Background: Since chlamydial cervicitis is not associated with specific complaints, screening asymptomatic women is an important initiative to prevent pelvic inflammatory disease and its sequelae. Compared with universal screening, selective screening is less costly but less effective so the cost savings vs the consequences of missing infected women need to be weighed carefully.
Methods: In two family planning clinics, 1002 women were surveyed for chlamydial infection (prevalence, 7%) and its predictors to determine whether universal or selective screening is the most efficient strategy. Two rules for the selection of patients were determined by logistic regression modeling and their relative efficiencies were compared by incremental cost-effectiveness and sensitivity analysis. The validity of the screening rules was tested in 191 students attending a university health clinic.
Results: If those with cervical friability, suspicious discharge, urinary frequency, or intermenstrual bleeding had been tested, 55.3% of all women would have been screened and 83.3% of all cases would have been detected. If those reporting a new sex partner in the preceding year had also been tested, 75.4% would have been screened, identifying 93.3% of all cases. The predictive power and practicality of the selection rules were validated in the university health clinic sample. Sensitivity analyses showed selective screening using cervical enzyme immunoassay with blocking confirmation was efficient if the prevalence of chlamydial infection was 16% or less, 11% or less, or 5% or less depending on whether base analyses, overestimated costs, or worst performance scenarios, respectively, were used.
Conclusions: Selective screening based on four or five predictors and confirmed cervical enzyme immunoassay is an effective and efficient strategy in low prevalence settings.