by study participants allowed these researchers to examine predictors of needle exchange participation and to estimate HIV seroconversion rates among needle exchange users and nonusers. According to these authors, this information allowed them to "assess whether the San Francisco needle exchanges are reaching their intended clientele and whether the exchanges affect risk reduction and HIV incidence" (p. 3).
This study recruited 1,093 injection drug users from nine methadone maintenance and 21-day detoxification programs from 1989 through 1990, following the November 1988 opening of the Prevention Point needle exchange program in San Francisco. Each participant completed structured interviews (questionnaire) and had an HIV test performed at each repeat visit. Of those recruited, repeated questionnaire data and serostatus test results were available from 412 participants (38 percent). Needle exchange participation was determined by examining participants' questionnaire responses (needle exchange participation items) for their most recent treatment visit (methadone maintenance, 21-day detoxification, or both).
Descriptive analyses based on the entire sample of 1,093 study participants showed that males, frequent injectors, homeless people, residents of hotels and shelters, homosexuals and bisexuals, and participants who were aware of their HIV status were more likely to use the needle exchange. No association was detected between needle exchange use and the education level of study participants, the age of first injection, the number of years injecting, bleach use, or prostitution.
During the 1989 through 1990 study period, nine seroconversions were detected among the 412 repeaters who were initially assessed as HIV negative. The time at risk for study participants who remained HIV negative was years from first interview to last interview, whereas the time at risk measure for those who seroconverted was years from first interview to first HIV-positive interview. This resulted in an estimated overall seroconversion rate of 1.1 per 100 person years (95 percent confidence interval [CI]: 0.5 to 2.0 percent per person years) among the 412 repeaters. The estimated seroconversion rates for those who had never used the needle exchange was 0.33 percent per person years (95 percent CI: 0.05 to 1.02 percent per person years) compared with an estimated rate of 3.49 percent per person years (95 percent CI: 1.49 to 6.75 percent per person years) for those who had ever used needle exchange. More specific estimates of seroconversion rates among needle exchange users were derived for those study participants who had used the needle exchange for fewer than 3 months (6.26 percent per person years) and for those who had used the needle exchange for 3 months or more (2.19 percent per person years).
Results of a proportional hazard model indicated that, among the 412 repeat study participants, the hazards ratio for seroconverting was higher for those who had used the needle exchange (10.60; 95 percent CI: 2.20 to 51.06) compared with those who had not.