[Bronchoalveolar lavage in the diagnosis of pulmonary tuberculosis].

In the January 1991 issue of Chest, Baughman et al’ reported the utility of bronchoscopy with bronchoalveolar lavage (BAL) in the diagnosis of Mycobacterium tuberculosis infections, and they concluded that BAL is useful for this indication. However, the M tuberculosis infection was diagnosed with only BAL in one of 29 cases, when BAL and bronchial wash specimens were obtained together. In a prospective study previously reported by our group,2 we performed fiberoptic bronchoscopy with BAL and bronchial wash and obtained postbronchoscopy sputum specimens in 20 of 222 patients suspected of having pulmonary tuberculosis (all patients had three consecutive early morning sputum specimens or gastric aspiration smears that were negative for acid-fast bacilli, and seven also had negative Lowenstein cultures). The BAL fluid specimens provided the highest yield for diagnosis of pulmonary tuberculosis (15/17 [88 percent]); in seven of the cases, the BAL fluid specimen was the only positive source. Bronchial wash was positive in only nine (53 percent) patients. Bronchial washes were performed immediately after inspection of the tracheobronchial tree, and the specimens obtained were collected for study before BAL fluids were obtained. The differences between the results of our study and those of the study by Baughman et al could be due to different methodology. In the latter study the bronchial wash included the aspirated fluid obtained after BAL; the diagnostic value of that fluid should be attributed to BAL, not to bronchial wash. Because of economic considerations, we flOW collect bronchial wash specimens and postBAL aspirated fluid in the same receptacle. Thus, we obtain fewer diagnoses with BAL exclusively than when we performed bronchial wash without aspirating fluid after BAL. In our experience, the diagnostic yield of bronchoscopy with BAL for M tuberculosis infection is significantly greater than that for bronchial wash (odds ratio = 6.67; confidence interval, 1.253 to 35.45). Therefore, BAL, fluid aspiration after BAL, and bronchial wash should be performed and a postbronchoscopy sputum specimen should beobtainedwhen bronchoscopy is performed in patients suspected of having pulmonary tuberculosis, especially if previous sputum smears were negative. Recently, the American Thoracic Society reported the utility of BAL for diagnosis of pulmonary tuberculosis.