Table 2. Multivariable analysis of index episode risk factors for subsequent smear-positive tuberculosis re-treatment (N = 1,733).
Total re-treatment cases | PY | Rate (per 100 PY) | Crude hazard-ratio* (95% CI) | P-value | |
Treatment outcome | <0.001 | ||||
Cured | 185 | 8863 | 2.09 (1.81–2.41) | 1 | |
Completed | 14 | 640 | 2.19 (1.29–3.69) | 1.08 (0.58–2.00) | |
Defaulted | 92 | 1341 | 6.86 (5.59–8.41) | 3.97 (3.00–5.26) | |
Smear grade † | 0.03‡ | ||||
Smear 1+ | 35 | 1588 | 2.20 (1.58–3.07) | 1 | |
Smear 2+ | 48 | 1576 | 3.05 (2.30–4.04) | 1.43 (0.93–2.22) | |
Smear 3+ | 154 | 4641 | 3.32 (2.83–3.89) | 1.61 (1.11–2.33) | |
Age | 0.003 | ||||
0–18 | 30 | 758 | 3.96 (2.77–5.66) | 1 | |
19–39 | 193 | 6646 | 2.90 (2.52–3.34) | 0.50 (0.32–0.78) | |
40+ | 68 | 3430 | 1.98 (1.56–2.51) | 0.40 (0.25–0.65) |
PY = Person-years.
CI = Confidence Interval.
Adjusted for the other factors shown in the table.
Smear grade at start of treatment:
Smear 3+: Any of the two initial smears was 3+, i.e. >10 acid-fast bacilli (AFB) per 1 high-power field (HPF).
Smear 2+: Any of the two initial smears was 2+, i.e. 1–10 AFB per 1 HPF, but none of them was 3+.
Smear 1+: Any of the two initial smears was 1+, i.e. 10–99 AFB per 100 HPF, but none of them was 2+ or 3+.
Test for trend.