Abstract
Background:
Among those at high risk for HIV, it is important to examine the ways in which someone who has recently tested for HIV might differ from someone who has not.
Methods:
In 2017–2018, 5,001 men-, trans women-, and trans men who have sex with men from across the U.S. completed an online survey about their recent testing behavior as well as self-collected oral samples for HIV testing.
Results:
In total, 3.8% tested HIV-positive and—among those with positive results—35% were recent HIV infections (i.e., self-reported an HIV-negative test result within the 12 months prior to enrollment). Those with HIV-positive results—regardless of how recent their HIV test was prior to enrollment—differed from those with negative results in ways that are known to be associated with HIV risk: racial and income disparities, housing instability, recent transactional sex, and recent methamphetamine use. Among those with HIV-positive results at enrollment, only having a primary care physician distinguished those who recently tested negative prior to enrollment versus not. Whereas, among those with HIV-negative results, there were numerous differences between those who had recently tested for HIV prior to enrollment, versus not, such that those who had not recently tested were significantly more likely to report being at higher risk for HIV.
Conclusion:
Strategies aimed at improving more frequent HIV testing among HIV-negative persons at high risk for HIV should address other needs including stable housing, transactional sex, access to a primary care provider, and methamphetamine use.
Keywords: HIV testing, men who have sex with men, gay and bisexual men, HIV results, methamphetamine, pre-exposure prophylaxis
INTRODUCTION
In the U.S., gay, bisexual, and other men who have sex with men (GBM) accounted for 67% of new infections in 2016 (82% of infections among men).1 Between 2005—2014, HIV diagnoses increased by 6% among GBM.2 And in 2016, the Centers for Disease Control and Prevention (CDC) estimated one-in-six U.S. GBM will acquire HIV in their lifetime, including one-in-two Black GBM and one-in-four Latino GBM.2,3
Dubbed a “Status-Neutral Approach to HIV,”4 HIV testing serves as a critical gateway to route persons into tailored prevention or treatment. That is, until you know someone’s HIV status, it is not possible to know whether HIV prevention (e.g., Pre-Exposure Prophylaxis, PrEP) or HIV treatment would be the appropriate course of action. Thus, HIV testing represents a pivotal moment in individualized sexual healthcare.
In order to locate HIV testing where it is needed most, it is important to examine the ways in which someone at risk for HIV who has recently tested for HIV might differ from someone who has not.5 The CDC recommends, minimally, annual HIV tests for GBM, though other guidelines have called for testing with greater frequency.6 Clark et al.7 reported on data from more than 68,000 HIV tests conducted with GBM during the CDC HIV Testing Initiative. Of 68,185 HIV tests, 8% were with GBM who never previously tested. Those having never tested were significantly more likely to be persons of color and younger. Such information informs how best to tailor initiatives that encourage testing.
Over the course of the HIV epidemic, efforts to increase HIV testing have ranged from incorporating HIV testing into routine health care, to mobile street vans that offer HIV testing, and, home-based HIV self-testing. In 2015, the NYC Department of Health launched a free, at-home HIV test kit giveaway via men-for-men geosocial networking apps.8 Over a 23-day period, 2,493 eligible participants were identified, of which 71% redeemed the coupon for an HIV test kit to be mailed to them. This program demonstrated the feasibility of using technology-mediated methods to identify persons at risk for HIV and successfully engage them in at-home self-testing.
There is an ongoing need to examine factors associated with HIV diagnoses, coupled with understanding factors that may differentiate those who have engaged in recent HIV testing versus not. In the present study, we report on findings from a U.S. national cohort of individuals at risk for HIV. Participants were identified via geosocial networking apps via ads marketing free at-home HIV testing. In order to inform HIV treatment, prevention, and testing initiatives, we examined group differences based on whether participants had negative versus positive results, as well as whether and when they had been tested for HIV in the year prior to enrollment.
METHODS
Cohort enrollment
Data were collected as part of the Together 5000 study; a national, internet-based cohort study of men, trans men, and trans women in the U.S..9,10 Enrollment began October 2017 using ads on men-for-men geosocial sexual networking phone applications, and concluded in June 2018. Eligibility criteria for the study are shown in Table 1.
Table 1.
Inclusion criteria for Together 5000 study
| Core eligibility criteria (all participants must meet all of these criteria) | |
|---|---|
| Aged 16 to 49 years | 8,755 (100%) |
| At least 2 male sex partners in the past 90 days | 8,755 (100%) |
| Not currently participating in a clinical trial for an HIV vaccine or pre-exposure prophylaxis | 8,755 (100%) |
| Not currently on pre-exposure prophylaxis | 8,755 (100%) |
| Never diagnosed with HIV (self-report) | 8,755 (100%) |
| Currently residing in the United States or territories | 8,755 (100%) |
| Not cisgender female | 8,755 (100%) |
| Additional eligibility criteria (participants must meet at least 1) | |
| >1 receptive condomless anal sex acts with a male partner in the last 3 months | 5,190 (59.2%) |
| >2 insertive condomless anal sex acts with a male partner in the last 3 months | 3,833 (43.7%) |
| Used methamphetamines in the last 3 months | 1,051 (12%) |
| Rectal gonorrhea/chlamydia in the last 12 months | 680 (7.8%) |
| Syphilis diagnosis in the last 12 months | 400 (4.6%) |
| Used postexposure prophylaxis in the last 12 months | 219 (2.5%) |
| Shared injection drug needles in the last 12 months | 180 (2.1%) |
Participants clicking one an ad were routed from the apps to a secured screening survey that collected information on demographic characteristics, sexual behavior, and substance use. Those who screened eligible were later sent a link to complete a secondary survey that collected additional information. Participants completing this secondary survey received a $15 gift card and were subsequently mailed an OraSure HIV-1 Oral Specimen Collection device. Using a self-addressed and stamped envelope, oral fluid samples were returned to a lab for analysis. Participants who returned a sample received another $15. Study procedures were approved by the CUNY Institutional Review Board.
Response rates
In total, 8,755 participants screened eligible. Of these 6,267 (71.5%) completed the secondary survey and were mailed at-home HIV kits, and 5,001 (79.8%) of those who received kits returned them to the laboratory and had valid test results. An additional 64 participants returned kits to the lab, but the sample was invalid (e.g., container opened in transit). Data used for the current analyses were participants who completed the second survey and had valid test results (n = 5,001). We have published a manuscript describing characteristics associated with returning an HIV test kit versus not.11 Using the results from that study, we created internal stabilized survey weights using propensity scores for the 5,001 participants who completed baseline testing then weighted these participants back to the originally enrolled 8,755. A map showing the distribution of the 5,001 is shown in.
Study measures:
The 4-level polytomous outcome of interest for the current study was the participants’ HIV test result at enrollment (HIV-positive vs. HIV-negative) in relation to their self-reported HIV testing history. Therefore, participants were organized into four groups: 1) those with HIV-negative results at enrollment who reported testing for HIV in the year prior to enrollment, 2) those with HIV-negative results at enrollment who reported they had not tested for HIV in the year prior, 3) those with HIV-positive results at enrollment who reported they had tested HIV negative in the year prior to enrollment, and 4) those with HIV-positive results at enrollment who reported they had not tested for HIV in the year prior.
Covariates included sociodemographic characteristics—age, race/ethnicity, gender identity, education, and income—and other factors known to be related to HIV risk, status, and testing—i.e. number of HIV-positive male partners, condomless anal sex (CAS), having a primary care physician (PCP), housing instability (past 5 years), transactional sex (past 3 months), prior experience taking PrEP and/or post-exposure prophylaxis (PEP), having ever spoken to a doctor about PrEP, whether a participant perceived themselves as an appropriate candidate for PrEP, and methamphetamine use (past 3 months). Variables were coded 1 = yes, 0 = no.
Analysis Plan
Descriptive statistics—frequencies, percentages, means, and standard deviations—were used to describe participant characteristics. Tests of differences between HIV status groups were conducted using chi-squared or ANOVA. To account for variable survey response, we conducted a weighted multinomial logistic regression to examine group differences in the aforementioned variables; to exhaustively compare all four groups, the regression resulted in six models. Initial variable selection for the adjusted models was determined by results from the bivariate analyses, and variables retained in the final models were determined by fit criterion changes (e.g., AIC). For all regressions, adjusted odds ratios, 95% confidence intervals, and p-values are reported. Analyses were completed using SAS 9.4.
RESULTS
Table 2 reports bivariate differences across the four HIV-status groupings. Across all groups, the average age was 31 and nearly all participants identified as cis-male. Nearly half of participants were persons of color and a majority had at least some college education. In total, 192 participants (3.8%) had HIV-positive results at enrollment, with 35.4% of those who tested HIV-positive (n = 68) reporting that they had tested HIV-negative in the year prior to enrollment (i.e., recent HIV infections). Among those testing HIV-negative at enrollment (n = 4809), 36.1% said they had not been tested for HIV in the past year (n = 1739). A majority of those who tested HIV-positive at enrollment were persons of color, compared to a majority of those testing HIV-negative being white. Those testing HIV-positive were significantly less likely to have ever used PrEP or PEP and were significantly more likely to report co-factors known to be associated with HIV risk such as housing instability (last 5 years), lower income, transactional sex (past 3 months), and methamphetamine use (past 3 months).
Table 2.
Prior HIV testing behavior and HIV testing results, USA, 2017–2018
| HIV-negative, recently tested1 | HIV-negative, not recently tested2 | HIV-positive, recently tested negative1 | HIV-positive, not recently tested2 | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| N = 3070 | N = 1739 | N = 68 | N = 124 | |||||||
| n | % | n | % | n | % | Chi-Sq or F | p | |||
| Age (M, SD) | 30.9 | (7.7) | 30.8 | (8.3) | 30.7 | (7.0) | 31.7 | (7.7) | 1.64 | 0.65 |
| Gender | 4.55 | 0.21 | ||||||||
| Cismale | 2995 | (97.6) | 1705 | (98.0) | 68 | (100.0) | 122 | (98.4) | ||
| Transgender men or women | 75 | (2.4) | 34 | (2.0) | 0 | (0.0) | 2 | (1.6) | ||
| Race/Ethnicity | 54.06 | <.0001 | ||||||||
| White | 1596 | (52.0) | 990 | (56.9) | 24 | (35.3) | 55 | (44.4) | ||
| Black or African American | 320 | (10.4) | 140 | (8.1) | 17 | (25.0) | 27 | (21.8) | ||
| Latino | 732 | (23.8) | 421 | (24.2) | 14 | (20.6) | 30 | (24.2) | ||
| All other | 422 | (13.8) | 188 | (10.8) | 13 | (19.1) | 12 | (9.7) | ||
| Highest level of Education | 105.50 | <.0001 | ||||||||
| < High school diploma | 47 | (1.5) | 47 | (2.7) | 3 | (4.4) | 8 | (6.5) | ||
| High school diploma or GED | 328 | (10.7) | 291 | (16.7) | 17 | (25.0) | 22 | (17.7) | ||
| Some college or associates degree | 1297 | (42.3) | 803 | (46.2) | 29 | (42.7) | 62 | (50.0) | ||
| College graduate or higher | 1398 | (45.5) | 598 | (34.4) | 19 | (27.9) | 32 | (25.8) | ||
| Income | 57.72 | <.0001 | ||||||||
| Less than $20,000 | 910 | (29.6) | 607 | (34.9) | 31 | (45.6) | 56 | (45.2) | ||
| $20,000-$49,999 | 1251 | (40.8) | 751 | (43.2) | 23 | (33.8) | 51 | (41.1) | ||
| $50,000 or more | 909 | (29.6) | 381 | (21.9) | 14 | (20.6) | 17 | (13.7) | ||
| Have primary care physician | 45.41 | <.0001 | ||||||||
| Yes | 1723 | (56.1) | 804 | (46.2) | 39 | (57.4) | 59 | (47.6) | ||
| No | 1347 | (43.9) | 935 | (53.8) | 29 | (42.7) | 65 | (52.4) | ||
| Housing instability | 68.70 | <.0001 | ||||||||
| Yes, within the last 5 years | 540 | (17.6) | 353 | (20.3) | 30 | (44.1) | 55 | (44.4) | ||
| No or not within the last 5 years | 2530 | (82.4) | 1386 | (79.7) | 38 | (55.9) | 69 | (55.7) | ||
| Transactional sex in the past 3 months | 44.27 | <.0001 | ||||||||
| Yes | 390 | (12.7) | 243 | (14.0) | 24 | (35.3) | 37 | (29.8) | ||
| No | 2680 | (87.3) | 1496 | (86.0) | 44 | (64.7) | 87 | (70.2) | ||
| Ever taken PrEP | 235.24 | <.0001 | ||||||||
| No | 2453 | (79.9) | 1644 | (94.5) | 62 | (91.2) | 120 | (96.8) | ||
| Yes | 617 | (20.1) | 95 | (5.5) | 6 | (8.8) | 4 | (3.2) | ||
| Ever taken PEP | 87.36 | <.0001 | ||||||||
| No | 2805 | (91.4) | 1696 | (97.5) | 64 | (94.1) | 122 | (98.4) | ||
| Yes | 265 | (8.6) | 43 | (2.5) | 4 | (5.9) | 2 | (1.6) | ||
| Ever spoken to a doctor about PrEP | 499.92 | <.0001 | ||||||||
| Yes | 1300 | (42.4) | 232 | (13.3) | 13 | (19.1) | 15 | (12.1) | ||
| No | 1770 | (57.7) | 1507 | (86.7) | 55 | (80.9) | 109 | (87.9) | ||
| Believe that they are a candidate for PrEP | 32.43 | <.0001 | ||||||||
| Yes | 2322 | (75.6) | 1195 | (68.7) | 44 | (64.7) | 81 | (65.3) | ||
| No/Not sure | 748 | (24.4) | 544 | (31.3) | 24 | (35.3) | 43 | (34.7) | ||
| Methamphetamine use (past 3 months) | 119.16 | <.0001 | ||||||||
| Yes | 273 | (8.9) | 160 | (9.2) | 29 | (42.7) | 46 | (37.1) | ||
| No | 2797 | (91.1) | 1579 | (90.8) | 39 | (57.4) | 78 | (62.9) | ||
| No. of times having receptive CAS < 3 months (M, SD) | 4.0 | (13.3) | 4.0 | (7.9) | 7.2 | (9.3) | 6.7 | (10.8) | 6.18 | 0.10 |
| No. of times having insertive CAS < 3 months (M, SD) | 4.3 | (7.7) | 4.0 | (7.2) | 6.8 | (13.8) | 4.0 | (6.9) | 8.87 | 0.03 |
| No. of HIV-positive male partners < 3 months (M, SD) | 0.6 | (2.3) | 0.6 | (3.8) | 1.2 | (2.4) | 1.1 | (2.6) | 4.51 | 0.21 |
Participant indicated last HIV negative results were ≤ 12 months ago
Participant indicated last HIV-negative results were > 12 months ago, or never tested for HIV
How HIV-negative, recent testers compared to other participants:
Table 3 (Part A) presents comparisons of the HIV testing groups. Compared to participants who had HIV-negative results and said they tested for HIV in the 12 months prior to enrollment, those who had negative results and had not tested in the 12 months prior to enrollment were older, more likely to be white, less likely to have attended college, not have a primary care provider, have experienced housing instability (last 5 years), less likely to have ever taken PrEP or PEP, less likely to perceive themselves as an appropriate candidate for PrEP, and less likely to have ever spoken to a doctor about PrEP. Next, compared to participants who had HIV-negative results and said they had tested for HIV in the 12 months prior to enrollment, those who had HIV-positive results at enrollment and not tested in the 12 months prior to enrollment were significantly more likely to be older, Black, have engaged in transactional sex recently, been unstably housed (past 5 years). They were also significantly more likely to have used methamphetamine recently. They were significantly less likely to believe they were an appropriate candidate for PrEP. Third, compared to participants who had HIV-negative results and said they had tested for HIV in the 12 months prior to enrollment, those with HIV-positive results at enrollment who said they had tested HIV-negative within the 12 months prior to enrollment were significantly more likely to be Black, have experienced housing instability (last 5 years), and have used methamphetamine recently. They were also significantly less likely to have a PCP or ever spoken to a doctor about PrEP.
Table 3.
Weighted multinomial logistic regression - factors associated with prior HIV testing behavior and HIV testing results, USA, 2017–2018
| HIV-negative, not recently tested2 | HIV-positive, not recently tested2 | HIV-positive, recently tested negative1 | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Part A. | |||||||||||||||
| Comparison Group: HIV-Negative, recently tested1 | AOR | 95% CI | p | AOR | 95% CI | p | AOR | 95% CI | p | ||||||
| Age in years | 1.01 | 1.00 | -- | 1.02 | 0.03 | 1.03 | 1.00 | -- | 1.06 | 0.02 | 1.00 | 0.97 | -- | 1.03 | 0.99 |
| Race/Ethnicity | |||||||||||||||
| White (ref.) | |||||||||||||||
| Black or African American | 0.61 | 0.50 | -- | 0.75 | <.0001 | 2.88 | 1.82 | -- | 4.58 | <.0001 | 3.97 | 2.14 | -- | 7.36 | <.0001 |
| Latino | 0.87 | 0.74 | -- | 1.01 | 0.07 | 1.18 | 0.74 | -- | 1.89 | 0.50 | 1.27 | 0.65 | -- | 2.49 | 0.48 |
| All other | 0.74 | 0.60 | -- | 0.91 | 0.004 | 0.93 | 0.49 | -- | 1.78 | 0.83 | 2.18 | 1.08 | -- | 4.38 | 0.03 |
| Highest level of Education | |||||||||||||||
| < High school diploma | 1.24 | 0.84 | -- | 1.84 | 0.28 | 1.98 | 0.92 | -- | 4.24 | 0.08 | 1.00 | 0.36 | -- | 2.80 | 1.00 |
| High school diploma or GED (ref.) | |||||||||||||||
| Some college or associates degree | 0.79 | 0.66 | -- | 0.95 | 0.01 | 0.87 | 0.55 | -- | 1.38 | 0.54 | 0.57 | 0.31 | -- | 1.02 | 0.0585 |
| College graduate or higher | 0.59 | 0.48 | -- | 0.71 | <.0001 | 0.61 | 0.35 | -- | 1.06 | 0.08 | 0.58 | 0.29 | -- | 1.16 | 0.12 |
| Have primary care physician | |||||||||||||||
| No | 1.26 | 1.11 | -- | 1.44 | 0.0004 | 1.02 | 0.71 | -- | 1.47 | 0.92 | 0.56 | 0.34 | -- | 0.91 | 0.02 |
| Housing instability in the last 5 years (yes) | 1.24 | 1.05 | -- | 1.46 | 0.01 | 2.59 | 1.75 | -- | 3.85 | <.0001 | 2.06 | 1.22 | -- | 3.48 | 0.01 |
| Transactional sex in the past 3 months (yes) | 1.06 | 0.88 | -- | 1.28 | 0.52 | 1.57 | 1.03 | -- | 2.39 | 0.04 | 1.66 | 0.96 | -- | 2.87 | 0.07 |
| Ever taken PrEP (yes) | 0.57 | 0.44 | -- | 0.74 | <.0001 | 0.38 | 0.15 | -- | 0.98 | 0.05 | 0.63 | 0.25 | -- | 1.56 | 0.32 |
| Ever taken PEP (yes) | 0.52 | 0.37 | -- | 0.74 | 0.0003 | 0.27 | 0.06 | -- | 1.11 | 0.07 | 1.08 | 0.40 | -- | 2.95 | 0.87 |
| Ever spoken to a doctor about PrEP | |||||||||||||||
| No | 3.61 | 3.02 | -- | 4.31 | <.0001 | 4.01 | 2.24 | -- | 7.16 | <.0001 | 2.77 | 1.44 | -- | 5.32 | 0.002 |
| Believe that they are a candidate for PrEP (yes) | 0.85 | 0.74 | -- | 0.97 | 0.02 | 0.67 | 0.46 | -- | 0.98 | 0.04 | 0.60 | 0.36 | -- | 1.00 | 0.05 |
| Methamphetamine use, past 3 months (yes) | 0.86 | 0.68 | -- | 1.08 | 0.18 | 4.06 | 2.64 | -- | 6.25 | <.0001 | 6.62 | 3.80 | -- | 11.56 | <.0001 |
| Part B. | |||||||||||||||
| Comparison Group: HIV-positive, recently tested negative1 | AOR | 95% CI | p | AOR | 95% CI | p | |||||||||
| Age in years | 1.02 | 0.98 | -- | 1.05 | 0.35 | 1.03 | 0.99 | -- | 1.08 | 0.12 | |||||
| Race/Ethnicity | |||||||||||||||
| White (ref.) | |||||||||||||||
| Black or African American | 0.17 | 0.09 | -- | 0.32 | <.0001 | 0.77 | 0.36 | -- | 1.63 | 0.49 | |||||
| Latino | 0.57 | 0.29 | -- | 1.14 | 0.11 | 0.82 | 0.37 | -- | 1.83 | 0.63 | |||||
| All other | 0.30 | 0.15 | -- | 0.63 | 0.001 | 0.41 | 0.16 | -- | 1.04 | 0.06 | |||||
| Highest level of Education | |||||||||||||||
| < High school diploma | 1.26 | 0.42 | -- | 3.80 | 0.68 | 2.18 | 0.64 | -- | 7.49 | 0.22 | |||||
| High school diploma or GED (ref.) | |||||||||||||||
| Some college or associates degree | 1.30 | 0.71 | -- | 2.41 | 0.397 | 1.54 | 0.74 | -- | 3.18 | 0.25 | |||||
| College graduate or higher | 1.11 | 0.53 | -- | 2.30 | 0.79 | 1.12 | 0.47 | -- | 2.70 | 0.80 | |||||
| Have primary care physician | |||||||||||||||
| No | 2.36 | 1.42 | -- | 3.93 | 0.001 | 1.86 | 1.02 | -- | 3.39 | 0.04 | |||||
| Housing instability in the last 5 years (yes) | 0.66 | 0.38 | -- | 1.15 | 0.14 | 1.31 | 0.69 | -- | 2.50 | 0.41 | |||||
| Transactional sex in the past 3 months (yes) | 0.66 | 0.37 | -- | 1.15 | 0.14 | 0.96 | 0.50 | -- | 1.86 | 0.91 | |||||
| Ever taken PrEP (yes) | 0.95 | 0.34 | -- | 2.63 | 0.92 | 0.58 | 0.15 | -- | 2.18 | 0.42 | |||||
| Ever taken PEP (yes) | 0.81 | 0.24 | -- | 2.72 | 0.73 | 0.27 | 0.05 | -- | 1.65 | 0.16 | |||||
| Ever spoken to a doctor about PrEP | |||||||||||||||
| No | 1.44 | 0.71 | -- | 2.90 | 0.32 | 1.53 | 0.64 | -- | 3.64 | 0.34 | |||||
| Believe that they are a candidate for PrEP (yes) | 1.34 | 0.80 | -- | 2.25 | 0.27 | 1.09 | 0.60 | -- | 2.01 | 0.77 | |||||
| Methamphetamine use, past 3 months (yes) | 0.12 | 0.07 | -- | 0.22 | <.0001 | 0.61 | 0.31 | -- | 1.19 | 0.15 | |||||
| Part C. | |||||||||||||||
| Comparison Group: HIV-positive, not recently tested2 | AOR | 95% CI | p | ||||||||||||
| Age in years | 0.99 | 0.96 | -- | 1.01 | 0.28 | ||||||||||
| Race/Ethnicity | |||||||||||||||
| White (ref.) | |||||||||||||||
| Black or African American | 0.23 | 0.14 | -- | 0.37 | <.0001 | ||||||||||
| Latino | 0.73 | 0.45 | -- | 1.18 | 0.20 | ||||||||||
| All other | 0.79 | 0.41 | -- | 1.55 | 0.50 | ||||||||||
| Highest level of Education | |||||||||||||||
| < High school diploma | 0.58 | 0.26 | -- | 1.28 | 0.18 | ||||||||||
| High school diploma or GED (ref.) | |||||||||||||||
| Some college or associates degree | 0.85 | 0.53 | -- | 1.37 | 0.51 | ||||||||||
| College graduate or higher | 1.00 | 0.56 | -- | 1.80 | 0.99 | ||||||||||
| Have primary care physician | |||||||||||||||
| No | 1.26 | 0.86 | -- | 1.86 | 0.23 | ||||||||||
| Housing instability in the last 5 years (yes) | 0.51 | 0.34 | -- | 0.76 | 0.001 | ||||||||||
| Transactional sex in the past 3 months (yes) | 0.71 | 0.46 | -- | 1.11 | 0.13 | ||||||||||
| Ever taken PrEP (yes) | 1.62 | 0.61 | -- | 4.34 | 0.34 | ||||||||||
| Ever taken PEP (yes) | 2.68 | 0.59 | -- | 12.25 | 0.20 | ||||||||||
| Spoken to a doctor about PrEP | |||||||||||||||
| No | 0.91 | 0.49 | -- | 1.67 | 0.75 | ||||||||||
| Believe that they are a candidate for PrEP (yes) | 1.27 | 0.86 | -- | 1.88 | 0.232 | ||||||||||
| Methamphetamine use, past 3 months (yes) | 0.20 | 0.13 | -- | 0.32 | <.0001 | ||||||||||
Participant indicated last HIV negative results were ≤ 12 months ago
Participant indicated last HIV-negative results were > 12 months ago, or never tested for HIV
How HIV-positive, recent testers compared to other participants:
Next (Table 3, Part B), compared to participants who had HIV-positive results at enrollment and said they had tested HIV-negative within the 12 months prior to enrollment, those who had HIV-negative results at enrollment and had not tested in the 12 months prior to enrollment were significantly less likely to be Black, to have recently used methamphetamine, or to have a primary care provider. And, compared to participants who had HIV-positive results at enrollment and said they had tested HIV-negative within the 12 months prior to enrollment, those with HIV-positive results at enrollment who had not tested for HIV in the last 12 months were more likely to report having a PCP.
How HIV-positive, not recently tested compared to other participants:
Finally (Table 3, Part C), compared to participants who had HIV-positive results at enrollment and had not tested for HIV in the 12 months prior to enrollment, those who had HIV-negative results at enrollment and also had not tested for HIV in the 12 months prior to enrollment were significantly less likely to be Black, have experienced housing instability (last 5 years), or have used methamphetamine (last 3 months).
DISCUSSION
In this US national sample of over 5,000 GBM, 3.8% tested positive for HIV and, among those with positive results, 35% were likely recent HIV infections (i.e., evidenced by self-reporting an HIV-negative test result within the 12 months prior to enrollment). Regardless of HIV testing results at enrollment, those who had tested in the last year were more likely to report having a primary care provider (~56% vs. ~47% among those not having tested in the year preceding enrollment). This highlights the important function that simply having a primary care provider has in access to basic healthcare like HIV testing.12,13 Meanwhile, those with HIV-positive results, regardless of how recent their previous HIV test was prior to enrollment, differed from those with negative results in many ways that are known to be associated with HIV risk. These included racial and income disparities, housing instability, recent transactional sex, and perhaps most notably, recent methamphetamine use.2,3,5,14–16 Approximately four out of every ten individuals in our study with HIV-positive results had recently used methamphetamine, compared with fewer than one in ten among those with HIV-negative results. This finding highlights the ongoing need to address what has been termed the “double epidemic” whereby methamphetamine use is integrally connected to risk for HIV.17
Furthermore, those with negative results at enrollment who recently tested for HIV were the most likely to have prior experience with both PrEP and PEP (20% and 9% respectively). HIV testing is a component associated with PrEP and PEP care, thus this finding is intuitive. However, we would add that both PrEP and PEP—during the time in which they are being used—greatly protect individuals from HIV, which likely also contributed to the fact that these individuals were the most likely to have HIV-negative results. That is, although these participants were not on PrEP or PEP at the time of enrollment, these individuals were more likely to have experienced time-periods in their lives in which they were pharmacologically protected against HIV.
In multivariate modeling, among those with HIV-positive results at enrollment, only having a PCP distinguished differences between those who recently tested negative prior to enrollment versus not. Whereas, there were numerous distinctions between those who recently tested negative prior to enrollment (versus not), among those with HIV-negative results at enrollment. Furthermore, the characteristics that distinguished those with negative results who had not recently tested were generally those considered to put a person at-risk for HIV including things like housing instability, transactional sex, and not having a primary care provider. Our findings highlight that strategies aimed at increasing HIV testing among this group might be well served to holistically and synergistically address broader needs of this population. And although it is beyond the scope of the present study, these findings also highlight the need to investigate interpersonal and psychological barriers that may prevent someone from seeking HIV testing, such as perceived stigma and avoidance behaviors.
However, our findings should be understood in light of their limitations. Aside from the laboratory-based HIV testing, much of our data are self-report. Furthermore, our findings are cross-sectional, thus causality cannot be inferred. We used the Orasure HIV-1 oral specimen collection device, which is a third generation HIV test kit. The window period for this test can be up to three months. As a result, it is possible that very recent HIV seroconversions were not detected in our study. Although fourth generation HIV tests have a narrower window period (and thus more accuracy), collection methods to conduct such a test require blood samples, which was not feasible in our study.
Next, although participants were geographically distributed across all 50 states, Puerto Rico, and Guam, the study itself is not designed to be representative of all gay and bisexual men. Rather, it was designed to enroll a sample of individuals at high risk for HIV. Furthermore, it is possible that there may be regional or even state-by-state variation associated with both independent and dependent variables (e.g., regional variation in the prevalence methamphetamine use)—participants located closer to one another may be more similar than those located further apart. Also, because HIV results were our outcome of interest, only those who returned kits were included in analyses for the present study. We have reported elsewhere demographic differences in rates of kit return.11 Briefly, although the difference was small, those having returned a kit were more likely to be white—a facet that has been observed in other studies. Nevertheless, this bias should be noted.
In addition, participants were enrolled using geosocial networking apps. Although these apps have emerged as the modal way in which gay and bisexual men meet their male sex partners today,18–22 it is important to remember that not all individuals use these apps or own smart phones. The Together 5,000 cohort study was designed to reach cisgender gay and bisexual men; however, we enrolled transgender men and transgender women met enrollment criteria. Nevertheless, the sample size for this group was relatively small and all of whom had HIV-negative results at enrollment. Although we are limited in the scope of what kinds of analyses can be conducted with a small sample of transgender individuals, we highlight the importance of including transgender individuals in HIV prevention research.
Finally, online studies can potentiate fraudulent participants and data.23–27 Although we took great lengths to avoid this (such as recording IP addresses, requiring unique and valid email and mailing addresses, preventing multiple entries from a given device, and geo-restricting advertisement to those GPS located within the US), we recognize there is some potential for these individuals to make their way into online research studies. Of note, however, some challenges regarding the accuracy of a given person’s data also exist even if a person were to participate face-to-face in a researcher’s office/lab.
CONCLUSION
Participants with HIV-positive results differed from those with HIV-negative results in ways that are known to both directly and indirectly put individuals at risk for HIV in the first place. The factor largest in magnitude that appeared to differentiate these two groups was recent methamphetamine use. Our findings demonstrate the continued need to incorporate methamphetamine treatment and education with HIV prevention and treatment. Next, among those with HIV-positive results, few factors differentiated whether these participants had recently tested for HIV prior to enrollment, or not. This would be an area for further investigation to determine if such a pattern was observed in other settings and contexts.
In contrast, among those with HIV-negative results, there were numerous notable differences between those who had recently tested for HIV prior to enrollment, or not. Those with HIV-negative results who had not recently tested for HIV comprised 35% of our sample, which is not an insignificant number, given the risk profile of the cohort. Strategies aimed to engage this group in HIV testing might be well served were they to dualistically address other needs such as stable housing, transactional sex, and access to a primary care provider.
SUMMARY BOX.
What is already known on the subject?
Men, trans women, and trans men who have sex with men are disproportionally impacted by HIV. HIV testing serves as an important gateway into the status neutral continuum of care. That is, only when someone knows their HIV status can appropriate measures be taken to either treat their HIV (if HIV-positive) or help prevent infection (if HIV-negative). Examining factors associated with different patterns of testing behaviors may serve to develop tailored interventions that improve HIV testing behaviors.
What does this study add?
Using a U.S. national geographically diverse sample of 5,001 men, trans women, and trans men who have sex with men, we compared patterns of HIV testing prior to enrollment relative to one’s HIV results (using an at-home self-collected HIV test kit). In total, 3.8% tested HIV-positive and—among those with positive results—35% were recent HIV infections (i.e., self-reported an HIV-negative test result within the 12 months prior to enrollment). Those with HIV-positive results—regardless of how recent their HIV test was prior to enrollment—differed from those with negative results in ways that are known to be associated with HIV risk: racial and income disparities, housing instability, recent transactional sex, and recent methamphetamine use. Among those with HIV-positive results at enrollment, only having a primary care physician distinguished those who recently tested negative prior to enrollment versus not. Whereas, among those with HIV-negative results, there were numerous differences between those who had recently tested for HIV prior to enrollment, versus not, such that those who had not recently tested were significantly more likely to report being at higher risk for HIV.
Acknowledgements:
Special thanks to additional members of the T5K study team: Sarah Kulkarni, Alexa D’Angelo, Pedro Carneiro, Chloe Mirzayi, Caitlin MacCrate, Corey Morrison, Matthew Stief, Javier Lopez-Rios, Viraj Patel, David Pantalone, Don Hoover, Greg Millett, Sabina Hirshfield, Adam Carrico, Michael Camacho, Demetre Daskalakis, & Jeremiah Johnson. Thank you to the program staff at NIH: Gerald Sharp, Sonia Lee, and Michael Stirratt. The NIH had no role in the production of this manuscript nor necessarily endorses its findings.
Funding:
Together 5,000 was funded by the National Institutes for Health (UG3 AI 133675 - PI Grov). Other forms of support include the CUNY Institute for Implementation Science in Population Health, the Einstein, Rockefeller, CUNY Center for AIDS Research (ERC CFAR, P30 AI124414)
Footnotes
Competing Interest: None to declare.
REFERENCES
- 1.CDC. HIV and Gay and Bisexual Men. 2018; https://www.cdc.gov/hiv/group/msm/index.html. Accessed October 17, 2018.
- 2.CDC. HIV Among Gay and Bisexual Men. 2016; http://www.cdc.gov/hiv/group/msm/. Accessed August 22, 2016.
- 3.CDC. Half of black gay men and a quarter of Latino gay men projected to be diagnosed within their lifetime. 2016; http://www.cdc.gov/nchhstp/newsroom/2016/croi-press-release-risk.html. Accessed April 26, 2016.
- 4.Myers JE, Braunstein SL, Xia Q, et al. Redefining prevention and care: A status-neutral approach to HIV. Paper presented at: Open forum infectious diseases 2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Grov C, Rendina HJ, Parsons JT. How different are men who do not know their HIV status from those who do? Results from an U.S. online study of gay and bisexual men. AIDS Behav. 2016;20(1989–1999). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.DiNenno EA, Prejean J, Irwin K, et al. Recommendations for HIV screening of gay, bisexual, and other men who have sex with men—United States, 2017. MMWR Morbidity and mortality weekly report. 2017;66(31):830. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Clark HA, Oraka E, DiNenno EA, et al. Men Who Have Sex with Men (MSM) Who Have Not Previously Tested for HIV: Results from the MSM Testing Initiative, United States (2012–2015). AIDS Behav. 2018:1–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Edelstein Z, Salcuni P, Khawja A, et al. Results from the HIV home test giveaway, New York City, 2015. 2016, November; Presentation given at the Annual meeting of the American Public Health Association, Nov 2, Denver, Co. https://apha.confex.com/apha/144am/meetingapp.cgi/Paper/353199#_blank. Accessed Feb 20, 2019. [Google Scholar]
- 9.Grov C, Westmoreland DA, Carneiro PB, et al. Recruiting vulnerable populations to participate in HIV prevention research: findings from the Together 5000 cohort study. Annals of epidemiology. 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Nash D, Stief M, MacCrate C, et al. A Web-Based Study of HIV Prevention in the Era of Pre-Exposure Prophylaxis Among Vulnerable HIV-Negative Gay and Bisexual Men, Transmen, and Transwomen Who Have Sex With Men: Protocol for an Observational Cohort Study. JMIR Research Protocols. 2019;8(9):e13715. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Grov C, Westmoreland DA, Carneiro PB, et al. Recruiting vulnerable populations to participate in HIV prevention research: Findings from the Together 5,000 cohort study. Ann Epidemiol. 2019;35:4–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Henny KD, Duke CC, Geter A, et al. HIV-Related Training and Correlates of Knowledge, HIV Screening and Prescribing of nPEP and PrEP Among Primary Care Providers in Southeast United States, 2017. AIDS Behav. 2019:1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.McNaghten AD, Valverde EE, Blair JM, Johnson CH, Freedman MS, Sullivan PS. Routine HIV testing among providers of HIV care in the United States, 2009. PLoS ONE. 2013;8(1):e51231. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Grov C, Parsons JT, Bimbi DS. In the shadows of a prevention campaign: Sexual risk in the absence of crystal methamphetamine. AIDS Educ Prev. 2008;20:42–55. [DOI] [PubMed] [Google Scholar]
- 15.Clatts MC, Goldsamt L, Yi H, Gwadz MV. Homelessness and drug abuse among young men who have sex with men in New York city: a preliminary epidemiological trajectory. J Adolesc. 2005;28(2):201–214. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Bond KT, Yoon IS, Houang ST, DJM J, Grov C, Hirshfield S. Transactional sex, substance use, and sexual risk: comparing pay direction for an Internet-based U.S. sample of men who have sex with men. Sexuality Research & Social Policy. in press. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Halkitis PN, Parsons JT, Stirratt MJ. A double epidemic: crystal methamphetamine drug use in relation to HIV transmission among gay men. J Homosex. 2001;41(2):17–35. [DOI] [PubMed] [Google Scholar]
- 18.Grov C, Breslow AS, Newcomb ME, Rosenberger JG, Bauermeister JA. Gay and bisexual men’s use of the Internet: research from the 1990s through 2013. The Journal of Sex Research. 2014;51(4):390–409. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Grov C, Cain D, Whitfield TH, et al. Recruiting a US national sample of HIV-negative gay and bisexual men to complete at-home self-administered HIV/STI testing and surveys: Challenges and opportunities. Sexuality Research and Social Policy. 2016;13(1):1–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Goedel WC, Duncan DT. Geosocial-Networking App Usage Patterns of Gay, Bisexual, and Other Men Who Have Sex With Men: Survey Among Users of Grindr, A Mobile Dating App. JMIR Public Health and Surveillance. 2015;1(1):e4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Phillips II G, Magnus M, Kuo I, et al. Use of Geosocial Networking (GSN) Mobile Phone Applications to Find Men for Sex by Men Who Have Sex with Men (MSM) in Washington, DC. AIDS and behavior. 2014:1–8. [DOI] [PubMed] [Google Scholar]
- 22.Phillips II G, Grov C, Mustanski B. Engagement in group sex among geosocial networking mobile application-using men who have sex with men. Sexual Health. 2015;12(6):495–500. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Teitcher JEF, Bockting WO, Bauermeister JA, Hoefer CJ, Miner MH, Klitzman RL. Detecting, preventing, and responding to “fraudsters” in Internet research: ethics and tradeoffs. The Journal of Law, Medicine & Ethics. 2015;43(1):116–133. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Bauermeister JA, Pingel E, Zimmerman M, Couper M, Carballo-Dieguez A, Strecher VJ. Data quality in HIV/AIDS web-based surveys: Handling invalid and suspicious data. Field Methods. 2012;24(3):272–291. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Pequegnat W, Rosser BS, Bowen AM, et al. Conducting Internet-based HIV/STD prevention survey research: considerations in design and evaluation. AIDS and behavior. 2007;11(4):505–521. [DOI] [PubMed] [Google Scholar]
- 26.Grey JA, Konstan J, Iantaffi A, Wilkerson JM, Galos D, Rosser BRS. An updated protocol to detect invalid entries in an online survey of men who have sex with men (MSM): how do valid and invalid submissions compare? AIDS Behav. 2015:1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Konstan JA, Simon BRS, Ross MW, Stanton J, Edwards WM. The story of subject naught: A cautionary but optimistic tale of Internet survey research. Journal of Computer‐Mediated Communication. 2005;10(2):00–00. [Google Scholar]