Objectives Partner concurrency is associated with STI acquisition, even when partners are sequential (ie, serial monogamy). However, little is known about the timing of STI relative to partner change. This is due, in part, to lack of documentation of STI before and after partner change. Using prospective behavioural and STI screening data, we compared the proportion of an STI occurring before and after partner change within a 12-week period.

Methods As part of a larger study of risk and protective factors for STI in a mid-sized US city, young women provided weekly vaginal swabs and completed daily diaries of sexual behaviours for a 12-week period twice a year for up to 8 years. Vaginal swabs were tested for chlamydia (CT), gonorrhoea (NG) and trichomonas (TV) using amplified DNA-based tests. We limited analysis to the 12-week periods in which young women reported one sex partner change (ie, 2 sequential partners, no overlap). STIs with partner 1 were defined as those diagnosed before 1st sex with partner 2; STIs with partner 2 occurred after 1st sex with partner 2. Published data shows a higher odds of infection after a partner change (compared to no change) with the ORs ranging from 2.0 to over 4.0. We used a non-equivalency test to identify the odds for not detecting differences in STI acquired with partner 1 vs partner 2. If there is a difference between rates for partner 1 and partner 2, it will not be higher than the upper bound of the CI.

Results Ninety-two women provided 111 12-week periods with one partner change. Mean age was 17.8 years; 94% were African American. Rates of STI were high for both Partner 1 and Partner 2 (see Abstract P1-S5.33 table 1). For CT, TV, and any STI, the upper bound of the 95% CI was lower previously reported rates. We have 95% confidence that the OR for infection with partner 2 vs partner one is not greater than 1.83 for CT, 1.47 for TV, and 1.68 for any STI. In contrast, the true OR for GC may fall within or higher than previously reported. We have 95% confidence that the OR for GC infection with partner 2 vs partner 1 is between 0 and 9.7.

Conclusion Many young women already have an STI when a partner change occurs. These data suggest that the relationship contexts of partner change–in addition to the risk characteristics of a new partner–are also relevant to the epidemiology of STI in a specific common form of partner concurrency (ie, serial monogamy).