September 30, 2002
Over $100 million is directed annually to state and local health departments to support HIV C&T sites. While such programs were created so that individuals could learn whether they were infected with HIV, an increasing proportion of tests are being performed on persons who have previously tested HIV negative. With such a substantial proportion of testing dollars going towards retesting of individuals, it is important to understand whether persons are retesting because they are continuing to have high-risk sexual practices or whether they are retesting to assure themselves that they have not seroconverted as a result of low-risk sexual practices, such as oral intercourse.
Repeat testers have been found to be more likely to have both high-risk and low-risk sexual practices. Better characterization of risk behavior of repeat testers is necessary to assess the impact of prior HIV C&T on subsequent risk behavior. In this study, the authors use data from San Francisco's publicly funded HIV C&T sites to estimate the seroincidence of HIV infection among repeat testers stratified by risk behavior; examine the association between risk level and repeat testing rates; and assess the association between tester characteristics and level of repeat testing.
From January 1995 through June 1997, a total of 57,505 publicly funded HIV tests were performed in anonymous and confidential testing sites in San Francisco. Of the 57,505 tests done during the 2.5-year study period, 33,770 (58.7 percent) were identified as repeat testers, having reported at least one prior HIV test. In total, 22,431 (66 percent) observations met inclusion criteria for repeat testers.
Overall, the HIV incidence was 1.3 infections per 100 person-years (py), with 4.6 infections/100 py among transgender testers, and 1.6 infections per 100 py among female testers. The very high-risk and high-risk levels include behaviors that are reported by MSM and TSX persons only, with incident infection in these levels significantly higher than in the low-risk level. During the six months between January and June of 1997, 8,785 individuals were tested at publicly funded sites. Based on self-reported testing history, these individuals had collectively accrued a total of 25,950 tests, since the initiation of testing in 1985, for a mean of 3.0 tests/person and a repeater rate of 66 percent. The greatest number of tests/person occurred in the high-risk level, at 5.3 tests per person and a repeater rate of 92 percent. This was followed by those in the very high-risk level with 5.0 tests/person and a repeater rate of 86 percent. In addition, the authors looked at the high level repeaters with 10 or more tests. Seventeen percent of high-risk level testers and 16 percent of very high-risk level testers had been tested at least 10 times. Moderate-risk level testers were older (mean age, 35.5 years) than low-risk testers (mean age, 29.6 years). In bivariate analyses, higher risk level, male gender, advancing age, increasing numbers of sex partners within the last 12 months and white race were all significantly associated with repeat testing. These associations held regardless of the cutoff number of repeat tests chosen.
In this retrospective analysis of repeat HIV testing at publicly funded C&T sites in San Francisco, the authors found that the majority of those who report testing more than once tend to be MSM between ages 25 and 34. Further, those with the highest rate of repeat testing are also those who have the highest incidence of HIV: gay/bisexual men who inject drugs, report both unprotected and protected receptive anal intercourse, and/or do not specify their sexual practices.
Journal of Acquired Immune Deficiency Syndromes
09.01.02; Vol. 31; No. 1: P. 63-70; Susan E. Fernyak; Kimberly Page-Shafer; Timothy A. Kellogg; William McFarland; Mitchell H. Katz