Current guidelines recommend that men who have sex with men be tested annually for HIV, and results of a survey conducted among such men in six U.S. cities suggest how important that recommendation may be.1 In a subsample of participants who had never tested positive for the virus before the survey, only 54% said that they had had an HIV test within the previous year; the rest had been tested in the more distant past or had never been tested at all. Tests taken in conjunction with the survey revealed that 10% of these men were infected; more than half of HIV-positive men had not been tested in the last year.

The analyses were based on data from the second phase of the Young Men's Survey, which was conducted in 1998–2000 in Baltimore, Dallas, Los Angeles, Miami, New York and Seattle. Survey respondents were 23–29-year-old men who were recruited at venues frequented by men who have sex with men; those who agreed to participate completed standard interviews, had blood drawn for HIV testing, received HIV counseling and, if appropriate, were given a referral for care. Only respondents who had never had a positive HIV test result before the survey were included in the analytic sample. The analysts used logistic regression to identify characteristics associated with recent HIV testing and, among those who had been tested recently, anonymous testing, testing because of exposure to risk and receipt of counseling with testing.

Of the 2,797 men in the sample, 50% were white, 24% Hispanic, 19% black and the rest members of other racial or ethnic groups. The majority had at least some postsecondary education and were employed. Sixty-three percent had a regular health care provider; of this group, 69% felt that it was important to receive HIV prevention services from their provider, and 57% had ever discussed HIV testing with their provider. Fifty-four percent of the sample had had an HIV test within the last year, and 46% had had one longer ago or had never been tested.

Men aged 23–25 were significantly more likely than those aged 26–29 to have been tested within the last year (odds ratio from multivariate analysis, 1.2), and those whose annual income was $30,000 or more had higher odds of recent testing than those with an income of less than $15,000 (1.3). Participants who had disclosed their sexual orientation to many persons and those who were aware of highly active antiretroviral therapy (HAART) had elevated odds of having been tested recently (1.3 and 1.5, respectively), as did respondents who had been recruited at a site other than a gay social organization, users of illicit drugs, men who had had six or more male partners, those who had had an STD, those who considered themselves unlikely to be infected with HIV and those who had told new sex partners in the previous six months their HIV status (1.2–1.8). Respondents who had a regular health care provider and said that their provider had discussed HIV testing with them were more likely than those who either did not have a provider or had not discussed testing to have had a test in the last year (1.9), and men who used a provider and considered prevention services important had greater odds of having been tested than those who did not have a provider or did not consider these services important (1.4).

In tests taken as part of the study, 10% of respondents were found to be HIV-positive; of this group, 46% had tested negative within the year preceding the survey, and 54% had not been tested recently. The 271 HIV-positive men reported having had a total of 1,796 male partners and 89 female partners in the six months before the survey; the majority of partners (68% of males and 70% of females) were reported by men who had not been tested within the last year.

Among men who reported a recent HIV test, half had undergone anonymous testing. The odds that HIV testing was anonymous were elevated for white men; those who had no health care provider or had not discussed testing with their provider; those who had been tested at a site other than a private physician's office, a health maintenance organization or a hospital; those who had sought the test because of exposure to risk and those who had received counseling (odds ratios, 1.3–11.6).

Half of men who had recently been tested had been motivated by concern about exposure to risk. Risk-based testing was particularly likely among men who were aware of HAART, men who had ever had anal intercourse without using a condom, men who had been tested anonymously, men who had asked at least one recent new sex partner about his HIV status and men who had reduced their risks after being tested (1.3–1.8).

Either before or after receiving their results, three-fifths of respondents reporting a recent HIV test had received counseling that covered HIV and AIDS, HIV therapy, reasons for testing, risk behaviors or risk reduction. The likelihood that men had received counseling along with their testing was increased for those whose provider had discussed testing with them, those who knew of HAART, those who had had unprotected anal sex, those tested at a public clinic, those who considered counseling important and those who had reduced their risks after testing (odds ratios, 1.5–2.9).

Nearly half of participants who had had an HIV test within a year before the survey said they had reduced their HIV risk because of the testing or counseling experience. Nevertheless, 8% were HIV-positive when tested for the study. The proportion who had seroconverted was 24% among blacks and 1–6% among other racial and ethnic groups; it did not differ by reason for testing or by whether the men had received counseling.

The analysts consider many of their findings encouraging, including the level of use of health care, the importance that respondents attached to receiving prevention services and counseling, and the association between testing and risk reduction. However, they speculate that seroconversion rates and risky behavior among respondents who had not been tested recently may suggest that some men underestimate their level of risk. The findings, the researchers conclude, "underscore the urgency for renewed initiatives and policies to increase HIV testing and risk reduction among young [men who have sex with men}, especially those who are black."—D. Hollander


1. MacKellar DA et al., Recent HIV testing among young men who have sex with men: correlates, contexts, and HIV seroconversion, Sexually Transmitted Diseases, 2006, 33(3):183–192.