Table 1.
Definitions and treatment regimens used under the Revised National TB Control Programme in 2015–2016.
Fig 1.
Flow diagram indicating the selection of study population.
*RNTCP = Revised National Tuberculosis Programme of India; §DOT = Directly Observed Therapy.
Table 2.
Demographic and clinical characteristics of TB-HIV co-infected patients and bivariable analysis showing association between these characteristics with unfavourable TB treatment outcomes in Delhi, 2015 (N = 816).
Table 3.
Comparison of clinical characteristics and treatment outcomes of TB-HIV patients with complete data vs incomplete data on co-variates* included for building a predictive model for unfavourable model in Delhi.
Table 4.
Variable combinations/iterations used for selecting the prediction model and its performance values for selecting the most appropriate model for predicting unfavourable outcomes in TB-HIV patients enrolled for TB treatment in 2015 in Delhi.
Table 5.
Coefficients (log of odds ratios), its standard errors, p-value and odds ratios (95% CI) of the prediction model for unfavourable TB treatment outcomes among TB-HIV co-infected patients enrolled for TB treatment in Delhi, 2015 (n = 448).
Fig 2.
Receiver operating characteristics curve and the area under the curve for various sensitivity and specificity values of the prediction model developed for identifying TB HIV patients with unfavourable TB treatment outcomes under RNTCP in Delhi, India, 2015.
Fig 3.
Sensitivity and specificity of various cut-off values of predicted probabilities derived from a prediction model for identifying unfavourable TB treatment outcomes in TB HIV patients under RNTCP in Delhi, 2015.