Study of different methods of susceptibility testing of M. tuberculosis

This communication is on a multi-centre collaborative study carried out by New Delhi Tuberculosis (NDTB) Centre, Delhi, to compare the different methods of drug susceptibility testing (DST) of Mycobacterium tuberculosis for Streptomycin, Rifampicin, Isoniazid and Ethambutol in various accredited laboratories 40 samples (selected by simple random sampling technique) from sputum positive out-patients presenting between December 2004 and January 2005 were collected at New Delhi Tuberculosis Centre from 24 males and 16 females (age ranging from 18 to 55 years) were taken up for culture and susceptibility testing. The proportion method of DST done at NDTB Centre was taken as standard for comparing the other methods after it was found to have a good agreement with the proportion method performed at Tuberculosis Research Centre, Chennai (Kappa coefficient 0.75 to 1). All 40 sub-cultures were processed for DST by the resistance ratio and proportion methods at NDTB Centre. Also, the 40 sub-cultures were sent to Central JALMA Institute for Leprosy (JALMA), Agra that used the resistance ratio method and 25 sub-cultures were transported to National Institute of Communicable Diseases (NICD), Delhi for performing the Bactec 460TB method. The resistance ratio and proportion methods were compared both intra-laboratory (NDTB Centre) and inter-laboratory (NDTB Centre and JALMA). The sensitivity and specificity of the resistance ratio method compared to the proportion method both inter and intra-laboratory, were very high for all the drugs (except Ethambutol inter-laboratory). This was corroborated by the good agreement between the resistance ratio performed at the two centres. Bactec 460TB method, on the other hand, did not have acceptable measures of validity (sensitivity and specificity) compared to the proportion method for Streptomycin and Ethambutol. Both resistance ratio method & proportion method are comparable and either can be used for making clinical decisions. But Bactec 460 method, despite yielding early results loses out on validity.