Can cord formation in BACTEC MGIT 960 medium be used as a presumptive method for identification of M. tuberculosis complex?

Serpentine cord formation in BACTEC MGIT 960 medium was evaluated as a rapid method for the presumptive identification of M. tuberculosis complex (MTBC). Total 2527 samples were processed for AFB culture using MGIT 960 TB system over a period of three months. AFB smears were prepared from 1000 MGIT tubes flagged positive by the MGIT instrument and stained by ZN method to examine presence or absence of serpentine cording. The cord formation was compared with PNBA [p-nitro benzoic acid] test on MGIT system and all controversial cases were further evaluated by NAP [p-nitro-a-acetylamino-phydroxypropiophenone] test on BACTEC 460 TB system. Of the 1000 culture positives, 904 (90.4%) were identified as mycobacteria, of which 869 (96%) showed cording by smear microscopy. One (0.1%) was identified as nocardia. In the remaining 95 (9.5%) cases, primary smear made from MGIT vial was negative. Of 869 cultures showing serpentine cord formation, 842 were confirmed as MTBC and 27 as NTM by PNBA assay on MGIT 960 TB system. The sensitivity, specificity, positive and negative predictive values are found to be 99.6%, 54%, 96% and 91% respectively. An average detection time for PNBA assay was found to be eight days whereas cording results were available on the same day of culture positivity. Though highly sensitive it is not very specific and hence cannot be the only test for presumptive diagnosis of MTBC.