Table 2. Weibull proportional hazard regression to measure the effects of predictors such as BMI, type of lineages, and HIV on the hazard rate using auramine conversion time.

	Unadjusted			Adjusted
Lineages	Haz. Ratio	p	[95% CI]	Haz. Ratio	p	[95% CI]
1	1.09	0.63	0.76–1.56	1.09	0.67	0.72–1.65
2	1.04	0.89	0.58–1.84	1.00	0.98	0.53–1.89
3	0.70	0.27	0.37–1.31	0.58	0.19	0.26–1.31
5	0.80	0.48	0.42–1.49	1.15	0.68	0.57–2.32
6	0.79	0.01	0.66–0.95	0.80	0.02	0.66–0.97
M. bovis	0.53	0.15	0.22–1.27	0.59	0.25	0.24–1.44
HIV positive	0.48	0.00	0.35–0.66	0.45	0.00	0.32–0.63
BMI (day 0)	1.05	0.00	1.03–1.08	1.06	0.00	1.03–1.09
Few AFB (day 0)	0.45	0.03	0.22–0.95	0.44	0.03	0.21–0.93
Moderate and Many AFB (day 0)	0.35	0.00	0.17–0.72	0.34	0.00	0.17–0.70

Proportions of smears converted and non-converted by Lineage to estimate the effect of predictors in smear conversion using Auramine Rhodamine staining technique; L = lineage 1, 2, 3, 5 and 6; HIV = Human Immunodeficiency Virus; BMI = Body Mass Index; CI = Confidence Interval; p = Probability. L4 is not shown as it is the reference for the others. In addition HIV+ was fitted against HIV-, and AFB smear grade were fitted against smear negative.