Background: Serosorting is the practice of preferentially having sex with partners of concordant HIV status or of selectively using condoms with HIV-discordant partners.
Methods: We evaluated the epidemiology of serosorting among men who have sex with men (MSM) seen in a sexually transmitted disease clinic, Seattle, WA, 2001-2007, and defined the percentage of visits during which MSM tested HIV positive based on whether they reported nonconcordant unprotected anal intercourse (UAI), UAI only with partners thought to be HIV negative (serosorters), no UAI, or no anal intercourse.
Results: Men reported serosorting during 3295 (26%) of 12,449 visits. From 2001 to 2007, the proportion of visits during which men reported serosorting increased (P = 0.02); this change was greater among HIV-infected MSM than among HIV-uninfected MSM. Among men who tested HIV negative in the preceding year, HIV tests were positive in 49 (3.5%) of 1386 who reported nonconcordant UAI, 40 (2.6%) of 1526 serosorters, 28 (1.5%) of 1827 who had only protected anal intercourse, and 0 of 410 who had no anal intercourse (P < 0.0001); 32% of new HIV infections occurred in serosorters. The prevalence of HIV was higher among serosorters tested during 2004-2007 than among those tested during 2001-2003 (0.85% vs. 3.2%, P = 0.03).
Conclusions: Serosorting offers MSM limited protection from HIV.