Abstract
Various diagnostic methods exist for the detection of Chlamydia psittaci. In the current study, the test performance of polymerase chain reaction (PCR) was compared with other testing methods used in the diagnosis of C. psittaci. Tissue and fecal specimens (n = 119) of avian and mammalian origin were tested by PCR and one or more of the following methods: cell culture, enzyme-linked immunosorbent assay, and direct fluorescein-conjugated monoclonal antibody staining. Several gold standards, based on results of testing methods other than PCR, were used to calculate the following test performance characteristics of PCR: sensitivity and specificity, with their 95% confidence intervals; kappa statistics, a measure of intertest agreement; and lambda statistics, a chance-corrected estimate of the sensitivity and specificity. Overall, the test performance characteristics of PCR were low compared with the other testing methods. Possible reasons for the poor test performance of PCR in the current study include destruction of the organisms during storage, interference with the PCR by other reagents, or technical errors.