Determinants and causes of mortality in HIV-infected patients receiving antiretroviral therapy in Burkina Faso: a five-year retrospective cohort study

Abstract

In this study, we investigated the causes of death and the factors associated with mortality in a cohort of patients receiving highly active antiretroviral therapy (HAART) in Burkina Faso, an African country with limited resources. This retrospective cohort study included patients aged 15 years and older who started HAART for the first time between January 2003 and December 2008 in 14 health districts. We used survival analyses, including the Kaplan–Meier method, to examine potential predictors of death and two Cox proportional hazard models to estimate hazard ratios for death, first from baseline covariates and then from time-dependent covariates. A total of 6641 patients initiated HAART during this period; of these, 5608 were included in the analysis. By the end of the study period, 4310 of those patients were still receiving HAART, 690 had died, 207 had been transferred and 401 were lost to follow-up. The median duration of follow-up was 23.2 months [interquartile range (IQR): 12.4–36.9], and the overall incidence of mortality was 6 per 100 person-years. The clinical stage, CD4 count, body mass index (BMI), haemoglobin level, HAART regimen, gender, age, profession and year of initiation were the primary risk factors associated with death. In the multivariate analysis, BMI, clinical stage, treatment regimen and CD4 count remained significantly associated with death. The most frequent causes of death were wasting syndrome, tuberculosis and anaemia. This result highlights the already advanced stage of immunodeficiency among patients in Burkina Faso when they start HAART. Testing patients for HIV and starting antiretroviral therapy earlier are necessary to further reduce the mortality of patients living with HIV. This study provides a solid evidence base with which future evaluations of HAART in Burkina Faso can be compared.

Keywords:

• HIV
• HAART
• mortality
• predictors

Acknowledgements

We would like to thank the “task force” of the Burkina Faso Country Coordinating Mechanism (CCM); the Executive Secretariat of the National Aids Council(S/P-CNLS); the financial unit of GFATM of Burkina Faso; MACRO, Inc.; WHO Headquarters; Médecins Sans Frontières (MSF)-Luxembourg; CANDAF and the medical staff in the local health centres. MACRO, Inc. and the WHO provided technical support. Thanks to Donna Riley and Naomi Ruff for translation and editing support.

This study was funded by the Global Fund to Fight AIDS, Tuberculosis and Malaria for district comprehensive assessment (letter SP/CNLS-GFATM, convention SP/CNLS-IRSS), MACRO, Inc. (Subcontract Number: 31463-7S).

The research protocol was reviewed by the ethical committee for health research of Burkina Faso.