Multivariate logistic regression analysis for predicators† of XDR-TB mortality in the study cohort with resistance to rrs A1401G.

†A priori variables included in both the univariate and mulitivariate analysis included i) Demographic and clinical: Age, gender, ethnicity, weight at diagnosis, HIV status, CD4 cell count and ART (in HIV-infected), previous MDR-TB treatment, previous pre-XDR diagnosis; ii) M. tuberculosis strain typing and drug-susceptibility testing: Beijing/non-Beijing strain, phenotypic DST for second line injectables (amikacin, kanamycin, capreomycin, streptomycin), ofloxacin, ethambutol, ethionamide, and capreomycin genotypying for rrs A1401G mutant, iii) XDR-TB drug treatments: amikacin, kanamycin, capreomycin, ciprofloxacin, ofloxacin, moxifloxacin, co-amoxicillin/clavulanic, ethambutol, ethionamide, pyrazinamide, PAS, clofazime, dapsone, thioacetone, terizidone/cycloserine.*The total numbers of observations included in each of the multivariate outputs. Numbers differ from previous patient group totals due to missing data. Multiple imputation was not used.Multivariate logistic regression analysis for predicators† of XDR-TB mortality in the study cohort with resistance to rrs A1401G.