Background. Polymerase chain reaction (PCR) techniques have revolutionized the diagnosis of tuberculosis (TB). PCR has significantly improved the sensitivity and specificity of existing diagnostic methods. In this study, we report our experience using a modified IS611O-based nested PCR assay for rapid diagnosis of pulmonary TB. Methods. A total of 327 respiratory specimens from 275 patients suspected of having pulmonary TB at Taipei Veterans General Hospital were tested using the nested PCR assay, acid-fast smear and culture for the presence of Mycobacterium tuberculosis complex (MTB). Nested PCR was performed with IS611O-based primers specific for MTB. We reviewed the medical records of patients and analyzed the clinical features. The PCR results were compared with the final clinical diagnosis. Results. We identified MTB in 167 of 327 samples by the nested PCR assay. No non-tuberculous Mycobacterium (NTM) was identified among the clinical samples. Diagnosis by PCR took about 6 hours in this study. The sensitivity and specificity compared with culture were 94.7% and 100%, respectively for the smear-positive, culture-positive samples, and 76.7% and 98.6% for the smear-negative, culture-positive samples. The overall sensitivity, specificity, positive and negative predictive values, compared with culture results, were 91.7%, 98.6%, 98.8% and 90.6%, respectively. Two specimens positive by PCR and negative by culture were taken from patients on anti-TB drug therapy. These specimens were culture-positive before anti-TB drug therapy. After resolution of the discrepancies by studying the patients' clinical data, both specificity and positive predictive value reached 100%. Conclusions. The results indicated that this in-house nested PCR assay is a rapid and sensitive method for diagnosing pulmonary TB. It is also good for excluding infections caused by NTM.
|Number of pages||6|
|Journal||Chinese Medical Journal (Taipei)|
|Publication status||Published - 2000|
- Polymerase chain reaction
- Pulmonary tuberculosis
ASJC Scopus subject areas