The Population-Level Impact of Doxycycline Postexposure Prophylaxis on Syphilis in King County, WA: An Interrupted Time Series Analysis

BACKGROUND: An initial evaluation found that doxycycline postexposure prophylaxis (doxy-PEP) has a population-level effect on syphilis among cisgender men and transgender women who have sex with men. Whether doxy-PEP's impact is generalizable across the United States and will affect syphilis incidence among cisgender women and pregnant people-populations who seldom receive doxy-PEP-and congenital syphilis (CS) is unknown. METHODS: We used public health surveillance data from January 2017 to June 2025 to conduct an interrupted time series analysis of monthly syphilis diagnoses (all stages) overall and by gender before and after doxy-PEP implementation in March 2023 in King County, WA. We estimated the observed and counterfactual number of syphilis cases postimplementation to calculate absolute and relative differences in cases. RESULTS: In the postimplementation period, the absolute difference in total syphilis cases was -3031 (95% CI: -4295, -1768) with a relative difference of -52.3% (95% confidence interval [CI]: -74.1%, -30.5%). Among cisgender men, the absolute difference in cases was -2248 (95% CI: -3321, -1175) for a relative difference of -53.1% (95% CI: -78.5%, -27.8%) in the postimplementation period. Cisgender women experienced an absolute difference of -755 cases (95% CI: -939, -572) and a relative difference of -46.9% (95% CI: -58.3%, -35.5%) after doxy-PEP implementation. Among transgender and nonbinary people, the absolute difference in syphilis cases was -56 (95% CI: -90, -23), a relative difference of -33.1% (95% CI: -53.3%, -13.6%), in the postimplementation period. Syphilis cases among pregnant persons rose from 13 in 2017 to 48 in 2024, while CS cases increased from 0 in 2017 to 23 in 2024. CONCLUSIONS: After the implementation of doxy-PEP, syphilis diagnoses declined dramatically, including among cisgender women. This decline has not affected syphilis in pregnant persons or CS.