Factors associated with death from tuberculosis and treatment default in a general hospital in the city of Rio de Janeiro, 2007 to 2014

Abstract

Background and Objectives: tuberculosis remains a relevant public health problem and the cure of diagnosed cases is a strategy to reduce high morbidity and mortality. The objective of our study is to identify factors associated with death due to tuberculosis and treatment default in patients diagnosed in the HFSE from 2007 to 2014. Method: we conducted an observational, analytical study of all 670 confirmed cases reported in this period at the local SINAN reporting base, whose outcome was registered as cure (383), tuberculosis death (159) or default (128). This system is fed from a standardized epidemiological investigation sheet containing socio-demographic and clinical variables. Statistical analysis involved variables available in SINAN and considered as unfavorable outcome that composed by death or default. Variables with high odds ratios in bivariate analysis and/or clinical relevance were considered for inclusion in a logistic regression model. As a measure of association for analysis of factors associated with unfavorable outcome versus cure, odds ratios and their 95% confidence intervals were estimated. Results: the final model identified as statistically associated with an unfavorable outcome: age ≥60 years, non-white race/color, history of previous treatment for tuberculosis, or unknown history of previous treatment, positive anti-HIV test, pulmonary form and presence of severe extrapulmonary forms. Conclusion: the results confirm the need to invest in policies that guarantee access and fair care to patients, especially those with conditions that predispose to default and severe forms.