To determine the accuracy of the Bactec system in the diagnosis of Tuberculous Pleural Effusion compared to histopathology and clinical course.
All patients in the Chest Clinic Registry with Pleural Effusion confirmed by x-ray and or ultrasound between January, 1995 to December, 2003 were studied. All subjects had clinical evaluation, Mantoux test, thoracentecis/pleural biopsy, effusion Kinyoun AFB smear, pleural tissue Bactec culture and histopathology. Laboratory and pathologist had no prior knowledge of actual diagnosis. Patients were assessed independently based on a modified ATS Classification for TB Disease status and followed-up by same attending pulmonologist.
Data of 175 of 185 registered patients were available for analysis. At 49.2% TB prevalence, Histopathology has 77.5% sensitivity, 98.9% specificity, 69.8% positive predictive value, 81.7% negative predictive value, 9.8 Likelihood ratio (+), 0.23 likelihood ratio (-) and 88.2% overall accuracy while Bactec has 30.9% sensitivity, 93.6% specificity, 83.3% positive predictive value, 55.6% negative predictive value, 4.8 likelihood ratio (+), 0.74 likelihood ratio (-) and 61.6% overall accuracy.
Using a single pleural biopsy specimen, histopathology appears more accurate and faster than Bactec culture in the diagnosis of TBPE. There is no significant morbidity with the Cope Needle pleural biopsy.
The etiology of Pleural Effusion is always a clinical dilema. Pleural biopsy with culture for M.TB often facilitates the diagnosis and treatment in a high prevalece setting.
Bernardo Briones, Grant monies (from sources other than industry) INCLEN Thesis grant on the original study of the serodiagnosis of PTB.