Predictors of Mortality among Human Immunodeficiency Virus infected Patients taking Antiretroviral Treatment in Public Health Facilities, Bahir Dar, Northwest Ethiopia

Abstract

Abstract Background: Ethiopia has an estimated adult Human Immunodeficiency Virus prevalence of 1.3% and annual Acquired Immunodeficiency Syndrome deaths 34,365 in 2013. Even though the use of Antiretroviral Treatment and prophylaxis for opportunistic infections has led to a reduction in mortality, it remains still higher compared to the general population especially in resource-constrained countries. Objective: To assess mortality and predictors of mortality among Human Immunodeficiency Virus infected adult patients taking Antiretroviral Treatment in public health facilities. Methods: Institution based retrospective follow-up study was conducted among 1901 HIV infected patients started antiretroviral treatment from January 2008 to December 2012, in four public health facilities, Northwest Ethiopia. Descriptive statistics were calculated. Kaplan-Meier curves to estimate mortality and Cox proportional hazards models to assess predictors of mortality were used. Data quality was assured during data retrieval, entry and analysis. Results: Treatment outcomes were active follow up 1497(78.7%), transfer out 254(13.4%), lost to follow up 73(3.8%), stop 33(1.7%) and deceased 44(2.3%). There were 44 deaths in 4476.4 person-years of retrospective follow up resulting in an incidence density of 9.8 per 1000 person-year observations (95% CI 7.3 to 13.2). In multivariate analysis independent predictors of mortality were bedridden functional status, WHO Stage IV, CD4 cell count less than 100cells/mm3, history of CNS toxoplasmosis, year of ART initiation and interruption of ART. Conclusion and recommendations: There was an overall low level of mortality (2.3%) of patients on antiretroviral treatment but high death rate in the first 5 months. Promoting early voluntary counselling and testing, followed by early initiation of antiretroviral treatment and close monitoring in the early course of treatment is crucial for better outcomes