Evaluation the practicability of Cord formation of MGIT positive smears for Presumptive Identification of Mycobacterium tuberculosis Shu-Fen Jan, Jenn-fuh Leu, Pen-Tsai Weng, I-Shan Wu, Yi-Tien Chen, Rui-Fu-Shi Medical Laboratory, Taichung40715, Taiwan   Background: More than 50% Mycobacterium isolates are Non-tuberculous Mycobacterium (NTM) in most mycobacterial laboratories in Taiwan. Species identification of mycobacterium requires the use of time-consuming conventional biochemical methods, or more rapid molecular techniques that are expensive and require special instrument. Cord factor is known of the specific toxic factor of Mycobacterium tuberculosis (MTB). Most of MTB can be examined the cord formation in acid fast staining smears (AFS) of direct specimens or liquid cultures, and on young culture colonies of Middlebrook 7H10 or 7H11. The aim of this study is to evaluate the practicability of microscopic examination of the cord formation in BACTEC MGIT (Becton Dickinson Diagnostic Systems, Sparks, MD) culture positive smears for presumptive identification of MTB. Method: A total 522 acid fast positive smears of MGIT cultures were collected on October 2007. Each of smears was examined the microscopic morphology, including cord, pseudo-cord, needle, dot, short, ladder. Species identification of all mycobacterium isolates were undertaken by conventional biochemical techniques. The sensitivity and the specificity of cord formation for discriminating MTB from NTM were calculated by using conventional biochemical techniques as standard. Result: Of 522 MGIT-positive cultures, 250 were identified as MTB and one mixed with MTB and M. avian complex(MAC) by conventional method. The remaining 271 cultures were NTM, including 112 MAC, 21 M. kansasii, 39 M. abscessus, 18 M. fortuitum group, 7 M. gordonae, 6 M. scrofulaceum, 2 M. mucogenicum, 3 M. marinum, 58 unidentified NTM, 3 mixed with two NTM species, 2 mixed with MAC and unidentified NTM. Of 251 MTB isolates, 249 were found cord formation, including the MTB and MAC mixed culture. OF 271 NTM cultures, only one was found cord formation, and 21 were found pseudo-cord. The overall sensitivity and specificity of the cord formation was calculated to be 99.2% and 99.6%. Conclusion: Our results manifested that the usefulness of the cording morphology as a rapid and simple microscopic examination test for presumptive identification of MTB in MGIT-positive cultures with high sensitivity and high specificity. Because pseudo-cord might be confounded as the cord formation of MTB, examiner must have high experience for microscopic morphologies of mycobacteria. We suggest that the cord formation can only be employed as a rapid screening examination for MTB. It still required further confirmation by other test techniques. Key words: Mycobacterium tuberculosis, cord formation, MGIT, microscopic examination Contact person email: jeff_lu@sancordon.com.tw