A successful introduction to a non-expert setting of the thin-layer agar Colour Test as an indirect phenotypic drug susceptibility test for Mycobacterium tuberculosis

Objectives: We evaluated the performance of the MDR/XDR-TB Colour Test (CT) as an in-house thin-layer agar-based indirect drug susceptibility test (DST) for Mycobacterium tuberculosis (MTB) in a non-expert setting in Estonia. Methods: After 2 days of hands-on training for laboratory technicians, 6 panels of 150 MTB isolates were cultured onto CT plates prepared in-house in 2 laboratories. Triplicate readings of 900 CT plates resulted in 18 DST patterns for each initial isolate. Time intervals to the results and for media preparation were estimated, and intra- and interobserver agreement, test sensitivities and specificities were calculated. BACTEC MGIT 960 DST was used as a reference. Results: The median time to produce DST results for isoniazid, rifampicin and levofloxacin was 13 days. CT sensitivity was 94.7% for levofloxacin, 95.8% for isoniazid and 97.3% for rifampicin. Test specificities were >97% for all 3 drugs. Interobserver agreement was 100% in Lab A and in Lab B >97% for levofloxacin and 99% for isoniazid and rifampicin. Conclusions: The implementation of the CT into a new laboratory was straightforward with only minimal guidance required. This study proves that the CT is highly reproducible and easily interpreted by previously inexperienced personnel. ? 2020 The Author(s). Published by Elsevier Ltd on behalf of International Society for Infectious Diseases. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc Objectives: We evaluated the performance of the MDR/XDR-TB Colour Test (CT) as an in-house thin-layer agar-based indirect drug susceptibility test (DST) for Mycobacterium tuberculosis (MTB) in a non-expert setting in Estonia. Methods: After 2 days of hands-on training for laboratory technicians, 6 panels of 150 MTB isolates were cultured onto CT plates prepared in-house in 2 laboratories. Triplicate readings of 900 CT plates resulted in 18 DST patterns for each initial isolate. Time intervals to the results and for media preparation were estimated, and intra-and interobserver agreement, test sensitivities and specificities were calculated. BACTEC MGIT 960 DST was used as a reference. Results: The median time to produce DST results for isoniazid, rifampicin and levofloxacin was 13 days. CT sensitivity was 94.7% for levofloxacin, 95.8% for isoniazid and 97.3% for rifampicin. Test specificities were >97% for all 3 drugs. Interobserver agreement was 100% in Lab A and in Lab B >97% for levofloxacin and 99% for isoniazid and rifampicin. Conclusions: The implementation of the CT into a new laboratory was straightforward with only minimal guidance required. This study proves that the CT is highly reproducible and easily interpreted by previously inexperienced personnel. (c) 2020 The Author(s). Published by Elsevier Ltd on behalf of International Society for Infectious Diseases. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Automated summary

The study evaluated the performance of the MDR/XDR-TB Colour Test as an in-house drug susceptibility test for Mycobacterium tuberculosis in Estonia. Results showed high sensitivity and specificity of the CT for levofloxacin, isoniazid, and rifampicin. High interobserver agreement was achieved in both laboratories.