Table 4.

Crude and multivariable analysis investigating correlates of community stigma scores among tuberculosis (TB) presumptives

TB presumptives
Variable	Crude analysis		Multivariable analysis
	Community stigma score
	$\beta$ (95% CI)	P-value	$\beta$ (95% CI)	P-value
District
Zululand	0.90 (− 0.75, 2.55)	0.28	1.02 (− 0.86, 2.89)	0.28
Gender
Male	0.99 (− 0.55, 2.54)	0.21	0.0075 (− 1.54, 1.55)	0.99
Age	− 0.025 (− 0.092, 0.043)	0.47
Race
Black	− 0.57 (− 2.39, 1.25)	0.54
Relationship
Union	− 1.04 (− 2.73, 0.64)	0.22	− 0.81 (− 2.48, 0.85)	0.34
Employment
Employed	0.37 (− 1.22, 1.96)	0.65
Community type
Rural/farming	0.89 (− 0.86, 2.63)	0.32
Education
12th grade and above	− 0.89 (− 2.45, 0.67)	0.26
Income
≥ ZAR 5000	− 0.26 (− 2.21, 1.69)	0.79
Causes of TB*
Good understanding	0.45 (− 1.23, 2.13)	0.60	− 0.23 (-2.00, 1.54)	0.80
Mixed understanding	− 1.59 (-4.01, 0.84)	0.20	− 1.28 (-3.60, 1.04)	0.28
Poor understanding	REF		REF
TB/HIV knowledge
TB increases chance of HIV	2.12 (0.62, 3.63)	0.01	1.35 (− 0.36, 3.06)	0.12
HIV increases chance of TB	2.90 (0.79, 5.02)	0.01	2.48 (0.020, 4.94)	0.05
TB contacts	− 1.37 (− 2.90, 0.17)	0.08	− 1.01 (− 2.58, 0.55)	0.20
Ever having TB	0.34 (− 1.68, 2.36)	0.74
HIV status
Positive	− 0.63 (− 3.00, 1.73)	0.60
Negative	− 1.19 (− 3.49, 1.12)	0.31
Unknown/no test	REF
HIV stigma	0.036 (− 0.16, 0.23)	0.71
Social support	− 0.093 (− 0.16, − 0.026)	0.01	− 0.077 (− 0.14, − 0.010)	0.02

^*Good understanding denotes participants who only identified accurate causes of TB (i.e. bacteria). Mixed understanding denotes participants who identified accurate and mis-informed causes of TB (i.e. infection and witchcraft). Poor understanding denotes participants who only identified incorrect causes of TB (i.e. exposure to the cold)