Table 3. Multivariable models for the association between regimen composition and sputum culture conversion.
| Predictor | aHR (95% CI) | |
|---|---|---|
| Model 1 | Model 2 | |
| Average number of potentially effective drugs received per day | Not included in model | |
| 0 to <4 | 0.36 (0.29–0.44) | |
| 4 to <5 | 0.56 (0.47–0.66) | |
| 5 to <6 | Reference | |
| 6 or more | 1.36 (1.09–1.69) | |
| Average number of effective drugs received per day * | Not included in model | 1.65 (1.48–1.84) per drug, when regimen contains one untested drug |
| Average number of ineffective drugs received per day * | Not included in model | 1.33 (1.18–1.51) per drug |
| Average doses of pyrazinamide received per day * | Not included in model | 2.00 (1.65–2.41) per drug, when regimen contains one untested drug |
| Average number of untested drugs received per day * | Not included in model | See Table 4 |
| Number of drugs to which TB resistant at baseline * | 1.09 (1.02–1.17) | 1.07 (0.99–1.16) per drug |
| Resistance pattern at baseline | ||
| MDR only | Reference | Reference |
| MDR with resistance to any second-line injectable | 0.51 (0.38–0.70) | 0.56 (0.42–0.76) |
| MDR with resistance to any fluoroquinolone | 0.52 (0.36–0.73) | 0.48 (0.34–0.68) |
| Previous treatment history | ||
| None | Reference | Reference |
| First-line drugs only | 0.84 (0.66–1.06) | 0.84 (0.67–1.07) |
| Second-line drugs | 0.62 (0.46–0.83) | 0.72 (0.54–0.97) |
| Unknown | 0.46 (0.24–0.87) | 0.53 (0.28–1.02) |
| Smear result | ||
| Negative | Reference | Reference |
| Positive | 0.72 (0.56–0.92) | 0.73 (0.56–0.94) |
| Unknown | 0.25 (0.12–0.50) | 0.22 (0.10–0.46) |
| Extent of disease on chest radiograph | ||
| Unilateral | Reference | Reference |
| Bilateral | 0.70 (0.59–0.83) | 0.68 (0.57–0.80) |
| Unknown | 1.16 (0.71–1.87) | 1.09 (0.67–1.78) |
aHRs result from two separate Cox proportional hazards regression models for the association between regimen composition and time to initial sputum culture conversion. All variables included in each model are shown, and analysis was stratified by country. Model 1 characterizes regimens using the average number of potentially effective drugs received, treated as a categorical variable. Model 2 characterizes regimens using the average number of drugs in each of four categories, treated as continuous variables. Model 2 included the following two interaction terms: average number of effective drugs received per day × average number of untested drugs received per day; average doses of pyrazinamide received per day × average number of untested drugs received per day. Initial sputum culture conversion was defined as at least two consecutive negative cultures of sputum samples collected at least 30 d apart. aHRs in bold are statistically significant.
*Continuous variables.
MDR, multidrug resistant.