Polymerase chain reaction test in induced sputum of patients with pulmonary tuberculosis.

INTRODUCTION: Induced sputum (IS) is an alternative method of obtaining sputum, but IS smears are frequently negative. Culture is more time consuming in its results, and less useful to guide the diagnosis. Polymerase chain reaction (PCR) is the most common methodology for rapid diagnosis of tuberculosis (TB), and few studies evaluated its role in IS samples.

OBJECTIVES: The objective of this study is to determine the diagnostic yield of PCR for TB compared with culture in IS samples.

MATERIALS AND METHODS: Prospective study. Inpatients and outpatients of >18 years with respiratory symptoms suggestive of PTB were invited to participate. The subjects were interviewed using a standardized questionnaire, and collected IS. Three samples were obtained for AFB smear and culture. A fourth sample was obtained for PCR test.

RESULTS: A total of 116 IS samples were evaluated. The sensitivity, specificity, positive predictive value and negative predictive values of PCR were 95.2%, 48.4%, 29.0% and 97.9%, respectively. The area under the receiver operating characteristic curve was .72 for the PCR test (P < .0001).

CONCLUSIONS: Although the PCR specificity could be underestimated, if we consider PCR to be more sensitive than the culture method used, we believed that these PCR-positive tests mean false positives. The results of PCR should always be interpreted carefully in conjunction with clinical information.