Table 3 Logistic regression for non-standard initial TB treatment.

From: The prevalence, clinical reasoning and impact of non-standard anti-tuberculosis regimens at the initial prescription

Factors

Multivariable

Adjusted HR (95% CI)

p value

Eye disease

8.869 (2.542–30.949)

0.001

Liver disease

12.790 (3.981–41.089)

 < 0.001

Gout/hyperuricemia

4.012 (1.196–13.425)

0.024

  1. Significant pre-treatment factors (p < 0.010) in univariate analysis are entered in the multivariable regression.
  2. The factors in the final model were selected by the forward conditional stepwise methods.
  3. TB tuberculosis.
Back to article page