Implementation of the new integrated algorithm for diagnosis of drug-resistant tuberculosis in Karnataka State, India: How well are we doing?

Abstract

BackgroundAs per national policy, all diagnosed tuberculosis patients in India are to be tested using Xpert® MTB/RIF assay at the district level to diagnose rifampicin resistance. Regardless of the result, samples are transported to the reference laboratories for further testing: first-line Line Probe Assay (FL-LPA) for rifampicin-sensitive samples and second-line LPA(SL-LPA) for rifampicin-resistant samples. Based on the results, samples undergo culture and phenotypic drug susceptibility testing. We assessed among patients diagnosed with tuberculosis at 13 selected Xpert laboratories of Karnataka state, India, i) the proportion whose samples reached the reference laboratories and among them, proportion who completed the diagnostic algorithm ii) factors associated with non-reaching and non-completion and iii) the delays involved.MethodsThis was a cohort study involving review of programme records. For each TB patient diagnosed between 1st July and 31st August 2018 at the Xpert laboratory, we tracked the laboratory register at the linked reference laboratory until 30th September (censor date) using Nikshay ID (a unique patient identifier), phone number, name, age and sex.ResultsOf 1660 TB patients, 1208(73%) samples reached the reference laboratories and among those reached, 1124(93%) completed the algorithm. Of 1590 rifampicin-sensitive samples, 1170(74%) reached and 1104(94%) completed the algorithm. Of 64 rifampicin-resistant samples, only 35(55%) reached and 17(49%) completed the algorithm. Samples from rifampicin-resistant TB, extra-pulmonary TB and two districts were less likely to reach the reference laboratory. Non-completion was more likely among rifampicin-resistant TB and sputum-negative samples. The median time for conducting and reporting results of Xpert® MTB/RIF was one day, of FL-LPA 5 days and of SL-LPA16 days.ConclusionThese findings are encouraging given the complexity of the algorithm. High non-reaching and non-completion rates in rifampicin-resistant patients is a major concern. Future research should focus on understanding the reasons for the gaps identified using qualitative research methods.