Aims and objectives Multidrug resistant tuberculosis (MDR-TB) management is expensive, prolonged and complicated. After starting anti-tuberculous treatment in MDR-TB patients, microbiological response to therapy is monitored by monthly sputum cultures. Conventional culture methods are costly, leading to poor compliance with this recommendation in resource-limited settings and reliance on less sensitive sputum microscopy. Published data suggests a less favorable outcome with sputum microscopy and stresses upon use of follow-up cultures. Micro-colony liquid culture method is an established method for diagnosis of new TB cases. As anti- tuberculous treatment may produce considerable changes in bacterial morphology, leading to pleomorphic shapes and sizes, the utility of micro colony broth culture for treatment follow-up cases is uncertain. Therefore, this study aims to evaluate the micro-colony liquid culture method in terms of its sensitivity, specificity, rapidity and cost to monitor response to second-line ATT in diagnosed MDR-TB patients. Methods Prospective cross-sectional study performed at Clinical Microbiology Laboratory of Aga Khan University. During the period of February 2013–September 2014, a total of 139 adult, MDR-pulmonary TB patients were enrolled in this study. For each patient an appropriate sputum specimen was collected for MTB culture initially and then monthly (maximum) up to the next 6 months. Samples were processed using the standard protocol for microscopy, routine MIGIT & LJ culture and micro colony broth culture method. Micro colony broth culture finally evaluated for sensitivity, specificity, rapidity, cost and contamination rate. Results To date, a total of 502 sputum samples were submitted from 139 enrolled patients. Out of these, 170 were smear-positive, while 310 were smear-negative. Sensitivity, specificity, positive predictive value and negative predictive value of micro colony broth culture for AFB smear positive samples were 100% (95% CI: 97.83–100), 95.45% (95% CI: 77.08–99.24), 99.42% (95% CI: 96.77–99.90), and 100.00% (95% CI: 83.75–100.00), respectively. Sensitivity, specificity, positive predictive value and negative predictive value of micro colony broth culture for AFB smear negative samples was 100.00% (95% CI: 93.78–100.00), 99.60% (95% CI: 97.80–99.93), 98.31% (95% CI: 90.88–99.72) and 100.00% (95% CI: 98.53–100.00), respectively. Average time of positivity of standard cultures for smear positive samples (n = 192) was 26.0 days, while micro colony broth culture average positivity time was 8.8 days. Average time of positivity of standard cultures for smear negative samples (n = 310) was 30.0 days, while micro colony broth culture average positivity time was 11.4 days. Average contamination rate of micro colony broth culture was 4.5% in comparison with 6.1% of routine cultures. Finally, the cost of the micro colony methods was about 35% of the combined BACTEC and LJ media cultures. Conclusions The preliminary data from this study indicates micro colony broth culture method for MTB detection to be highly sensitive, specific, rapid and cost-effective for treatment monitoring of MDR-TB cases. In resource-limited settings this method can be used safely either alone or along with LJ medium for monitoring of second-line anti-tuberculous treatment. Acknowledgement This work was supported by a grant from Aga Khan University Research Council.
