Whether Americans seek HIV testing is linked to race and perceived risk. (Digests).

Nearly one-third of American adults have ever been tested for HIV, excluding those who have been tested to qualify as blood donors. According to a review of HIV testing data from the 1999 National Health Interview Survey (NH1S), (1) the proportion ever tested differs widely by race, with blacks being most likely to have had a test (46%), followed by Hispanics (33%) and whites (29%). Similar racial and ethnic differences characterize recent testing: Twenty percent of blacks have been tested for HIV within the past 12 months, compared with 12% of Hispanics and 8% of whites. Although Americans who perceive themselves to be at risk for HIV or who have engaged in high-risk behaviors are more likely than others to have ever been tested, substantial proportions of individuals in either of these categories--39% of whites, 35% of similar Hispanics and 26% of blacks-have never been tested for the virus.

The NHIS is an annual household-based survey of a representative sample of the civilian, noninstitutionalized U.S. adult population. Of the 30,801 respondents, 2% reported having participated in at least one of the five HIV-risk behaviors * that the survey asked about. Respondents were also asked whether they perceived their risk to be high, medium, low or nonexistent; 2% perceived themselves to be at medium or high risk for contracting HIV. Thus, 4% of the total sample fell into either the perceived or the actual HIV-risk category.

Thirty-one percent of respondents had ever been tested for HIV (excluding testing that was a requirement for donating blood). This proportion was significantly higher among blacks (46%) than among either Hispanics (33%) or whites (29%). Overall rates of HIV testing were consistently higher among Americans who were at risk than among those who were not. For example, roughly three-quarters (73%) of the sample who reported a risky behavior had ever been tested, compared with 30% of those not citing any such behavior. Similarly, 54% of those who perceived themselves to be at medium or high risk had ever been tested, compared with 30% of others. Finally, 61% of respondents who fell into either the actual or the perceived risk category had ever been tested, compared with 30% of those in neither risk classification.

Among Americans who reported having engaged in an HIV-risk behavior, the proportion ever tested was higher among blacks (82%) than among whites and Hispanics (73-74%). Blacks who characterized their personal risk as medium or high were also more likely to have been tested (70%) than were similar Hispanics (63%) or whites (51%). These differences by race also characterized the proportions ever tested who were in either at-risk category--73% among at-risk blacks, compared with 65% among at-risk Hispanics and 60% among similar whites.

Rates of recent HIV testing (i.e., within the past 12 months) were also higher among blacks (20%) than among Hispanics (12%), whites (8%) or the U.S. population as a whole (10%). Blacks who either reported a risk factor or who perceived themselves to be at risk were also more likely to have been tested in the past year (40%) than were similar Hispanics (28%) or whites (23%).

When asked why they decided to get tested for HIV in the past year, 43% of recently tested blacks said they did so "just to find out their infection status," compared with 34% of similar Hispanics and 26% of similar whites. The proportion who noted that they had been tested to fulfill a requirement (i.e., for hospitalization, surgery, health or life insurance, compliance with provider guidelines, a new job, admission into the military or compliance with immigration law) was highest among recently tested whites (39%); 32% of similar Hispanics and 26% of similar blacks gave this response. Hispanics were the most likely to have been tested because it had been recommended by a doctor, a sexual partner or a health department (35%, compared with 29% of the other groups).

The researchers caution that these data are likely to be underestimates. First, like all self-reported data, the findings from the NHIS may be affected by recall or other personal bias. Second, highly sensitive behaviors may be underreported, given that these questions were asked face-to-face; further, respondents may not know all there is to know about their partners' past behaviors and thus underestimate their own risk status. Third, hospitalized and institutionalized persons are excluded from the sampling frame. The findings are further limited by the fact that respondents' actual infection status is unknown.

The researchers note that the 31% prevalence rate for HIV testing in 1999 represents an important increase from past years (i.e., from rates of 5% in i987 and 26% in 1995), and while testing did not vary much by race in the late 1980s, important racial and ethnic differentials were apparent by the late 1990s. Moreover, substantial proportions of men and women who either perceived themselves to be at medium or high risk for HIV or had engaged in an HIV--risk behavior had never been tested for HIV-26% of at-risk blacks, 35% of at-risk Hispanics and 39% of at-risk whites. The investigators conclude that these important proportions of at-risk individuals who remain untested have important public health implications; thus "prevention programs should continue to develop innovative methods for counseling and testing at-risk persons."

* The five behaviors that qualified as high-risk were, for hemophiliacs, having received clotting factor concentrations; for men, ever having had sex with another man since 1980; ever having taken street drugs by needle since 1980; ever having traded sex for money since 1980; and ever having had sex with a person who fits any of the above descriptions.

REFERENCE

(1.) Centers for Disease Control and Prevention, HIV testing among racial/ethnic minorities-Unit ed States, 1999, Morbidity and Mortality Weekly Report, 2001, 50(47): 1054-1057.