Objectives: To assess biological and behavioral factors in HIV-uninfected female sex workers (FSWs) accessing postexposure prophylaxis (PEP) and to characterize the circumstances preceding PEP, time to access, and adherence.
Methods: Participants were HIV-uninfected FSWs enrolled in an HIV care and prevention program in Nairobi, Kenya. Those accessing PEP between 2009 and 2010 were enrolled and compared with controls. Multiple logistic regression models were used to compare PEP-related biological and behavioral parameters.
Results: PEP users (n = 326) had been involved in sex work for a shorter duration than nonusers [n = 2570; 3.3 vs. 5.1 yrs, AOR: 0.92, 95% confidence interval (CI): 0.89 to 0.95] and were less likely to report a regular partner (54.9% vs. 72.5%, AOR: 0.52, 0.39 to 0.68). PEP use was associated with gonorrhea infection (6.9% vs. 2.6%, AOR: 2.37, 95% CI: 1.34 to 4.21) and alcohol use (84.3% vs. 76.1%, AOR: 1.58, 95% CI: 1.09 to 2.31), but with increased condom use (85.1% vs. 68.2%, AOR 1.80, 95% CI: 1.38-2.35) and a history of prior HIV testing (89.2 vs. 76.2%, AOR: 2.22, 95% CI: 1.45 to 3.40). Reasons for PEP access centered on issues of client mistrust. The median time from exposure to PEP was 18 hours, with an estimated adherence of 49%. Precise PEP efficacy could not be calculated, but HIV incidence was 0.6% in users (2/326) compared with 2.1% (30/1417) in nonusers (Cox regression P = 0.35).
Conclusions: “PEP services were accessed by 10% of FSWs during the study period and were not implicated in any incident HIV cases.” Users had indicators of increased sexual risk and higher health care literacy. Increasing PEP access and compliance in FSWs may be an important HIV prevention strategy.
*Department of Medical Microbiology, Kenya AIDS Control Project, University of Nairobi, Nairobi, Kenya
†Department of Medicine, University of Toronto, Toronto, Ontario, Canada
‡Department of Medical Microbiology, University of Manitoba, Winnipeg, Manitoba, Canada
§Department of Public Health, Erasmus Medical Center, Rotterdam, Netherlands
‖Community Medicine, United Arab Emirates University, Al Ain, United Arab Emirates
¶National AIDS and STI Control Programme, Ministry of Public Health and Sanitation, Nairobi, Kenya.
Correspondence to: Dr Lyle R. McKinnon, Department of Medical Microbiology, University of Nairobi, PO Box 19676-00202, Nairobi, Kenya (e-mail: email@example.com).
Supported by grants from the US President's Plan for AIDS Relief (PEPFAR) grant no. U62/CCU024510-05 to J.K. and L.G., and by the Canadian Institutes of Health Research (CIHR MOP 89983) to R.K.
Equal contributions were made for the article by P.I. and L.R.M. and also R.K. and J.K. The authors P.I., L.R.M., R.K., and J.K. designed the study; P.I., J.P., S.K., C.W., and J.K. provided and managed clinical care; P.I., L.R.M., J.M., S.K., W.K., and N.N analyzed the data; P.I., L.R.M., J.M., S.K., W.K., J.P., S.K., R.K., N.N., G.G., C.W., N.M., H.M., L.G., R.K., and J.K wrote the article.
The authors have no conflicts of interest to disclose.
Received May 21, 2012
Accepted September 10, 2012
Female sex workers (FSWs) represent a vulnerable group at high risk of HIV infection, and sex work is an important driver of HIV transmission in the broader community, even during a mature epidemic.1 Therefore, interventions that prevent HIV infection in FSWs will not only protect vulnerable women but can also reduce HIV transmission at a population level.2,3 Antiretroviral medications have been shown to reduce HIV transmission in several ways. The provision of effective antiretroviral therapy (ART) to an HIV-infected individual not only improves their own health but also dramatically reduced the chance of subsequent HIV transmission to their sexual partner(s).4,5 In addition, the provision of oral and/or vaginal pre-exposure prophylaxis (PrEP) to an HIV-uninfected individual also reduces the risk of HIV acquisition.6,7
However, the results from recent clinical trials suggest that the efficacy of PrEP and antiretroviral-based microbicides is highest in the context of long-term HIV-serodiscordant couples, but remains less clear in the setting of high-risk individuals.8,9 The FEM-PrEP trial of Truvada in high-risk African women was halted early due to futility,10 as were the oral and vaginal tenofovir arms of the VOICE study.9 Therefore, there may be a “prevention gap” for FSWs and other high-risk women, which has implications for the eventual effectiveness of HIV control efforts.
Despite high rates of condom use, FSWs have frequent exposures that could lead to HIV infection, and postexposure prophylaxis (PEP) represents a potential prevention tool to avert infection in these circumstances. PEP was first used in health care settings after exposure to HIV via needlestick injuries or contact with other infectious body fluids, and a case–control study demonstrated that azidothymidine alone after a needlestick injury resulted in an 81% reduction in transmission risk.11 There is also evidence that PEP can prevent mucosal transmission in nonhuman primates.12,13 Therefore, PEP is commonly recommended after known or possible sexual HIV exposure, particularly after sexual assault, although breakthrough HIV infections can occur.14 The potential for HIV exposure during sex work arises in several contexts, including coerced sex, inability to negotiate safe sex, and condom breakage. It is currently recommended in many countries that PEP be given after high-risk sexual exposures,15 but there are no specific guidelines for PEP use in FSW populations. A barrier to the development of such guidelines include the many questions remain surrounding the access to, compliance with, and efficacy of PEP in the FSWs context.
We have been providing PEP to FSWs as a core component of a comprehensive HIV/sexually transmitted infection (STI) care and prevention package in a large FSWs community outreach program in Nairobi. To better understand the utility and potential efficacy in this context, we have explored the timing and circumstances of PEP use, adherence to the regimen, and subsequent HIV incidence.
We recruited HIV-uninfected FSWs who enrolled in a large outreach program in the central business district of Nairobi, the Sex Worker Outreach Program City Clinic, between August 2008 and August 2010 (n = 2896). This clinic is open from 8 AM–5 PM, Monday to Friday. Recruitment of sex workers into the clinic utilized several strategies, including (1) identification and mapping of “hot spots” where sex work takes place, via key informants and the prevention team, (2) hot spot–driven mobilizations with the assistance of peer leaders, and (3) continual “snowball” recruitment via provision of cards to clinic attendees, which they can pass on to their peers. Hot spots included bars, lodgings, and street-based locations, and mobilization has been continual since the clinic opened. PEP was offered to FSWs starting in August 2009, and from August 1, 2009, to July 31, 2010. PEP was requested by approximately 11% (n = 326) of FSWs (PEP users). FSWs who did not access PEP during the same period (PEP nonusers) were included as controls. Institutional Review Boards at Kenyatta National Hospital (Kenya) and the University of Manitoba (Canada) approved the study.
Clinical Protocol for PEP Provision
As part of health education during demand creation and clinic enrollment, FSWs were provided with information on use of PEP after risky sexual exposure. All women were told about PEP during mobilization and at clinic enrollment, including circumstances where PEP would be appropriate, the time frame in which it should be accessed, and what was included in the PEP package. The PEP provision protocol incorporated risk-reduction counseling, HIV counseling and testing, Combivir (azidothymidine + lamivudine) provision twice daily for 28 days, emergency contraception, and STI prophylaxis (200 mg doxycycline for 7 days, 800 mg norfloxacin stat, and 2 g tinidazole stat). PEP was provided in 2 installments, with 10 days of Combivir provided at initial presentation and the remaining 18 days of doses provided at day 10. At the latter visit, FSWs were assessed clinically for symptoms/signs of acute HIV infection (flu-like illness, rash, lymphadenopathy, fever), asked about medication side effects, and adherence was assessed. Routine clinical blood work (blood count, liver function, renal function, and biochemistry) and a pelvic exam for STI diagnostics (gonorrhea, chlamydia, trichomonas, syphilis, bacterial vaginosis) were performed at the time of PEP request. HIV/STI testing was repeated at 6 weeks and 3 months, where possible. In addition to an enrollment questionnaire administered to all cohort participants, a specific questionnaire was administered to each PEP user, capturing data on demographics, behavior, and specific information about PEP.
We compared baseline (cohort enrollment) characteristics between PEP users and PEP nonusers using χ2 and Mann–Whitney U tests for univariate analyses. We incorporated any variable at P < 0.1 in univariate analysis into logistic regression models to control for possible confounding variables (Table 1). The final model included only variables associated with PEP, as shown in Table 1. PEP users were compared in a similar manner with respect to poor vs. good adherence and early vs. late PEP access. Qualitative data regarding the nature of the exposure was compiled and presented with similar answers collapsed into single variables. Survival analysis was carried out on those with prospective follow-up (56% of the initial sample). Cox proportional hazard regression was used, with PEP as a time-dependent covariate. A second model added as possible confounders additional variables associated with HIV acquisition in this cohort (described in Results).
Characteristics of FSWs Who Accessed PEP
During the study period, 326 HIV-uninfected FSWs requested PEP (PEP users) and were compared with 2570 HIV-uninfected FSWs who attended the clinic during the same period and did not request PEP (PEP nonusers). A small proportion of participants requested PEP twice (n = 35) and thrice (n = 2) during the study period. No HIV seroconversions were observed in temporal association with PEP use in this study (ie, HIV acquisition within 12 months after a PEP request). Annual HIV incidence in the cohort over a similar time frame was approximately 2.2% [95% confidence interval (CI): 1.6% to 3.1%, McKinnon et al, unpublished data, 2012].
We analyzed a number of variables in univariate analyses, including age, religion, marital status, education level, reproductive and medical history, duration of sex work, and current sex work behavior (condom use, client volume and type). In multivariate logistic regression, several variables were associated with PEP use (Table 1). In univariate analysis, PEP users were younger (median 27 vs. 29 years, P = 0.003) and had been involved in sex work for a shorter duration; only shorter duration of sex work remained significant in the multivariate model (mean 3.3 vs. 5.1 years, P < 0.0001, AOR: 0.92, 95% CI: 0.89 to 0.95). Age was therefore removed from the final model. In univariate analyses, PEP users reported a higher number of casual clients the day before cohort enrollment (P < 0.0001) and more frequent sex with regular partners in the past week (P = 0.005). In multivariate analysis, there was no association with the number of casual clients during the preceding day (mean 0.92 vs. 0.68, P = 0.11, AOR 1.06), but PEP users were less likely to have a regular partner (55% vs. 73%, P < 0.0001, AOR 0.52, 95% CI: 0.39 to 0.68). PEP users were more likely to report condom use with casual clients 100% of the time (85 vs. 68%, P < 0.0001, AOR: 1.80, 1.38 to 2.35) and also more likely to have been HIV tested before cohort enrollment (89% vs. 76%, P < 0.0001, AOR: 2.22, 95% CI: 1.45 to 3.40). Gonorrhea rates were higher at enrollment in PEP users compared with controls (6.9% vs. 2.6%, p=0.003, AOR 2.37, 95% CI: 1.34 to 4.21), and PEP users were more likely to report alcohol use (84 vs. 76%, P = 0.017, AOR: 1.58, 95% CI: 1.09 to 2.31). Finally, PEP users were more likely to report a prior therapeutic abortion (mean 0.74 vs. 0.62, P = 0.004, AOR: 1.21, 95% CI: 1.06 to 1.37). Together these data demonstrate several differences between PEP users and controls.
Circumstances Associated With Seeking PEP
We next examined characteristics of the specific exposure that prompted the PEP request, including the location, type of client, and type of sex (Table 2). The majority of encounters leading to requests for PEP were with first-time casual clients (69.2%), whereas repeat casual clients represented 24% of PEP encounters. Most sex occurred in “lodgings” (73.3%) and was vaginal (98.8%). “A lodging is a place that people rent out for a relatively short time (generally less than a day or night), often for the purpose of having paid sex.” This is similar to the remainder of the cohort, where 96.3% and 80.4% reported never having anal and oral sex with casual clients, respectively. Similar frequencies were found with regular partners. The ratio of casual to regular clients associated with PEP use was also similar to the overall cohort, where FSWs have more casual than regular clients in a typical week (mean 11.8 casual vs. 2.2 regular clients, respectively). We do not routinely collect data on location of sex, so it is unclear whether sex in lodgings is more likely to lead to PEP.
We asked the reason for seeking PEP as (1) a categorical question, with options given (Table 2), and (2) as an open-ended question (Fig. 1). Reasons for seeking PEP varied widely. Most women (84.7%) reported that a condom breakage was the reason for seeking PEP, “possibly because this reason is suggested at clinic enrollment when describing when PEP should be used.” However, when asked “what specifically about this exposure” led to seeking PEP, the most common reasons revolved around trust issues (19.7%) and not knowing the client's HIV status (15.2%). This was also reflected by a number of participants (14.4%) who reported that the client was not worried after a condom burst. Also relatively common were abusive sexual behavior including intentional condom breakage (6.8%) and rape (4.5%). A number of other responses were given less frequently. These data show that PEP use was associated with condom ruptures, but also a number of other reasons surrounding the exposure, including (but not limited to) characteristics of the client involved.
Timing of PEP Access
The timing of PEP initiation after exposure is believed to be critical, with the efficacy of PEP presumed to decline over time. Most guidelines recommend it not be offered after 72 hours, except in pediatric cases.15 Furthermore, it is believed that PEP is most effective if given within the first 24 hours after exposure. In this cohort, we found a median time of 18 hours from exposure to initiation, but also that 25% of FSWs initiated >36 hours after exposure. Timing of PEP access is clearly an issue in this setting, particularly because many exposures occur at night or during the weekend. We compared participants who accessed PEP on Monday (n = 90) to other days of the week (n = 236). PEP access on Monday was associated with a longer time since exposure (median 29.5 vs. 16 hours, P < 0.001).
We next determined correlates of late presentation (>24 hours after perceived exposure, because this corresponds to decreased efficacy in nonhuman primate models12). Late presenters were older (30.5 vs. 28 years, P = 0.007), had more prior induced abortions (0.92 vs. 0.61, P = 0.048), had prior self-reported history of genital ulcer disease (38.9% vs. 27.3%, P = 0.040), and differed in their use of contraception (Table 3). No other behavioral or demographic variables, at baseline or the time of PEP request, were associated with late PEP presentation. These data suggest late presenters may differ from those who present within the optimal 24-hour window after possible HIV exposure.
Adherence to PEP
To increase adherence and reinforce prevention messages, we provided PEP in 2 doses; a 10-day dose of PEP was provided at presentation and a second 18-day dose at the day 10 visit. Therefore, we have measured adherence in 2 ways as follows: (1) the proportion of PEP users who presented at day 10 to receive their remaining dose and reported adherence during the first 10 days of PEP; and (2) the proportion of PEP users who returned within 2 months for follow-up evaluation and HIV testing. In total, 56% of PEP users returned for their day 10 visit; 88% of these participants reported complete adherence (159/181). For those with incomplete adherence at day 10, participants did not take pills on a median of 2 of 10 days (interquartile range: 2–5). Approximately one-quarter (24%) of all PEP users returned for their 6-week follow-up HIV test. We compared those with poor adherence at day 10 (did not attend or attended but reported <80% adherence) vs. those who reported good adherence at day 10 and received the remainder of their 28-day course. The only association of PEP adherence was age. FSWs adherent to PEP (for at least 10 days) were older with a median of 29 years compared with 26 years for nonadherent participants (P = 0.001, not shown).
PEP Use and HIV Seroconversion
To determine if PEP was a risk factor for HIV over a longer period, we performed Cox proportional hazard analyses with PEP access as a time-dependent co-variate for HIV acquisition. These analyses incorporated 1639 FSWs, including 222 PEP users (68%) and a 1417 HIV-uninfected non-PEP controls (54%); this is the subset of participants who had the prospective data required for survival analysis. Follow-up duration among PEP users was longer compared with nonusers (mean 385 vs. 330 days, P < 0.001). A total of 32 HIV infections were observed during follow-up. Two PEP users seroconverted, in each case >1 year after PEP had been accessed, whereas 30 HIV seroconversions were observed in non-PEP users (P = 0.56, not shown). In the control group, HIV acquisition occurred on an average of 366 days after cohort enrollment. To control for known confounders, other covariates that have been associated with HIV acquisition in this cohort, including gonorrhea infection, a shorter duration of sex work, sex during menses, and hormonal contraception use (McKinnon et al, unpublished data) were added to the model. Condom use and client numbers were not associated with HIV risk in this cohort, and their addition did not affect PEP associations (not shown). PEP use was also not associated with HIV acquisition in this model (P = 0.35, not shown).
FSWs are at very high risk for HIV infection and remain an important contributor to population-level HIV transmission even in a generalized epidemic.1 Behavioral interventions, such as peer-led counseling to increase condom use and a decrease in the number of sexual partners, can reduce HIV risk.16,17 However, it is not yet clear whether emerging biomedical HIV prevention technologies will be effective in this important most at-risk population.8 The efficacy of technologies such as PrEP, microbicides, and vaccines in HIV prevention has been optimal in the context of stable serodiscordant couples and has been variable and incomplete at best as individual participant risk increases.18–20 Clearly, additional prevention tools to prevent HIV transmission to and from FSWs are urgently needed and could have a major impact on the HIV epidemic.21
PEP after a high-risk sexual HIV exposure is an established prevention technology, but data regarding efficacy and acceptability in any context—and particularly the FSW context—are very sparse. We found that PEP was relatively well accepted by this population, with >10% of FSW participants requesting PEP at least once during the year after its introduction. Although many unmeasured confounders could complicate an accurate determination of PEP efficacy in an observational study, no incident HIV infections were observed during the year after PEP access. PEP was not associated with reduced HIV acquisition in these analyses, including a multivariable model incorporating known or possible confounders. Although the observational nature of our study precludes definitive conclusion on PEP and HIV risk, it remains possible that PEP use was a marker of higher perceived risk and an increase in other unmeasured protective behaviors. This hypothesis is supported by higher rates of abortion and HIV testing before cohort enrollment in PEP users, suggesting that that PEP users are more medically literate and better able to access the local health care system. Previous studies of men that linked PEP for sexual exposure and subsequent HIV seroconversion also found no increased risk.22,23 It should be noted that our study was not powered to detect a difference in HIV incidence; longer follow-up of PEP users may provide important data in this regard.
In addition to these possible increases in perceived HIV risk, there was evidence that PEP users have characteristics associated with a higher actual HIV risk. PEP users were less experienced in sex work, had a higher prevalence of gonorrhea, reported more casual (compared to regular) partners, and reported higher levels of alcohol use. Some of these variables, such as gonorrhea and shorter duration of SW, have been associated with increased HIV risk in this (McKinnon et al, unpublished data) or other Nairobi FSW cohorts.24,25
Previous work has shown that regular clients are associated with decreased condom use; it may also be that regular clients are associated with a decreased risk perception by the FSWs. When asked what in particular about an exposure led to the perception of HIV exposure, many responses revolved around mistrust of a casual “first-off” client. This was linked to something about the client (ie, occupation), in others, it was the client's behavior (ie, not worried after a condom burst), and in other cases, it was not clear why the mistrust existed. Despite these findings, there remains a need for future studies of what leads FSWs to request PEP. This might represent an opportunity to better “target” PEP to highest risk situations, particularly because the greatest HIV risk may exist within the context of a “trusted” regular sexual partner (Chakkalackal et al, unpublished data, May 15, 2012).
Our results show that PEP compliance was incomplete among FSWs, with approximately half of participants reporting optimal PEP usage at 10 days, and that PEP access was not always timely. Prior studies have found comparably low PEP adherence (20%–50%) in many settings, including highly informed health care workers and victims of sexual assault,26–28 whereas other studies have found higher adherence (up to 95%).29 A majority of PEP users in previous studies report side effects, in some cases relatively serious,30,31 and factors like stigma and psychological trauma, particularly in sexual assault victims, can also decrease PEP adherence.32 Although these factors did not seem to impact HIV acquisition, this remains a significant concern. FSW are a mobile population and may require additional counseling to counter diminished risk perception over time and to improve PEP adherence. Although PEP was generally accessed in a timely fashion, at a median of 18 hours after perceived exposure, more than one-third of FSW presented more than 24 hours after exposure. Both clinic and client factors were involved, with reduced clinic access over the weekend being a clear association of the former. These data suggest that PEP availability could be expanded beyond the normal clinic working hours to include weekend access. Timing of access was similar to another study of PEP for sex in non–sex workers in Kenya, which found access in 19 hours; although acceptable, this is for obvious reasons much slower than access to PEP after occupational exposure (3 hours).33
Although novel and timely, our study does have limitations that should be acknowledged. Although all FSWs were informed regarding the availability of PEP at the time of clinic enrollment, it remains possible some women accessed PEP elsewhere, and these participants would have been included in our study as controls. However, the large number of control participants in our study means that this is unlikely to substantially bias the results. It would be ideal to have gathered additional qualitative data regarding the circumstances that led FSWs to seek or not seek PEP. In this initial study, an open-ended question was posed, and many different reasons were given. Future studies should gather more extensive data about why PEP was accessed in certain situations and (perhaps even more importantly) why it was not accessed by controls in other high-risk situations. “Barriers to PEP access and knowledge of the denominator of women who might have benefited from PEP are important areas that require further study.” Such data will provide an important entry point for further risk-reduction counseling, in particular, among FSW who are relatively new to sex work. Finally, our ability to determine the causal efficacy of PEP in preventing HIV is limited in this observational study due to unmeasured bias and relatively low HIV incidence. Nonetheless, it is important that we observed very few HIV infections in PEP users despite their high-risk profile compared with PEP nonusers.
This represents the first large-scale study of PEP access and acceptability in the context of FSWs. Based on our sample size and overall clinic recruitment approach, we believe that these data are generalizable to FSWs in East Africa. PEP was accessed by 10% of the overall FSW cohort, with a number of indicators suggesting enhanced sexual risk and seemed to be effective despite suboptimal timing and adherence. The time of PEP access represents an important opportunity to provide enhanced risk-reduction counseling to a subset of FSWs at particularly high risk of HIV acquisition. PEP represents a useful tool within the HIV prevention package for FSWs, and further research should delineate ways to enhance access and efficacy. There remains an urgent need to develop guidelines for PEP in the setting of FSWs in sub-Saharan Africa.
The authors would like to acknowledge all those who participated in the study, and the staff at Sex Worker Outreach Program City Clinic and Kenya AIDS Control Project, Universities of Nairobi/Manitoba, for their support.
1. Chen L, Jha P, Stirling B, et al.. Sexual risk factors for HIV infection in early and advanced HIV epidemics in sub-Saharan Africa: systematic overview of 68 epidemiological studies. PLoS One. 2007;2:e1001.
2. Plummer FA, Nagelkerke NJ, Moses S, et al.. The importance of core groups in the epidemiology and control of HIV-1 infection. AIDS. 1991;5(suppl 1):S169–S176.
3. Jha P, Nagelkerke JD, Ngugi EN, et al.. Public health. Reducing HIV transmission in developing countries. Science. 2001;292:224–225.
4. Cohen MS, Chen YQ, McCauley M, et al.. Prevention of HIV-1 infection with early antiretroviral therapy. N Engl J Med. 2011;365:493–505.
5. Donnell D, Baeten JM, Kiarie J, et al.. Heterosexual HIV-1 transmission after initiation of antiretroviral therapy: a prospective cohort analysis. Lancet. 2010;375:2092–2098.
6. Karim QA, Karim SS, Frohlich JA, et al.. Effectiveness and safety of tenofovir gel, an antiretroviral microbicide, for the prevention of HIV infection in women. Science. 2011;329:1168–1174.
7. 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.
8. Cohen MS, Muessig KE, Smith MK, et al.. Antiviral agents and HIV prevention: controversies, conflicts and consensus. AIDS. 2012;26:1585–1598.
9. van der Straten A, Van Damme L, Haberer JE, et al.. Unraveling the divergent results of pre-exposure prophylaxis trials for HIV prevention. AIDS. 2012;26:F13–F19.
10. Karim SS, Kashuba AD, Werner L, et al.. Drug concentrations after topical and oral antiretroviral pre-exposure prophylaxis: implications for HIV prevention in women. Lancet. 2011;378:279–281.
11. Cardo DM, Culver DH, Ciesielski CA, et al.. A case-control study of HIV seroconversion in health care workers after percutaneous exposure. Centers for Disease Control and Prevention Needlestick Surveillance Group. N Engl J Med. 1997;337:1485–1490.
12. Tsai CC, Emau P, Follis KE, et al.. Effectiveness of postinoculation (R)-9-(2-phosphonylmethoxypropyl) adenine treatment for prevention of persistent simian immunodeficiency virus SIVmne infection depends critically on timing of initiation and duration of treatment. J Virol. 1998;72:4265–4273.
13. Otten RA, Smith DK, Adams DR, et al.. Efficacy of postexposure prophylaxis after intravaginal exposure of pig-tailed macaques to a human-derived retrovirus (human immunodeficiency virus type 2). J Virol. 2000;74:9771–9775.
14. Fisher M, Benn P, Evans B, et al.. UK Guideline for the use of post-exposure prophylaxis for HIV following sexual exposure. Int J STD AIDS. 2006;17:81–92.
15. Benn P, Fisher M, Kulasegaram R. UK guideline for the use of post-exposure prophylaxis for HIV following sexual exposure. Int J STD AIDS. 2011;22:695–708.
16. Gregson S, Garnett GP, Nyamukapa CA, et al.. HIV decline associated with behavior change in eastern Zimbabwe. Science. 2006;311:664–666.
17. Ngugi EN, Wilson D, Sebstad J, et al.. Focused peer-mediated educational programs among female sex workers to reduce sexually transmitted disease and human immunodeficiency virus transmission in Kenya and Zimbabwe. J Infect Dis. 1996;174(suppl 2):S240–S247.
18. Rerks-Ngarm S, Pitisuttithum P, Nitayaphan S, et al.. Vaccination with ALVAC and AIDSVAX to prevent HIV-1 infection in Thailand. N Engl J Med. 2009;361:2209–2220.
19. Baeten JM. What can the twisted tale of PrEP results teach us? Presented at: 19th Conference on Retroviruses and Opportunistic Infections; March 5–8, 2012; Seattle, WA.
20. Van Damme L, Corneli A., Ahmed K, et al., FEM-PrEP Study Group. The FEM-PrEP trial of emtricitabine/tenofovir disoproxil fumarate (truvada) among African Women. Presented at: 19th Conference on Retroviruses and Opportunistic Infectons; March 5–8, 2012; Seattle, WA.
21. Ministry of Public Health and Sanitation K. National guidelines for HIV/STI services for sex workers. Available at: nascop.or.ke/library/Marps/Sex%20Worker%20Guidelines.pdf. Accessed May 15, 2012.
22. Pierce AB, Yohannes K, Guy R, et al.. HIV seroconversions among male non-occupational post-exposure prophylaxis service users: a data linkage study. Sex Health. 2011;8:179–183.
23. Donnell D, Mimiaga MJ, Mayer K, et al.. Use of non-occupational post-exposure prophylaxis does not lead to an increase in high risk sex behaviors in men who have sex with men participating in the EXPLORE trial. AIDS Behav. 2010;14:1182–1189.
24. Fowke KR, Nagelkerke NJ, Kimani J, et al.. Resistance to HIV-1 infection among persistently seronegative prostitutes in Nairobi, Kenya. Lancet. 1996;348:1347–1351.
25. Yadav G, Saskin R, Ngugi E, et al.. Associations of sexual risk taking among Kenyan female sex workers after enrollment in an HIV-1 prevention trial. J Acquir Immune Defic Syndr. 2005;38:329–334.
26. Chacko L, Ford N, Sbaiti M, et al.. Adherence to HIV post-exposure prophylaxis in victims of sexual assault: a systematic review and meta-analysis. Sex Transm Infect. 2012;88:335–341.
27. Loutfy MR, Macdonald S, Myhr T, et al.. Prospective cohort study of HIV post-exposure prophylaxis for sexual assault survivors. Antivir Ther. 2008;13:87–95.
28. Bentz L, Enel P, Dunais B, et al.. Evaluating counseling outcome on adherence to prophylaxis and follow-up after sexual HIV-risk exposure: a randomized controlled trial. AIDS Care. 2010;22:1509–1516.
29. Tissot F, Erard V, Dang T, et al.. Nonoccupational HIV post-exposure prophylaxis: a 10-year retrospective analysis. HIV Med. 2010;11:584–592.
30. Armishaw J, Hoy JF, Watson KM, et al.. Non-occupational post-exposure prophylaxis in Victoria, Australia: responding to high rates of re-presentation and low rates of follow-up. Int J STD AIDS. 2011;22:714–718.
31. Day S, Mears A, Bond K, et al.. Post-exposure HIV prophylaxis following sexual exposure: a retrospective audit against recent draft BASHH guidance. Sex Transm Infect. 2006;82:236–237.
32. Abrahams N, Jewkes R. Barriers to post exposure prophylaxis (PEP) completion after rape: a South African qualitative study. Cult Health Sex. 2010;12:471–484.
33. Siika AM, Nyandiko WM, Mwangi A, et al.. The structure and outcomes of a HIV postexposure prophylaxis program in a high HIV prevalence setup in western Kenya. J Acquir Immune Defic Syndr. 2009;51:47–53.