Article Text

Download PDFPDF

Awareness of post-exposure HIV prophylaxis in high-risk men who have sex with men in New York City
  1. Sapna A Mehta1,
  2. Richard Silvera2,
  3. Kyle Bernstein3,
  4. Robert S Holzman1,
  5. Judith A Aberg1,4,
  6. Demetre C Daskalakis1,4
  1. 1Division of Infectious Diseases, Department of Medicine, New York University School of Medicine, New York, New York, USA
  2. 2Department of Environmental Medicine, New York University School of Medicine, New York, New York, USA
  3. 3San Francisco Department of Public Health, STD Prevention and Control Service, San Francisco, California, USA
  4. 4Bellevue Hospital Center, New York, New York, USA
  1. Correspondence to Dr Sapna A Mehta, Greenberg Hall, SC1-174, 545 First Avenue, New York, NY 10016, USA; sapna.mehta{at}


Objectives To understand the factors associated with knowledge of non-occupational post-exposure prophylaxis (nPEP) and pre-exposure prophylaxis (PrEP), bathhouse patrons in New York City (NYC) were surveyed.

Methods 554 men who have sex with men (MSM) at two NYC bathhouses were given a standardised survey focused on nPEP and PrEP at the time of HIV testing.

Results In the previous 90 days, 63% of respondents reported unprotected sex with a male partner and 7% reported any sex with a known HIV-positive male partner. Less than half reported having a primary provider (primary care practitioner) who was aware of their MSM behaviour. 201 men (36%) were aware of nPEP or PrEP. In univariate analyses, race/ethnicity, previous HIV testing, gay self-identification, higher education level, having a primary provider aware of MSM behaviour, reported interaction with the healthcare system, use of the internet for meeting sex partners, reporting unprotected sex in the previous 90 days, reporting any sex with an HIV-positive male partner in the previous 90 days and having a higher number of sex partners were each significantly associated with being aware of nPEP or PrEP. In multivariate analysis, having a higher number of sex partners was significantly associated (OR 5.10, p=0.02) with post-exposure prophylaxis (PEP)/PrEP knowledge and disclosure to a primary care provider was also associated, although less robustly (OR 2.10, p=0.06).

Conclusions Knowledge of nPEP or PrEP among sexually active MSM in NYC is low and is associated with having a primary provider aware of their patient's same-sex behaviours. These findings show the need for improving education about nPEP among high-risk MSM in NYC and the role of providers in these efforts.

  • AIDS
  • bathhouse
  • commercial sex venue
  • epidemiology
  • HIV prevention
  • MSM
  • post-exposure prophylaxis
  • pre-exposure prophylaxis
  • prevention
View Full Text

Statistics from

Antiretroviral agents have been used to prevent HIV infection in both occupational and non-occupational exposures for many years. Since 2005, the Centers for Disease Control and Prevention have published specific guidelines for non-occupational post-exposure prophylaxis (nPEP) in the setting of significant sexual or injection drug-associated exposures.1

Consideration has also been given to antiretroviral agents as pre-exposure prophylaxis (PrEP) in select populations at enhanced risk of infection.2–4 Despite the lack of data and approval, off-label use of antiretroviral agents for PrEP in the USA has been observed but remains to be fully characterised.5

Studies to date have attempted to survey men who have sex with men (MSM) to obtain a better understanding of the awareness and use of both nPEP and PrEP. A recent study by Liu et al4 reported higher rates of awareness for both nPEP (47%) and PrEP (16%) among men in California than previously reported. Factors associated with awareness included older age, being more affluent and higher-risk sexual behaviour. Kalichman et al6 reported an association with intention to use nPEP by MSM in Atlanta with younger age, less education, polysubstance use and higher-risk sex behaviour. Two previous multicity survey studies have interviewed minority gay pride event attendees regarding their awareness and use of nPEP and PrEP.2 3 In the first study, Kellerman et al2 found that 25% of men were aware of PrEP and that awareness was associated with having been tested for HIV in the previous 12 months and with reporting meeting sex partners at sex clubs. The following year, Voetsch et al3 found that 21.4% of men were aware of post-exposure prophylaxis (PEP) or PrEP and was associated with being HIV positive and with having a high number of sexual partners in the previous year.

Data to guide more effective risk-reduction interventions are limited on factors associated with both the knowledge and use of nPEP and PrEP among MSM populations. We aimed to study factors associated with nPEP and PrEP awareness among a highly sexually active population of MSM at risk of HIV, who elected to be tested for HIV at two commercial sex venues (bathhouses) in New York City (NYC) in a programme described elsewhere.7


Study design and subjects

Subjects were recruited between February 2006 and December 2007 on-site at two bathhouses in NYC by study staff to complete a structured questionnaire and receive rapid HIV testing or testing for a sexually transmitted infection. Recruitment was passive through signage throughout the bathhouse and overhead announcements. Bathhouse clients were not directly approached to participate in testing. On any single night, given limits of space and staff, four to nine people were tested during a shift. Only a small fraction of the estimated 50–200 men at the venue on an average night participated in testing. Of those that accepted testing, 12% declined survey completion. Informed consent was obtained utilising an institutional review board-approved research consent as well as New York State consent for HIV testing. New York State mandated HIV pre-counselling and post-counselling was also provided. No financial or material incentives were provided for participation.

The study survey was available in English and in Spanish. Individuals who believed themselves to be HIV negative were tested and surveyed. Repeat participants were identified by matching name and birthdate to previous respondents and were excluded from this analysis. In addition, participants were excluded if they self-identified as a repeat participant during the survey. As this was a descriptive analysis of a pre-existing database, no a priori power calculations were performed. To provide some information on the ability of logistic regression to detect associations in the data we performed a power calculation post hoc using data derived from the cross tabulation of a person's awareness of nPEP/PrEP and their primary provider's awareness of that person's MSM behaviour. We used the program Gpower (v 3.12)8 to compute the power of univariate logistic regression to detect the observed OR of 2.16 using a one-tailed test with α=0.05, and assuming that, as observed, 124 (39.2%) of all 316 patients for whom data were available were aware of nPEP/PrEP. It was additionally assumed that the distribution of provider awareness was binomial with a mean of 0.465 (ie, we observed that 147 of 316 patients' providers were aware). Under these assumptions the power of univariate logistic regression to detect the observed OR was 0.96.


Data regarding gender, race, sexual orientation, education, healthcare utilisation, previous HIV testing and substance use were collected during the risk assessment. Validated survey questions from the EXPLORE study9 were adapted to assess awareness and utilisation of PEP and PrEP. One survey item evaluated knowledge of PEP and PrEP and did not differentiate between these chemoprophylactic approaches. Subjects reported sexual activity in response to a series of questions asking about oral, insertive anal and receptive anal intercourse, condom use and the serostatus of partners in the previous 90 days. Unsafe sex was defined as any unprotected anal intercourse in the previous 90 days. Primary provider refers to all licensed healthcare workers who provide primary care, such as nurse practitioners, physician assistants and physicians.

Data analysis

Data entry, management and analyses were performed using SPSS for Windows version 15.0 and R.10 Some responses were missing because patients did not always answer all the questions. Differences between medians of continuous variables were tested for statistical significance using the Mann–Whitney U test. OR for categorical variables were tested for statistical significance by univariate or multivariate logistic regression. A p value of less than 0.05 was considered significant. For the regression analyses age was grouped in quintiles and the total number of sex partners in the past 90 days was grouped by whether the number of partners was at or below the mode of three, between four and 10, 11 and 20, or over 20. This grouping was selected as providing the most informative OR from among several possible categorisations that were examined. Variables that had p≤0.20 associated with their univariate OR were included in a multivariate analysis.


Six hundred and eighty-eight surveys were completed. Of these 688 surveys, 134 were identified as being from repeat participants and were excluded from the study in an attempt to limit the influence of our previous counselling on survey responses. Five hundred and fifty-four first-time participants seeking HIV testing were included in this analysis. Seventy-eight (12%) of patients tested for HIV or a sexually transmitted infection declined survey participation.

Baseline characteristics

The average age of participants was 39.8 years (IQR 24.8–54.8). Four hundred and twenty-two (77%) participants self-identified as gay, 482 (87%) had a college-level education,11 and 251 (49%) were non-white. Ninety-eight per cent (534/543) reported previous interaction with the healthcare system in one of several venues. The majority (89%, 494/554) reported that they had previously tested for HIV. Less than half (47%, 147/316) of these MSM reported that they had disclosed that they have sex with men to their primary provider. Many of these men, both provider disclosed and undisclosed, reported behaviours that may warrant nPEP; 63% (206/328) reported having unsafe sex in the previous 90 days and 7% (39/538) reported having any sex with an HIV-positive partner in the previous 90 days (see table 1).

Table 1

Baseline characteristics among surveyed MSM in NYC

Knowledge of nPEP or PrEP

Two hundred and one (36%) subjects were aware of either nPEP/PrEP. White men were more likely than non-white men to be aware of nPEP/PrEP (44% vs 30%, respectively). Gay-identified men were also more likely to be aware of nPEP/PrEP. Only 39% of men with a college degree, 36% of those individual who had had any interaction with the healthcare system and 49% of those men whose primary provider was aware of their MSM status were aware of nPEP/PrEP. Among men with higher-risk sex practices, 44% of those who used the internet for meeting partners, 40% of men who reported unsafe sex in the previous 90 days and 51% of those who had sex with an HIV-positive partner in the previous 90 days were aware of nPEP/PrEP (see table 2).

Table 2

Univariate and multivariate analyses of baseline characteristics and factors related to awareness of PrEP or nPEP among MSM in NYC

Recreational substance use was common among surveyed subjects. Thirty-six per cent of men reporting alcohol use were aware of nPEP/PrEP. Less than half (range 20–43%) of patients reporting other illicit drug use were aware of nPEP/PrEP. Neither alcohol use nor the use of any individual drug was associated with awareness of nPEP/PrEP, with non-significant OR ranging from 0.43 to 1.33.

In univariate logistic regression analyses, white race (OR 1.81, p=0.001), previous HIV testing (OR 3.15, p=0.001), gay-self-identification (OR 2.11, p=0.001), higher education level (OR 2.85, p=0.001), having a provider aware of MSM status (OR 2.16, p=0.001), interaction with the healthcare system (OR 4.56, p=0.15) and use of the internet for meeting partners (OR 1.73, p=0.004) were each significantly associated with being aware of nPEP/PrEP. Trends were identified for having sex with an HIV-positive partner (OR 1.92, p=0.052) in the previous 90 days and having a higher number of sex partners (p=0.05) as factors predicting awareness of nPEP/PrEP (see table 2).

In multivariate analysis, a significant association emerged between the number of sex partners and awareness of nPEP/PrEP. OR increased such that those with more than two partners had a 5.1-fold increase in the probability of being aware relative to those at or below the mode of three partners. While the OR associated with having interactions with health care, previous HIV testing and self identification as gay were reduced, suggesting that they contributed information correlated with the number of sex partners, the OR associated with having a provider aware of MSM status (decrease from 2.16 to 2.10), white race (unchanged at 1.81) and college or higher education level (decrease from 2.85 to 2.28), remained relatively unchanged suggesting that these variables contributed information to the model that was largely independent of the number of sex partners.

Use of nPEP or PrEP

Eighteen respondents (3.2%) reported previous use of nPEP. All had previous HIV testing, previous interaction with the healthcare system and a college education. The mean number of partners in the past 90 days for those who reported previous use of nPEP was 17.4 partners compared with 11.7 partners for those who had never used nPEP (p=0.0009). No respondents reported previous use of PrEP.

Testing results

Seventeen (3.1%) of 543 patients tested for HIV infection had a positive rapid test. Two of these 17 individuals were identified as having acute HIV infection by pooled nucleic acid (or viral load) testing after having a negative rapid HIV serological test.


PEP following occupational exposures has been shown to be an effective and safe intervention to prevent new HIV infections.12 These data together with various cohort studies have provided support for the use of PEP following non-occupational sexual exposures. Despite this observed efficacy and the Centers for Disease Control and Prevention guidelines for recommendations on the use of nPEP, awareness and use among high-risk populations such as MSM remains low. Our data show this is true even among highly sexually active MSM at risk of HIV in NYC. Almost all subjects in our analysis had had some contact with the healthcare system, yet awareness remains at 36% and lower than the 47% rate of awareness recently reported in California.4

These results suggest a need for increasing awareness and access to nPEP at various points of healthcare contact for populations at risk such as MSM. The majority of our surveyed participants had previously been tested for HIV and therefore had contact with some form of HIV-related counselling, which ideally should include education about prevention. Yet, only 39% of these men who had been tested for HIV were aware of nPEP or PrEP, highlighting a missed opportunity for targeted counselling. Nonetheless, our univariate analyses showed that having been tested for HIV or having interaction with the healthcare system are associated with a greater likelihood of being aware of nPEP or PrEP. In multivariate analysis, higher number of sex partners remained statistically associated with nPEP/PrEP knowledge. Having a primary provider aware of MSM behaviours appeared to weaken as an association, but remains near significance at a p value of 0.06.

Studies show the greatest cost-effectiveness when nPEP services are targeted to high-risk individuals, such as MSM.13 In the current era of universal testing for HIV, providers should therefore also integrate sexual history-taking for all patients, as understanding patients' sexual practices can allow for appropriately targeted HIV prevention and counselling. It is heartening that men with more partners appear to have more awareness of nPEP/PrEP than those with fewer partners. The observation that men who engaged in unprotected sex were not more likely to know about these interventions highlights the need for detailed history-taking and education to ensure risk and risk-based counselling is appropriately delivered. In addition, the observation that openness about MSM behaviours with a primary provider may also be associated with knowledge of nPEP/PrEP further emphasises the need for a detailed sexual health focus in caring for MSM.

Furthermore, delivery of nPEP services remains suboptimal14 and is probably not reaching those populations most likely to benefit from nPEP. In the context of public resource constraints, primary providers may serve as an important linkage as more effective nPEP delivery models are designed. Health departments can target providers who see large patient populations of MSM for training on nPEP education and referral resources. This may help improve awareness of and access to nPEP for at-risk men and must be balanced with provider clinical decision support for the use of nPEP.

A number of limitations require consideration. First, all data were collected through face-to-face interviews, and not audio computer-assisted self-interview, which has been shown to improve the self-report of stigmatised behaviours.15 Furthermore, we assessed nPEP and PrEP awareness together, and were not able to examine each separately. As PrEP is currently not recommended and is under current evaluation as an HIV prevention intervention, it is possible that most of the knowledge that was assessed in our analysis was regarding nPEP. It would be worthwhile to re-introduce survey items that differentiate nPEP and PrEP given recent publication and lay press coverage of the iPREX study.16 Survey results in that study are solely from MSM who presume themselves to be HIV uninfected at the time of study entry. In addition, our data were collected from a group of fairly affluent and well-educated men seeking HIV testing services in two NYC bathhouses. Given the very specific demographics of these study participants, they probably do not represent all MSM in NYC or in other areas. Further strategies should be developed to reach less well-educated and affluent MSM to evaluate their knowledge and utilisation of HIV medications for primary or secondary prevention.


Knowledge of nPEP or PrEP among sexually active MSM is low in NYC. Awareness of nPEP or PrEP is associated with having a primary provider who is aware of their patient's MSM status. These findings show the need for and the role of providers in improving education about nPEP among high-risk MSM in NYC.

Key messages

  • Knowledge of nPEP or PrEP among MSM is low in NYC.

  • There is a need for greater targeted education about nPEP among high-risk MSM in NYC.

  • Primary providers knowledgeable of their patients' sexual practices may play an important role in improving education about nPEP and other HIV prevention interventions.


The authors would like to thank the management of the two bathhouses for their cooperation. They also thank Drs Joel Ernst, Jeffrey Klausner, Diane Binson and James Woods for their intellectual support.


View Abstract


  • Competing interests None.

  • Patient consent Obtained.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.