Abstract

Introduction. Drug resistant tuberculosis remains a worldwide problem that requires prompt diagnosis. Hypothesis/Gap statement. The WHO recommended direct, rapid Xpert MTB/RIF is prohibitively costly, therefore, there is a need to validate a rapid, affordable DST for use in low- and middle-income settings. Aim. The technical performance and time to results of a simple, direct microscopy-based slide DST (SDST) assay for diagnosis of rifampicin-resistant TB was evaluated in Uganda. Methodology. Sputum samples from 122 smear-positive re-treatment TB patients presenting to the TB treatment centre at Uganda’s National Referral Hospital, Mulago, Kampala, Uganda were examined. The sputum samples were tested by the direct SDST which was compared to the indirect Lowenstein Jensen Proportion Method (LJDST) method as the gold standard. The time to results was defined as the time from DST setting to results interpretation. The results were further analysed for sensitivity and specificity as well as agreement between LJDST and SDST for rifampicin resistance determination. Results. A total of 117 smear positive sputum samples with valid results for both tests were compared. The median time to results for SDST was 14 days with an interquartile range (IQR) of 10–14 days compared to 60 days with IQR of 60–75 days for LJDST. The number for rifampicin resistance by the gold standard LJDST was 26. The SDST had a sensitivity of 96 % (95 %; CI 81–99 %) and a specificity of 97.8 % (95 %; CI 93–100 %). The Positive Predictive and Negative Predictive values for SDST were 92.3 % (95 %; CI 76.8–99 %) and 98.9 % (95 %; CI 94–100 %), respectively. The kappa agreement between SDST and LJDST was 92.3 %. Conclusion. The SDST was found to be a rapid and accurate direct test for the detection of rifampicin resistance among retreatment TB cases in low-income settings.