Differences in the temporal trends of HIV seroincidence and seroprevalence among sexually transmitted disease clinic patients, 1989-1998: application of the serologic testing algorithm for recent HIV seroconversion

Abstract

The authors compared temporal trends in the prevalence and incidence of human immunodeficiency virus (HIV) infection based upon 34,866 specimens from patients who attended the San Francisco, California, municipal sexually transmitted disease clinic between 1989 and 1998. HIV infection data were collected during annual blinded HIV serologic surveys. Incidence was determined by applying a serologic testing algorithm for recent HIV seroconversion that uses both a sensitive and a less sensitive enzyme immunoassay to stored HIV positive sera. The HIV seroprevalence declined from 15.2% in 1989 to 7.2% in 1998 (odds ratio per year = 0.92, 95% confidence interval (CI): 0.91, 0.94). Among homosexual men, the HIV prevalence declined from 50.9% in 1989 to 19.9% in 1998 (odds ratio per year = 0.86, 95% CI: 0.85, 0.88). The pooled seroincidence was 1.6% and did not change significantly over time (odds ratio per year = 1.0, 95% CI: 0.98, 1.1). The pooled seroincidence among homosexual men was 6.6% per year and remained steady between 1989 and 1998 (odds ratio per year = 0.99, 95% CI: 0.92, 1.1). During a dramatic, 10-year decline in seroprevalence of HIV infection, the incidence of HIV infection remained remarkably stable.