Conference abstract

Predictors of attrition in patients on anti-retroviral treatment in Ghana an analysis of National HIV Programme data

Pan African Medical Journal - Conference Proceedings. 2017:3(3).11 Dec 2017.
doi: 10.11604/pamj-cp.2017.3.3.114

Contact the corresponding author
Keywords: Antiretroviral therapy, HIV, attrition
Oral presentation

Predictors of attrition in patients on anti-retroviral treatment in Ghana an analysis of National HIV Programme data

Abdulai Marijanatu1,&, Stephen Ayisi Addo1, Akosua Baddoo1, Alfred Yawson2, Nibretie Workneh3, Zhao Jinkou3, Wiah Ekow1, Okae Ivy1, Rebecca Balira4

1Ghana Health Service, National AIDS/STI Control Program, Accra, Ghana, 2University of Ghana, School of Public Health, Accra, Ghana, 3The Global Fund to fight AIDS TB and Malaria, Geneva, Switzerland, 4National Institute for Medical Research - Tanzania, Dares Salaam, Tanzania

&Corresponding author
Abdulai Marijanatu, Ghana Health Service, National AIDS/STI Control Programme, Accra, Ghana


Introduction: the availability and use of antiretroviral treatment (ART) has made great progress in reducing the burden of HIV across the world. In view of this, countries have adopted strategies that will increase enrollment and retention of HIV positive individuals on ART. However, a good number of HIV positive patients drop out of treatment along the continuum of care which adversely affects national goals of reducing the disease burden. This study aimed at determining predictors of attrition in patients after initiating ART.

Methods: a retrospective cohort analysis of patients enrolled in ART between 2003 and 2015 across all ART centers in Ghana was conducted. Data was extracted from a nationwide Health Information Management Database and analyzed using STATA 13. Cox proportional hazards regression identified baseline factors associated with attrition. The probabilities of ART attrition after ART initiation were estimated by the Kaplan-Meier method.

Results: of the 78,816 clients registered in the database, attrition was among 32, 974 (42%) at the end of August 2015. Attrition rate was significantly lower among females compared to males [HR=1.20 (1.17-1.23)]. Attrition was however higher among patients with junior secondary education [HR=1.07(1.04-1.11)], who used medical insurance [HR=1.08(1.05-1.10)], who attended clinic at health centers [HR=1.51(1.32-1.73)] who were initiated into care at WHO clinical stage 4 [HR=1.35(1.29-1.41)] and patients who self-reported of missing 5 pills and above [HR=2.04(1.99-2.09)]. However, patients who had no formal education, funded their medical bills out of own pocket, attended clinic at Teaching/Regional hospitals, initiated into care at WHO stage 1 were patients who adhered to their ARVs. Additionally, attrition was low among those who had disclosed their status to their partners compared to those who did not [HR=1.07 (1.04-1.10)]. Similarly, patients attending Faith-based health facilities had a higher attrition rate [HR=1.07 (1.01-1.13)] compared to patients who used government owned, quasi-government and private owned facilities.

Conclusion: socio-demographic factors, clinical staging, patient adherence, source of funding, and disclosure were key predictors of attritions among patients on ART in Ghana. These findings provide a baseline guide to ART clinicians and programme managers to recognize key factors that may influence continuity in care for ART.