Objectives Peruvian men who have sex with men (MSM) and transgender women (TGW) are highly vulnerable to HIV infection (HIV), but stigma, access issues and fear of venipuncture hamper testing. The oral HIV test—which uses oral fluids and provides results in 20 minutes—could reduce these barriers. The objective of this study was to determine the acceptability of the oral HIV test and the individual-level factors associated with its acceptability among MSM and TGW.
Methods We conducted a cross-sectional secondary analysis among Peruvian MSM and TGW attending a community-based health centre between February 2012 and February 2013 to determine the individual-level factors associated with oral HIV test acceptability.
Results Of 334 participants, 88% were MSM and 12% TGW. Overall, 85% of participants indicated their acceptability of the oral HIV test. Acceptability was higher in MSM than TGW (85.7% vs 80.0%) but this difference was not significant. Factors associated with acceptability in MSM were: tertiary or higher education (prevalence ratio (PR)=1.18, 95% CI 1.06 to 1.32 and PR=1.16, 95% CI 1.03 to 1.30, respectively); sex with drug use (PR=1.19, 95% CI 1.05 to 1.36); believing that HIV is transmitted by saliva (PR=1.20, 95% CI 1.08 to 1.33); and potential use of the oral test at home (PR=1.56, 95% CI 1.32 to 1.85). The only factor associated with lower acceptability was having had first anal intercourse between 14 and 19 years of age (PR=0.89, 95% CI 0.80 to 0.98).
Conclusions We identified the individual factors associated with oral HIV test acceptability among Peruvian MSM and TGW. Expanded use of the oral HIV test to increase testing rates among Peruvian MSM and TGW is recommended.
Trial registration number NCT01387412, post-results.
- men who have sex with men
- primary prevention
Statistics from Altmetric.com
In 2014, United Nations Programme on HIV/AIDS (UNAIDS) announced the ‘90-90-90’ treatment targets to be reached by 2020, aiming to identify 90% of all people living with HIV, of which 90% would receive sustained antiretroviral therapy (ARV) and of those, 90% would reach viral suppression.1 While progress towards these goals has been made,2 3 a significant bottleneck lies in achieving the first ‘90’: getting 90% of all people living with HIV to know their HIV status so that ARV treatment can be offered.4 5 Globally, only about half of the estimated 36.7 million people living with HIV are aware of their serostatus.6 Delayed HIV diagnosis not only is associated with excess mortality7 and increased medical treatment expenditures compared with those who are diagnosed earlier,8 but persons unaware of their HIV infection also disproportionally transmit the virus to uninfected persons.9
In Peru, progress towards the ‘90-90-90’ goals is illustrative of the problem. The Peruvian HIV epidemic is concentrated in men who have sex with men (MSM) and transgender women (TGW), with prevalences estimated at 12.4% and 20.8%, respectively,10compared with 0.40% in the general population.11 Yet in 2013, only 6% of MSM were tested for HIV during the previous year and knew their HIV status.10 For TGW, the proportion that know their HIV serostatus is unknown because studies in Peru have often historically combined data specific to TGW under the ‘MSM’ label.10 A recent scoping literature review found that among Peruvian MSM and TGW living with HIV, only 24% knew their infection status, of which 13.6% were receiving ARV and 12% had achieved viral suppression.12 The authors point out that the largest ‘drop-off’ occurred at the beginning: identifying those who are infected with HIV. Strategies to address HIV testing barriers are therefore crucially needed.
There are multiple barriers to HIV testing, including access issues, HIV/AIDS stigma and, for blood tests, venipuncture, considered invasive by many people.13 14 Point-of-care testing using an oral HIV test—which uses oral fluids collected from the gumline and provides results in 20 minutes—could address many of these barriers and boost testing rates to promote primary prevention. First, the test can be self-administered in private (eg, at home or another private space), thereby increasing both test access and reducing real or perceived stigma encountered in public testing settings. Also, the oral HIV test is non-invasive (needleless), which can allay fears related to a needle stick for blood-based tests.
High acceptability of the oral HIV test has been documented in several countries. In a 2015 review of data from the USA, Canada, China, Malawi and Kenya, oral HIV self-testing had an acceptability of 81%–91%; ‘acceptability’ was measured as the confidence in the results and future use by study participants.15 However, much less is known about the acceptability of the oral HIV test in Latin America, and research in the region has generally focused more on home-based testing than the type of HIV test used. An internet study among Brazilian MSM found that 47% preferred home-based testing compared with clinic based testing; however, the study did not examine views on the oral test itself.16 In another study examining the acceptability of home-based HIV testing among Brazilian and Peruvian MSM, acceptability of the approach was again high (97%) but oral HIV testing was not one of the options assessed.17 A third study, conducted in Peru among 147 MSM and 45 TGW, focused solely on the hypothetical acceptability of self-testing using the oral HIV test (if free of charge) and reported a high (82%) acceptability.18 The oral HIV test price was the primary barrier to acceptability. In all of these studies, user characteristics (eg, socio-behavioural aspects, knowledge and views on the oral testing method itself and so on) were not assessed as related to oral test acceptability. Understanding how individual-level behaviours and characteristics influence oral HIV testing acceptability could inform strategies to augment its use.
The oral HIV test was licensed in Peru in October 2012, but its use has been limited to physicians only and not for self-use (consumer level), and it has not been widely implemented as part of the Peruvian National HIV strategy. Instead, its use is rare and many still consider HIV testing using oral fluids ‘new’ in Peru.
We conducted a secondary analysis of data collected from a study investigating the role of genital warts on HIV acquisition among a convenience sample of 600 HIV-uninfected MSM (84.0% of the original sample) and TGW (16% of the original sample) in Lima, Peru.19 20 Between February 2012 and February 2013, study participants were recruited at Epicentro (a community-based non-governmental organisation for MSM and TGW that provides HIV/STI testing, medical care and community activities), as well as at bars, clubs, volleyball courts, via social media and by snowball sampling. Eligibility criteria for participation were: born anatomically male; ≥18 years old; reported any anal intercourse with a man during the previous 12 months; residing in metropolitan Lima; willing to commit to twice-yearly clinic visits for 24 months; and had not previously participated in an HIV or human papillomavirus vaccine study. At enrolment, potential participants were screened for HIV using the Determine HIV-1/2 Ag/Ab Combo rapid test. HIV-positive participants were referred to the national HIV programme for free care and treatment, whereas HIV-negative participants were invited to participate in the study. Enrolled participants were first screened for syphilis using the Determine TP test (ALERE Medical Co).20 At the first follow-up visit (month 6 post enrolment) participants completed a 20 minute, computer-assisted self-interview which had 80 questions, of which 10 were related to the hypothetical use of an oral HIV test.
Acceptability of using an oral HIV test was defined as an affirmative answer to the question: ‘Would you use the oral HIV test for your usual future check-ups?’ Answer options were ‘yes’, ‘no’ or ‘no answer’. Individual characteristics assessed for MSM (gay, bisexual, straight, other) and TGW were demographic variables (including age, education, place of birth and relationship status); behavioural factors (including sexual role, sex of partner, age of first anal intercourse, condom use, transactional sex, smoking, sex and alcohol or drugs in the last month, and STI symptoms); and, HIV testing knowledge and attitudes (including confidence in results of blood tests, knowledge of the oral test, transmission of HIV by saliva and potential use of the oral HIV test at home). Sex with drugs or alcohol was measured by two questions, ‘Within the last month, have you had sex under the influence of alcohol?’ and ‘Within the last month, have you had sex under the influence of any kind of drugs?’ Response choices were: ‘yes’, ‘no’, ‘never drink alcohol’ or ‘never use drugs’ and ‘no answer’. The presence of STI symptoms was measured by the question, ‘Do you currently have: pus or discharge from your penis?; burning while urinating?; genital sores?; warts on your penis?; perianal warts or warts on your testicles?; pus or discharge from your anus?; blood in your anus?; anal pain or burning?; anal ulcers or sores?’. Response choices were: ‘yes’, ‘no’ and ‘don’t know/no answer’.
Participants with missing HIV testing question data were excluded from our model. Univariate analysis was performed to describe variable frequencies. Education was defined as primary/secondary, tertiary (no degree) and university (degree). The responses regarding confidence in the blood test for HIV were originally on an ordinal scale of 1–10 points and then categorised as low confidence (1–4 points), medium confidence (5–7 points) and high confidence (8–10 points). We conducted a bivariate analysis of oral HIV testing acceptability in dichotomous form. This analysis was performed separately for MSM and TGW using a χ2 test, Mann-Whitney or Fisher exact test according to the fulfilment of the assumptions for each test. Due to the small TGW sample size, multivariate analysis was performed only on MSM using Poisson distribution in a generalised linear model with a robust log-variance link to estimate the prevalence ratio (PR) after the binomial model failed to converge. Variables with a significance level of ≤0.05 in the crude analysis were used in the multivariate analysis in addition to the confounding variable age. One model was proposed that was adjusted for the variables: age; education level; age at first anal intercourse; past month sex and drugs; belief in HIV transmission by saliva; confidence in blood rapid HIV test; and potential use of oral test at home. All analyses were performed using the statistical software Stata V.14.2.
All participants provided written informed consent prior to study participation.
Of the 600 participants enrolled, 18% (105/600) had a positive result in the syphilis screening, indicative of current or past syphilis infection. A total of 509 participants completed their first follow-up visit, of which 175 participants were excluded because of missing data due to non-response (82 for the outcome variable, 30 for self-reported current STI symptoms, 29 for belief in HIV transmission by saliva and 34 for other variables) leaving 334 participants for our analysis. There were no statistically significant differences in survey responses between the population surveyed (n=509) and the population analysed (n=334) after performing a logistic regression using as the outcome completed/not completed data for all variables of table 1 (data not shown). Overall acceptability of the oral HIV test was 85% (284/334). Eighty-eight per cent (294/334) of participants were MSM while 12% (40/334) were TGW. The mean participant age of MSM and TGW was 26.6 years (range 18–46) and 25.5 years (range 19–33), respectively. Sixty per cent (176/294) of MSM had an education level of tertiary/university compared with 25% (10/40) of TGW. Fifty-four per cent (159/294)of MSM were born in Lima compared with 10% (4/40) of TGW (see table 1).
In the bivariate analysis (table 2), acceptability of the oral HIV test was significantly associated with higher education (p=0.020) only for MSM; having a current relationship with a man (p=0.037) only for TGW; higher confidence in a HIV blood test (p=0.003) only for MSM; and potential home use of the oral HIV test (p<0.001) for both MSM and TGW. Age and associated variables were used in the multivariate analysis for MSM, only.
We performed a multivariate analysis only among MSM using a generalised linear model (GLM) including age and significant variables in the bivariate analysis. The crude model included all variables of interest and showed significantly higher acceptability of the oral HIV test among participants with a higher education; those who had sex with drugs in the last month, those who believed that HIV was transmitted through saliva; and, those who would potentially use an oral HIV test at home (p<0.05). The only variable that was significantly associated with lower acceptability was first anal intercourse between 14 and 19 years old (p<0.05). See table 3.
All models were estimated with GLM family Poisson link log using robust variance due to binomial models failing to converge.
Our model was adjusted for age; education level; age at first anal intercourse; past month sex and drugs; belief in HIV transmission by saliva; confidence in rapid HIV blood test; and those reporting that they would use an oral test at home. We found high acceptability congruent with the crude analysis among those with the highest education levels (p<0.05); those who had sex under the influence of drugs in the last month (p=0.008); in those who believed that HIV is transmitted by saliva (p=0.001); and in participants who said they would use the oral HIV test at home (p<0.001). Those who had their first anal intercourse between 14 and 19 years old maintained their lower acceptability (p=0.021).
The oral HIV test was highly acceptable by most participants in our study. Factors most highly associated with oral HIV test acceptability were: having a higher education level; having sex under the influence of drugs in the past month; the belief that HIV is transmitted by saliva; and potentially using the oral test at home. Participants reporting first anal intercourse between 14 and 19 years of age had a lower acceptability of the oral HIV test.
Our findings corroborate those reported in similar populations on oral HIV test acceptability18 21–23 but extend those findings by providing information on the user characteristics associated with acceptability. For example, despite the near zero risk of HIV transmission via saliva,24 we found that the belief that HIV can be transmitted via saliva was common among our study participants and positively associated with oral HIV test acceptability. Although we did not ask participants in our study about the reasons for this belief, it is possible that the belief that saliva can transmit HIV was related to concerns about oral sex as a risk for HIV infection. For example, a study among MSM in the USA found considerable worry among participants regarding HIV transmission via oral sex.25 Though erroneous, the belief that saliva can transmit HIV may, paradoxically, bolster the acceptability of the oral HIV test that technically uses oral fluids gathered from the gumline which are slightly different than saliva. Clear messaging and education will be essential to explain how the rapid oral HIV test works so that misinformation on HIV transmission via saliva is not inadvertently perpetuated.
Sexual intercourse while using drugs is significantly associated with behaviours that increase risk of HIV infection among MSM.26 27 While our study participants reported a very low frequency of sex under the influence of drugs in the last month, sex with drug use was nonetheless associated with oral HIV test acceptability. Similarly, a recent study in China on the determinants of HIV self-testing found a high association between HIV self-testing and lifetime drug use among MSM.28 Another study of MSM in the USA using injectable drugs during sex found a greater predisposition to non-invasive HIV testing using oral fluids or urine which was explained by their already constant use of needles.29 The drugs reported by participants in our study were not injectable; however, it is possible that among MSM who use drugs during sex, whether injectable or not, a perception of heightened HIV risk may promote the desire for any HIV screening, including oral HIV testing.
We also found a greater acceptability for the oral HIV test among participants with a non-university or college education compared with those with a primary/secondary education, congruent with other studies with similar populations.23 30 However, this result contrasts with that found in a study of 5025 US MSM internet users where home testing acceptability, but not specifically oral HIV testing, was associated with a lower education.22 This finding warrants further research to disentangle the drivers of HIV test acceptability in general, including: rapid versus non-rapid testing; blood versus oral fluids and self-testing versus clinic-based testing. Whenever possible, acceptability should be assessed in the context of user characteristics as we have attempted to do in the present study.
The potential use of oral test at home was also associated with a high acceptability. In a similar study in India, willingness to use an HIV self-test in the future was associated with HIV self-testing performed currently.28 With our result we hypothesise that MSM who agree to future check-ups with an oral HIV test will prefer to perform a self-test at home.
Finally, we found that MSM who had their first anal intercourse between 14 and 19 years of age had a lower acceptability of an oral HIV test. This result contrasts with a study on the willingness of MSM to use oral fluid HIV tests which found that higher willingness among MSM reporting first insertive sexual intercourse with men <25 years of age.30 However, another study found no association between HIV self-testing age of debut of anal intercourse with other men.28
There were several limitations to our study. First, the HIV testing questions were posed in a study that used convenience sampling to enrol participants into a larger study that included regular HIV testing. As such, study participants may have been more motivated to receive HIV tests or try new technologies than MSM and TGW outside of a research context. Next, the sample size for this analysis was smaller than the number of people who completed the first follow-up visit due to missing data, leading to potential selection bias. However, we analysed complete versus incomplete data for all covariables in our models (oral HIV test acceptability, education, HIV transmission in saliva, and past month sex and drug use), finding no significant differences. Thus, we believe that participants in our analysis were distributed similarly to all surveyed participants; nonetheless, the study findings may not be generalised to all MSM or TGW outside of the study population. A further limitation, common in acceptability research, is that we surveyed participants who had not actually used the oral HIV test, thereby only measuring the predisposition to use the oral HIV test in future HIV screenings.
We identified the individual factors associated with oral HIV test acceptability among Peruvian MSM and TGW and recommend the expanded use of the oral HIV test to increase testing rates among Peruvian MSM and TGW. More broadly, these data can inform the national HIV plans in Peru and similar settings in Latin America as they work towards the goal of getting 90% of all people living with HIV to know their HIV status so that ARV treatment can be offered.
Oral HIV testing was highly acceptable among Peruvian populations extremely vulnerable to HIV.
Factors associated with oral HIV test acceptability were: believing that saliva transmits HIV, sex under the influence of drugs and higher education level.
To reach the 90-90-90 United Nations Programme on HIV/AIDS goals, multiple HIV testing modalities including oral testing should be widely deployed.
The authors are grateful to the men and women who participated in this study as well as the staff at the community centre Epicentro in Lima, Peru where the study was conducted. Special thanks to Emiliano Valle for his help with translation of the original manuscript text from Spanish to English.
Handling editor Stefan Baral
Contributors BJB and JTG are the principal investigators who conceived the study design, supervised the study and provided critical revision of this manuscript. JAF wrote the first draft, and performed the statistical analyses. SRL provided revision of the manuscript. HS helped coordinate study at the community-based health centre ‘Epicentro’. All authors contributed to interpretation of the results and the revision of the manuscript, and all approved the final draft.
Funding Supported in part, by the research grant 39619 from the Merck Investigator–Initiated Studies Program of Merck & Co, Inc. Additional funding by National Institutes of Health Grant T32 MH080634. Juan A. Flores is a doctoral student studying Epidemiological Research at Universidad Peruana Cayetano Heredia under FONDECYT/CIENCIACTIVA scholarship EF033-235-2015.
Disclaimer The opinions expressed in this study are those of the authors and do not necessarily represent those of Merck & Co, Inc.
Competing interests None declared.
Patient consent Obtained.
Ethics approval Institutional Review Boards at the University of California, Los Angeles, USA (approval number: 11-003060) and the Asociación Civil Impacta Salud y Educación in Lima, Peru (approval number: 00183-2011) reviewed and approved all study procedures and documents prior to implementation and all participants provided written informed consent prior to study participation.
Provenance and peer review Not commissioned; externally peer reviewed.
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.