Comparison of sputum induction with broncho-alveolar lavage in the diagnosis of smear-negative pulmonary tuberculosis.

Sputum microscopy and AFB-culture being gold standard and a fundamental tool for diagnosis of pulmonary tuberculosis (PTB) has got its limitation of low sensitivity. Fibreoptic bronchoscopy (FOB) has been widely recommended as the diagnostic procedure of choice in smear negative patients. But bronchoscopy is an invasive procedure, costly, not readily available in our country and needs expertise. Several studies abroad have directly compared the yield of sputum induction (SI) with 3% saline (NaCl solution) with Bronchoalveolar lavage (BAL) through FOB in smear-negative suspected PTB patients and showed that SI was a low cost, safe and well tolerated procedure with equal efficacy to BAL through FOB for the diagnosis of PTB in such patients. For the first time a prospective comparison was conducted in Bangladesh to see the yield of sputum induction (SI) and BAL in 52 selected smear- negative patients of suspected PTB. Each of the samples of induced sputum and BAL fluid were examined for AFB by Ziehl-Neelsen's method. Samples of both SI and BAL from 20 patients were cultured for AFB in Lowenstein-Jensen medium for 6 weeks irrespective of their induced sputum smear being positive or negative for AFB. Data were managed and analyzed using computer program SPSS version 10.0. Agreement of SI and BAL was tested using Pearson Chi-square and Kappa test. The results showed that the yield of SI were significantly more than that of BAL (p<0.05).The AFB smear results from specimens obtained by SI and BAL were in agreement in 75% cases (p=0.02).Statistical analysis of the yield of culture results from SI and BAL group with Fishers Exact test showed they were in agreement in 90% cases (p=0.0001) and was measured by Kappa test as significant (p=0.0004). The sensitivity of AFB-smears in samples from SI and BAL were 74% and 58% respectively. The specificity of smear positivity and of culture was assumed to be 100%. SI is a safe procedure with considerable diagnostic yield and a high agreement with the results of BAL through FOB for the diagnosis of PTB. SI offers an alternative or additional approach to the diagnosis of smear-negative suspected PTB patients and would enhance sensitivity for the diagnosis of tuberculosis.