Open AccessAgreement of Middle brook 7H10 with Lowenstein Jensen and accuracy of the Sensititre MYCOTB plate using either method as a reference standard for Mycobacterium tuberculosis first line drug susceptibility testing
Author(s) -
Willy Ssengooba,
Germine Nakayita,
Carolyn Namaganda,
Moses Joloba
Publication year - 2018
Publication title -
plos one
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.99
H-Index - 332
ISSN - 1932-6203
DOI - 10.1371/journal.pone.0199638
Subject(s) - ethambutol , isoniazid , streptomycin , rifampicin , mycobacterium tuberculosis , tuberculosis , medicine , microbiology and biotechnology , antibiotics , biology , pathology
Introduction Although Sensititre Mycobacterium tuberculosis (MYCOTB) plate offers both drug susceptibility testing (DST) and minimum inhibitory concentration (MIC) results, it has not been evaluated against both Lowenstein Jensen (LJ) and Middlebrook 7H10 (MB7H10) DST methods at standard critical concentrations. Materials and methods We analyzed 76 M . tuberculosis isolates consisting of 54 isolates from the Uganda National TB drug resistance survey done December 2009–February 2011 and 22 isolates from the World Health Organization External Quality Assessment panel for the year 2011. All isolates were tested for LJ, MB7H10 and MYCOTB plate based DSTs for streptomycin, isoniazid, rifampicin and ethambutol anti-tuberculosis drugs. The agreement of MB7H10 with LJ and accuracy of MYCOTB plate using either LJ-DST or MB7H10 as a reference standard were determined. Results The agreement (kappa) of MB7H10 with LJ was; 0.687 for rifampicin, 0.498 for isoniazid, 0.275 for streptomycin and 0.082 for ethambutol which as almost similar when compared with MYCOTB plate. The sensitivity (95% confidence interval; CI) of MYCOTB plate when LJ was used as a reference standard was higher for streptomycin 87.5% (81.6–98.4) followed by isoniazid 75.9% (65.1–95.6) and rifampicin 73.1% (52.2–88.4). When MB7H10 was used as reference standard, the sensitivity of MYCOTB plate improved significantly; isoniazid 96.2% (80.3–99.9), rifampicin 94.0 (83.4–98.7) and 93.8% (69.7–99.8). There was good agreement between MYCOTB plate and MB7H10; 1.00 for ethambutol, 0.959 for streptomycin, 0.915 for rifampicin and 0.778 for isoniazid. Conclusions The performance of the two culture-based reference standards for phenotypic first-line drug susceptibility testing methods, LJ and MB7H10, varied much even with acceptable MYCOTB plate MICs. There was acceptable agreement and accuracy of MYCOTB plate for drug susceptibility testing when MB7H10 was used as reference standard than with LJ-DST. Results from MIC information makes the MYCOTB plate more suitable for guiding clinicians on the choice of the most appropriate TB treatment regimen as well as limits of detection for TB drug resistance.