Follow-up Syphilis Testing Behavior After Diagnosis with Early Syphilis Among Men Who Have Sex with Men—San Francisco, 2005–2008

Background: In San Francisco (SF), men who have sex with men (MSM) diagnosed with syphilis are at high risk of reinfection within one year. SF Department of Public Health recommends regular syphilis serologic follow-up testing for 12 months following diagnosis.

Objectives: To develop and validate a prediction model to identify MSM unlikely to test for syphilis within six months following diagnosis.

Methods: Surveillance data from 2005–2008 on reactive serologic tests were used to assess follow-up syphilis testing behavior among MSM with early syphilis in SF with nontreponemal titer ≥1:16 at diagnosis. Using multivariate analysis, we analyzed data from contact-tracing interviews to identify predictors of not having a syphilis test within six months following diagnosis. We developed a model in a training sample (2005–2006) and applied it to a validation sample (2007–2008), calculating area under the receiver operating characteristic curve (AUC) to assess model performance.

Results: Among 795 persons, 202 (25.4%) did not have a syphilis test within six months following diagnosis. Predictors of not testing within six months following diagnosis were being diagnosed at the municipal STD clinic, being HIV-uninfected, and not residing in a largely gay male neighborhood. In the validation sample, 14.6% had all 3 characteristics; of those, 47.5% did not have a syphilis test within six months after diagnosis (27.6% of all non-testers). The model had modest accuracy in the training sample (AUC, 0.656) and the validation sample (AUC, 0.657).

Conclusions: This model can identify MSM unlikely to test for syphilis within six months following diagnosis. Compliance with follow-up syphilis testing can improve epidemiologic surveillance of syphilis as well as the clinical management of cases and reinfection.

Implications for Programs, Policy, and/or Research: Prediction models can help target syphilis prevention and control activities to persons who are at risk of experiencing syphilis and unlikely to test for syphilis within six months following diagnosis.