Factors associated with tuberculosis by HIV status in the Brazilian national surveillance system: a cross sectional study

Table 3 **Hierarchical* multivariate analysis of the association of tuberculosis by HIV status**

	Characteristics		OR**	95%CI***
	Gender	Female	Ref.
Level 1		Male	1,06	1.03-1.10
	Age	<20 years	Ref.
		20-39 years	4,82	4.34-5.36
		40-59 years	3,67	3.30-4.09
		≥60 years	1.04	0.92-1.18
	Race	White	Ref.
		Black	1.08	1.04-1.13
		Browns	0,77	0.75-0.80
		Other	0,66	0.57-0.77
	Schooling	<4 years	Ref.
		4-7 years	1.23	1.19-1.28
		≥ 8 years	0,96	0.92-1.00
	Area of residence	Urban	Ref.
		Rural	0,44	0.41-0.48
Level 2		Periurban	0,66	0.61-0.77
	DOT****	No	Ref.
		Yes	0.65	0.98-1.12
	Institutionalization	No	Ref.
		Yes	1.07	1.00-1.15
Level 3	Diabetes	No	Ref.
		Yes	0.45	0.40-0.50
	Alcoholism	No	Ref.
		Yes	1.04	0.99-1.09
Level 4	Type of entry	New case	Ref.
		Relapse	2.23	1.00-4.94
		Return after default	2.65	1.13-6.25
		Unknown	1	-
		Transferred	0.79	0.35-1.81
	Tuberculin skin test	Negative	Ref.
		Positive	0.73	0.50-1.07
	TB form	Pulmonary	Ref.
		Extra pulmonary	1.28	0.75-2.20
		Pulmonary + Extra pulmonary	2.80	1.56-5.02
	Smear	Negative	Ref.
		Positive	0.55	0.32-0.95
	Histopathologic examination	AFB***** positive	Ref.
		Suggestive	2.15	1.13-4.07
		Not suggestive	0.71	0.37-1.34
	X-Ray	Negative	Ref.
		Suspicious of TB	0.62	0.35-1.09
Level 5	Outcome	Cure	Ref.
		Default	2.79	2.45-3.18
		Death from TB	3.64	2.92-4.53
		Death from other causes	7.93	6.66-9.43
		MDR-TB	1.41	0.74-2.68

*The multivariate analyses: step 1 (Gender + age + schooling + race + Area of residence); step 2 (variables retained from step 1 + Institutionalization + DOT); step 3 (variables retained from step 2 + Diabetes); step 4 (variables retained from step 3 + treatment type+ TB form + Tuberculin Skin Test + smear + + culture + histopathologic examination + X ray suspicious for TB); and step 5 (variables retained from step 4 + outcome). ** Adjusted odds ratio. ***CI: confidence interval. ****DOT: Directly Observed Therapy. *****AFB: acid-fast bacilli.

ISSN: 1471-2334