Evaluation of risk score algorithms for detection of chlamydial and gonococcal infections in an emergency department setting

Objectives: To develop and evaluate screening algorithms to predict current chlamydial and gonococcal infections in emergency department (ED) settings and assess their performance. Methods: Between 2002 and 2005, adult patients aged 18 to 35 years attending an urban ED were screened for Chlamydia trachomatis (Ct) and Neisseria gonorrhoeae (GC) and completed a brief demographic and behavioral questionnaire. Using multiple unconditional logistic regressions, the authors developed four separate predictive models and applicable clinical risk scores to screen for infection. They developed models for females and males separately, for Ct and GC infections combined, and for Ct infection alone. The sensitivities and specificities of the clinical risk scores at different cutoffs were used to examine performance of the algorithms. Results: Among 5,537 patients successfully screened for Ct and GC, the overall prevalence of infection was 9.6%. Age was the strongest predictor of infection. Adjusting for other predictors, the prevalence odds ratio (POR) was 2.2 (95% confidence interval [CI] = 1.7 to 2.8) for Ct and GC combined and 2.9 (95% CI = 2.1 to 4.1) for Ct alone comparing females 25 years and younger to females older than 25 years. Among males, the association was stronger with an adjusted POR of 3.3 (95% CI = 2.3 to 4.7) for Ct and GC combined and 3.2 (95% CI = 2.1 to 4.7) for Ct infection alone. Conclusions: If the decision to incorporate Ct and GC screening into routine ED care is made, age alone appears to be a sufficient screening criterion.