Table 2. Risk factors for acquired resistance to second-line drugs among patients treated for multidrug-resistant tuberculosis, Georgia, March 2009–October 2012*.
| Risk factor | Univariate analysis, OR (95% CI) | p value | Multivariate analysis, aOR (95% CI) | p value |
|---|---|---|---|---|
| Baseline characteristic | ||||
| Median age >35 y | 1.96 (0.72–5.30) | 0.19 | – | – |
| Male sex | 2.15 (0.59–7.83) | 0.25 | – | – |
| BMI ≤18.5 kg/m2 | 1.96 (0.70–5.45) | 0.20 | 3.73 (0.98–14.14) | 0.053 |
| History of TB | 1.50 (0.57–3.94) | 0.42 | – | – |
| Prior receipt of second-line TB drugs | 3.08 (0.72–13.13) | 0.13 | – | – |
| Diabetes | 0.40 (0.05–3.19) | 0.38 | – | – |
| Hepatitis C | 1.57 (0.40–6.13) | 0.52 | – | – |
| HIV | 1.30 (0.14–11.78) | 0.82 | – | – |
| Cavitary disease | 7.45 (2.65–20.96) | <0.01 | 5.21 (1.56–17.38) | <0.01 |
| No. of drugs to which baseline isolate was resistant/drug (IQR) | 1.63 (1.05–2.51) | 0.03 | – | – |
| Resistant to ≥6 drugs by baseline DST | 4.63 (1.56–13.68) | <0.01 | 5.31 (1.50–18.77) | 0.01 |
| Baseline ofloxacin resistant | 6.24 (1.51–25.83) | 0.01 | – | – |
| Baseline capreomycin or kanamycin resistant | 1.20 (0.44–3.27) | 0.73 | – | – |
| Known active drugs in initial regimen per drug | 0.58 (0.35–0.99) | 0.045 | – | – |
| Follow-up characteristic | ||||
| Initial MDR TB treatment | ||||
| Capreomycin | 2.92 (1.04–8.18) | 0.04 | – | – |
| Treatment interruption | 2.93 (1.08–7.99) | 0.04 | – | – |
| >30 d to start SLDs | 0.48 (0.17–1.34) | 0.16 | – | – |
| Baseline AFB sputum smear value >3+ | 4.44 (1.61–12.22) | <0.01 | 2.21 (0.66–7.48) | 0.20 |
| Sputum smear positive, mo† | ||||
| 4 | 8.78 (1.95–39.66) | <0.01 | 6.54 (1.23–34.88) | 0.03 |
| 6‡ | 15.31 (4.66–50.32) | <0.01 | – | – |
*OR, odds ratio; aOR, adjusted OR; –, not included in multivariate analysis; BMI, body mass index; TB, tuberculosis; IQR, interquartile range; DST, drug susceptibility testing; MDR, multidrug resistant; SLDs, second-line drugs; AFB, acid-fast bacilli. †Time from initiation of second-line drug treatment for MDR TB. ‡Significant by an alternative multivariate analysis model when replacing the variable sputum smear positive at 4 mo.