Operational feasibility and multi-centric evaluation of ‘TBDetect sputum microscopy kit’ for the direct detection of Mycobacterium tuberculosis in field settings

Abstract

Background India relies primarily on direct smear microscopy for tuberculosis (TB) diagnosis. However, the low sensitivity of smear microscopy emphasizes the need to improve its performance. We recently described the development of ‘TBDetect’ kit which showed improved performance over direct smear microscopy at National Reference Laboratories (NRLs) in India. Methods The present study was aimed to assess the operational feasibility of ‘TBDetect’ microscopy in field settings. This was evaluated by (i) assessing the performance of ‘TBDetect’ microscopy vs. LED-fluorescence microscopy (LED-FM) on consecutive presumptive pulmonary TB patients (n = 5300) who attended Designated Microscopy Centres (DMCs, n = 13) under 4 NRLs at Bhubaneswar, Bhopal, Chennai, and New Delhi, and (ii) obtaining feedback from Scientists (n = 10) and laboratory technicians (n = 42) using semi-structured questionnaires under the following parameters: feasibility of initiation of ‘TBDetect’ microscopy in DMCs, sample preparation and testing, training, time-to-result, logistics, and troubleshooting. A scoring questionnaire was also used to assess ‘TBDetect’ microscopy vs. LED-FM and statistical significance of the scores was calculated using paired t-test. Results The overall positivity of ‘TBDetect’ microscopy was 10.32% (547/5300) vs. 8.96% (475/5300) of LED-FM at all sites and the increment in positivity was significant (p = 0.019). In addition, ‘TBDetect’ microscopy yielded an increment in smear grade status over LED-FM (p = 0.043). The feedback from the study-in-charge and kit users indicated that ‘TBDetect’ microscopy was easily adapted in point-of-care settings. An analysis of scoring feedback suggested that it was easy to perform and observe in comparison to LED-FM (p < 0.005). Conclusions This study established the feasibility of ‘TBDetect’ microscopy in field settings.