BackgroundBacteriological confirmed active case detection remains the corner stone for diagnosing tuberculosis. Non-radiometric liquid culture system Mycobacterium Growth Indicator Tube with automated interface had been recommended by expert groups in addition to conventional solid culture media such as Lowenstein–Jensen. However in high burden resource limited countries advanced non-radiometric based tuberculosis diagnostic methods such as MGIT 960 is limited. Therefore we have evaluated the performance of MGIT 960 system compared to LJ for recovery of Mycobacterium complex (MTBC) from clinical specimens.MethodsA cross sectional study was conducted from a total of 908 samples between January 1st, 2013 to December 31st, 2014. Clinical specimens were processed following standard procedures and the final suspension was inoculated to MGIT tubes and LJ slant. Identification and confirmation of MTBC was done by ZN staining and SD Bioline test. Data was analyzed by SPSS version 20. The sensitivity, specificity, recovery rate and the average turnaround time to recover the organism was computed.ResultsFrom a total of 908 clinical specimens processed using both LJ and BACTEC MGIT liquid culture methods the recovery rate for LJ and MGIT, for smear positive samples was 66.7% (74/111) and 87.4% (97/ 111) respectively while for smear negative samples was 13.4% (108/797) and 17.4% (139/797) for LJ and MGIT methods respectively. The overall recovery rate for MGIT is significantly higher than LJ methods [26% (236/908; vs. 20%, 182/908, P = 0.002)]. The average turnaround time for smear positive samples was 16 and 31 days for MGIT and LJ respectively. Turnaround time for smear negative samples was 20 and 36 days for MGIT and LJ respectively. The overall agreement between MGIT and LJ was fairly good with Kappa value of 0.59 (P < 0.001). In the present study the contamination rate for MGIT is higher than the LJ methods, 15 and 9.3% respectively.ConclusionsThe BACTEC MGIT liquid culture system has better MTBC recovery rate with shorter turnaround time for both smear positive and negative clinical specimens compared to Conventional LJ method. However, efforts should be made in order to reduce the high contamination rate in BACTEC MGIT system and to lesser extent to LJ methods.