RESUMEN
The objective of this study was to compare the diagnostic yield of the Kudoh-Ogawa (K-O) swab method for the culturing of Mycobacterium tuberculosis from clinical samples with the standard Petroff-Lowenstein-Jensen (P-LJ) procedure. A total of 2,287 sputum samples and 685 extrapulmonary clinical specimens were processed with both decontamination methods and compared for M. tuberculosis detection rate, recovery of M. tuberculosis colonies, and culture contamination. Overall, 23.9% and 23.5% of the samples, processed with, respectively, the K-O swab method and the P-LJ procedure, yielded M. tuberculosis after 8 weeks of incubation. The K-O swab method and the P-LJ procedure provided comparable diagnostic yields for extrapulmonary clinical specimens (P = 0.688), but the K-O method showed a slightly but statistically significantly higher diagnostic yield for pulmonary samples (P = 0.002). No significant difference for culture contamination or colony recovery was found for either method. The turnaround time for the isolation of M. tuberculosis was significantly shorter for the K-O swab method, with 77% of the M. tuberculosis cultures being positive within 3 weeks of incubation, and only 6.1% positivity for the P-LJ method. Concerning the workload, the K-O swab method needs a minimum sample manipulation and takes less than 4 min per sample, as the samples are not centrifuged in this procedure. The K-O swab method is an efficient and fast (in terms of sample processing and culture growth) alternative for culturing M. tuberculosis from either pulmonary or extrapulmonary clinical specimens. The method is particularly suitable for laboratories with a high workload and for laboratories lacking a special infrastructure.