RESUMO
OBJECTIVE: To determine test characteristics of provider judgment for empiric antibiotic provision to patients undergoing testing for a sexually transmitted infection. STUDY DESIGN: We conducted a retrospective cross-sectional electronic health record review of all patients aged 13-19 years who had Neisseria gonorrhoeae (GC) and Chlamydia trachomatis (CT) testing sent from an urban, academic pediatric emergency department in 2012. We abstracted data, including patient demographics, chief complaint, sexually transmitted infection test results, and treatment. We calculated test characteristics comparing clinician judgment for presumptive treatment for a sexually transmitted infection with the reference standard of the actual results of testing for a sexually transmitted infection. RESULTS: Of 1223 patient visits meeting inclusion criteria, 284 (23.2%) had a positive GC and/or CT test result. Empiric treatment was provided in 615 encounters (50.3%). Provider judgment for presumptive treatment had an overall sensitivity of 67.6% (95% CI, 61.8-73.0) and a specificity of 55% (95% CI, 51.7-58.2) for accurate GC and/or CT detection. CONCLUSIONS: Many adolescents tested for GC and CT receive empiric treatment at the initial emergency department visit. Provider judgment may lack sufficient sensitivity and specificity for identifying infected patients, resulting in the potential for undertreatment of true disease, overtreatment of uninfected patients, or both.