ABSTRACT
We evaluated a low-cost diagnostic strategy for detecting Chlamydia trachomatis in a low-prevalence population. We used an amplified enzyme immunoassay (EIA) with a reduced-cutoff "negative gray zone" to identify reactive specimens for confirmation by a nucleic acid amplification test. As part of the Chlamydia Screening Studies project, men provided a first-pass urine specimen, which they returned by post for testing. We tested 1,003 specimens by IDEIA PCE EIA (Dako) and Cobas PCR (Roche). There were 32 (3.2%) true positive specimens according to a combined standard using an algorithm requiring concordant results from at least two independent tests. All of these were positive by Cobas PCR and 24 were confirmed to be positive by PCE EIA, including 2 that gave results in the negative gray zone. There were 971 true negative specimens, 2 of which were positive by Cobas PCR and 19 of which were initially inhibitory for PCR. The relative sensitivity, specificity, positive predictive value, and negative predictive value of PCE EIA with PCR confirmation were 75.0% (95% confidence interval [CI], 56.6 to 88.5%), 100% (95% CI, 99.7 to 100%), 100% (95% CI, 88.3 to 100%), and 99.2% (95% CI, 98.4 to 99.6%), respectively. The corresponding values for Cobas PCR were 100% (95% CI, 89.1 to 100%), 99.8% (95% CI, 99.3 to 100%), 94.1% (95% CI, 76.9 to 98.2%), and 100% (95% CI, 99.6 to 100%), respectively, with 1.9% (19/1003) of the samples being initially indeterminate. When the prevalence of C. trachomatis is low, the use of an amplified EIA on urine specimens, with confirmation of results in the negative gray zone by use of a nucleic acid amplification technique, is not suitable for screening asymptomatic men. In addition, positive nucleic acid amplification test results should be confirmed and an inhibition control should be used.