Purpose: To study the number of sputum cultures by MGIT system (Becton-Dickinson) needed to diagnose pulmonary tuberculosis.
Object and method: Prospective study of all patients who visited our hospital, and were strongly suspected of pulmonary tuberculosis during the period from Jan. 2002 to Sept. 2003. In these patients, 3 consecutive sputum cultures were done by both MGIT-system and egg-based Ogawa medium (1 slant).
Results: Altogether 290 cases of sputum-culture positive pulmonary tuberculosis were available for analysis. In 210 first-sputum-smear positive cases, incremental yield of 3rd sputum culture in 3 consecutive MGIT cultures was equal to or less than 1.0% and 98.1% of culture positive cases were detected by 2 consecutive MGIT cultures. In 80 first-sputum-smear negative cases, incremental yield of 3rd sputum culture in 3 consecutive MGIT cultures was equal to or more than 5.0%, and 90.0% of culture positive cases were detected by 2 consecutive MGIT cultures. This detection rate was almost the same as the calculated detection rate (91.4%) by 3 consecutive Ogawa (2 slant) cultures (previous standard method).
Conclusion: It was suggested that 2 consecutive sputum cultures by MGIT were sufficient to detect M. tuberculosis in first-sputum-smear positive cases, but 3 consecutive sputum cultures by MGIT were relatively useful in first-sputum-smear negative cases.