Survival Time to Death and its Predictors among Tuberculosis Patients Who Have Started Anti-TB Treatment in Selected Health Centers of Addis Ababa, Ethiopia: A Retrospective Cohort Study
No Thumbnail Available
Date
2020-01
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Abeba University
Abstract
Background: Tuberculosis remains the leading cause of morbidity and mortality
globally and in Ethiopia despite different strategies have been designed and implemented
to combat it. Survival time and its predictors among tuberculosis patients who enrolled to
care in selected health centers in Addis Ababa were assessed.
Method: A retrospective cohort study was conducted in 20 selected health centers of
Addis Ababa city, Ethiopia. Data were collected from April 1 to August 30, 2018 by
reviewing medical records of tuberculosis patients who were registered from May 2016
to May 2017. Statistical software STATA version 14 was used for analysis of the data.
Kaplan–Meier curve and log-rank test was employed to investigate the statistical
significance of the difference in survival experience among different categories of
patients. Cox-proportional hazard and accelerated failure time model was used to assess
the relationship between baseline variables and mortality. The strength of association was
presented by hazard ratio with 95% CI and results were reported significant at P ≤ 0.05.
Results: The medical records of 371 patients were included in the analysis of which
136 (36.7 %) died during the treatment period. Majority of TB deaths occurred within
two months of the start of treatment and the overall estimated median survival time was
157 days. Based on akaikie information criterion, weibull accelerated failure time model
manifested better results as compared with other models. In multivariable weibull model,
age (HR=0.98, P=0.04), baseline weight (HR=0.96, P=0.03), tuberculosis treatment
phase (continuation phase, HR=0.48, P≤0.01) and tuberculosis type (pulmonary negative
TB, HR=19.92, P≤0.01) were found to be independent predictors of time to death of
tuberculosis patients.
Conclusions: Most of the patients died at the end of study period. This warrants that,
special attention and follow up with nutritional support for pulmonary negative patients
and underweight patients to reduce deaths and for better clinical and treatment outcome.
Description
Keywords
Mortality, Predictors, Survival model, Survival time, Tuberculosis