Table 2.

Predictors of death during treatment (n = 565).

Predictors		Death during treatment N (%)	Unadjusted HR (95% CI), p value	Adjusted HR (95% CI), p value
N		37
Mean age at diagnosis ± SD (hazard/10 year increase)		45 ± 16	1.26 (1.03-1.55), 0.03	1.52 (1.18-1.95), 0.001
Gender	Male	29 (6.4)	0.98 (0.44-2.14), 0.95
	Female	8 (7.1)
Ethnicity	Hispanic	7 (4.2)	0.55 (0.24-1.27), 0.16
	Non-Hispanic	30 (7.5)
Place of birth	U.S.	31 (7.9)	2.92 (1.10-7.74), 0.03	2.70 (0.96-7.63), 0.06
	Foreign	6 (3.5)
HIV status	Positive	24 (11.8)	2.30 (1.14-4.61), 0.02	2.57 (1.17-5.64), 0.02
	Negative/Unknown	13 (3.6)
DOT1	Yes/Partial*	32/373 (8.6)	2.65 (1.02-6.87), 0.04	1.31 (0.47-3.64), 0.61
	No	5/190 (2.6)
Episode of inadequate therapy during intensive phase2	Yes	24/153 (15.7)	3.39 (1.68-6.83), 0.001	3.15 (1.52-6.52), 0.002
	No	13/409 (3.2)
Episode of inadequate therapy during continuation phase3	Yes	10/126 (7.9)	0.72 (0.32-1.65), 0.44
	No	23/430 (5.4)
Sputum smear status	Positive	27 (10.4)	2.97 (1.43-6.17), 0.004	3.07 (1.44-6.56), 0.004
	Negative	10 (3.3)
Extrapulmonary in addition to pulmonary	Yes	18 (11.2)	1.69 (0.85-3.34), 0.13
	No	19 (4.7)
Cavitary disease	Yes	4 (3.0)	0.43 (0.15-1.23), 0.12
	No	33 (7.6)
Any prior medical condition	Yes	25 (8.0)	1.74 (0.85-3.54), 0.13
	No	12 (4.8)

1) Excluding 2 with missing data for directly observed therapy. * Partial DOT = DOT used for less than the full duration of treatment (clinicians opted to transition select patients to self-administered therapy (SAT) at the end of the intensive phase of treatment, or initiated DOT in patients previously treated by SAT in whom clinical, radiographic or microbiologic response to treatment was delayed). 2) Excluding 3 with missing data for intensive phase. 3) Excluding 9 with missing data for continuation phase. Note-HR = hazard ratio