This report presents data collected through the HIV Testing Survey, conducted during 2002 in the states of Florida, Illinois, Michigan, New Jersey, and Washington and the cities of Los Angeles (California), New York City (New York), Philadelphia (Pennsylvania), Houston (Texas), and Milwaukee (Wisconsin). Men who have sex with men (MSM) were recruited at gay bars, high-risk heterosexuals (HRHs) were recruited at sexually transmitted disease (STD) clinics, and injection drug users (IDUs) were recruited through street outreach or at needle exchange programs (NEPs). For each state, the largest city and other key metropolitan areas were included. For each venue type (bar, clinic, street/NEP), specific sites were identified through formative research, which included review of reports, such as local HIV/AIDS surveillance reports (“secondary data review”), key informant interviews, and observations at some of the potential interview sites. Site selection by project staff was based on the feasibility of conducting interviews at each of the locations and on criteria for obtaining a diverse sample of each risk group.

Persons at the venues were eligible to participate in HITS if they were at least 18 years of age, had been a resident of the state for at least 6 months, and gave informed consent. Further details of selection and sampling processes within venues have been described elsewhere [1]. After eligibility was assessed and informed consent obtained, participants were administered a face-to-face interview by trained study personnel. No personal identifiers were collected. This study was reviewed by institutional review boards at CDC and in participating areas. For each project area, the intended sample size was 100 each of MSM, HRHs, and IDUs. In addition, sites attempted to recruit approximately equal numbers of male and female heterosexual adults from STD clinics; there were no requirements for gender distribution of IDUs. Of persons approached and determined to be eligible, 3127 (83%) completed an interview: 1185 (86%) MSM, 1140 (91%) HRHs, and 802 (72%) IDUs. Of the total number of interviews, 2 were missing age, 21 (1%) were missing residence information, and 5 were missing sex: all 28 were excluded from analysis.

Behaviors reported during the survey were used as selection criteria for analysis. During the 12 months before interview, MSM must have had sex with a man, HRHs must have been sexually active only with members of the opposite sex, and IDUs must have injected drugs. Excluded from analysis were 602 (15%) persons who completed an interview but did not report the behaviors used as selection criteria. Of MSM interviewed in bars, 186 (13%) had not had sex with a man during the past year. Of the HRHs interviewed in STD clinics, 104 (8%) reported that they had not had heterosexual sex or that they had sex with a same-sex partner. Of the IDUs recruited on the street or at NEPs, 312 (25%) reported that they had not injected drugs during the past year.

For this report, we used several additional criteria for exclusion from analysis. Although 20 transgender persons were interviewed, they were excluded from analysis because they were not consistently asked the questions about sexual risk behavior. All persons who reported being HIV infected were excluded from analysis (n = 161, 5%), as were those without data on HIV testing (n = 23, 1%) and those who never received their HIV test results (n = 111, 4%). Because of a lack of appropriate interviewers, Site C did not conduct the component for IDUs. Site D was unable to collect data from an adequate number of IDUs for the purposes of this report.

As all participants were administered the same questionnaire, information about risk behaviors other than those pertaining to the population recruited (e.g., sex with men among male IDUs, injection drug use among MSM and HRHs) are available. However, we present risk behavior data by venue because we used venue-based sampling as a means of reaching persons who engaged in a specific high-risk behavior (e.g., injection drug use only for persons recruited at street/NEP venues).

The findings in this report are subject to several limitations. Data stratification in some instances may produce numbers in each category that are too small for reliable inferences. The study was not population based but was designed to enroll equal proportions of each of 3 groups recruited from specific venues; thus, it may not represent all at-risk populations or their distribution in the general population. Findings from the states or cities in this study may not be generalizable to all other states or cities. Because the survey was administered by an interviewer, some respondents may not have reported their behavior accurately. For example, some respondents may not have reported a less socially desirable behavior in which they were engaging (e.g., sharing needles) or may have reported a more socially desirable behavior that they did not engage in (e.g., using a condom during intercourse).

Reference

• Hecht FM, Chesney M, Lehman JS, et al. Does HIV reporting by name deter testing? AIDS 2000;14:1801-1808.