydia increased 65% (from 292 to 483) [22]. If the widespread acceptance of PrEP contributes to a drop-off in condom usage, we would expect that these sorts of cases will become more common in MSM in the future.Fig. 1: Trends of sexually transmitted infections and preexposure prophylaxis among MSM, King County, Washington State, USA.We should emphasize that none of this evidence is intended to prove that PrEP leads to risk disinhibition and increased STIs, as almost all clinical trials have failed to detect an increase in rates of STIs following PrEP introduction [2,3,7,8,12]. Furthermore, rates of syphilis were gradually increasing nationwide at least as early as 2005, well before PrEP had been introduced [23]. However, recent years have shown an acceleration in rates of primary and secondary syphilis cases, with MSM showing the largest increases [23]. We believe that there is enough evidence that introduction of PrEP contributes to risk compensation in real-world settings, which public health advocates should treat as a legitimate concern. In conclusion, there is no question that PrEP is a valuable new tool in the fight against HIV. However, if PrEP comes to be erroneously interpreted as a substitute for safe sex practices, then it may lead to careless condom use and the spread of other STIs that were previously under control. An important new idea in public health is the idea of quaternary prevention, which suggests that physicians should avoid giving treatments that indirectly lead to other public health problems [24]. If PrEP does indeed lead to risk compensation for other STIs, then the quaternary prevention perspective suggests that physicians should not prescribe it without including adequate safe sex education alongside it, just as was done prior to PrEP's introduction or even more given the national trend of dropping condom use among target populations. A critical issue that needs to be addressed in future research is the long-term effects of PrEP uptake in real-world settings, as well as other key populations in addition to MSM. Most of the evidence for lack of risk compensation comes from placebo-controlled clinical trials [1,6–8,10,11], but it is not known if the same results will apply once PrEP is integrated into actual community healthcare systems over time. Also, few studies have used follow-up periods longer than 1 year for the impact of PrEP uptake. The study by Liu et al.[3] found that at the end of the follow-up period at 48 weeks, STIs were increasing sharply rather than decreasing. This suggests that PrEP could increase unsafe sex practices when used over time horizons longer than what is typically captured in clinical trials. Therefore, public health officials should continue to gather evidence on how the adoption of PrEP operates in the real-world so that the health of the general public is not inadvertently threatened. Acknowledgements K.A. suggested the topic, reviewed the relevant studies, and guided interpretation of the findings. C.A.P. searched the literature and wrote the main body of the manuscript. S.-C.J. and A.A. reviewed the manuscript and provided valuable feedback. Disclosure: This work was not supported by any grant money. Conflicts of interest There are no conflicts of interest.
No takes yet. Share an insight, caveat, or question.
Alaei et al. (2016) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: