TABLE 1

Multivariable logistic regression assessing the association between uptake of social protection and financial hardship

Financial hardshipOR (95% CI)p-value
Total costs ≥20% of household income#0.37 (0.14–0.94)ƒ0.04
Used coping strategy1.31 (0.50–3.47)##0.58
Impoverished+0.16 (0.04–0.54)¶¶0.002
All three financial hardships§0.01 (0.00–0.07)++<0.001

Socioeconomic and clinical characteristics associated (p<0.1) with outcomes in univariable logistic regression were included in multivariable logistic regression models. #: pre-illness annual household income; three participants with pre-illness annual household income USD 0 were excluded, so n=116. : one participant with missing data on use of coping strategies was excluded, so n=118. +: defined using Brazil's 2016 poverty line of monthly household income per capita of USD 48.6 per month [18]; five participants in poverty pre-illness were excluded, so n=114. §: six participants with either pre-illness annual household income USD 0, missing data on use of coping strategies, or in poverty pre-illness were excluded, so n=113. ƒ: mutually adjusted for sex, age, education, acquired drug-resistant tuberculosis (TB), and time to directly observed therapy (DOT) clinic. ##: mutually adjusted for sex, age, pre-illness household head, pre-illness household income per capita, acquired drug-resistant TB, and time to DOT clinic. ¶¶: mutually adjusted for sex, age, education, and pre-illness household income per capita. ++: mutually adjusted for sex, age, pre-illness household head, pre-illness household income per capita, acquired drug-resistant TB, and time to DOT clinic.