March 8–11, 2020


Conference Dates and Location: 
February 23-26, 2015 | Seattle, Washington
Abstract Number: 

Population Viral Load in Three High HIV Prevalence Settings in Sub-Saharan Africa


David Maman1, Helena Huerga1, Gilles Van Cutsem2, Irene Mukui3, Benson Chilima4, Beatrice Kirubi5, Ruggero G. Giuliani2, Elisabeth Szumilin6, Charles Masiku7, Jean-François Etard1
1 Clinical Research, Epicentre/Médecins Sans Frontières, Paris, France. 2 Médecins Sans Frontières, Cape Town, South Africa. 3 National AIDS and STDs Control Program, Nairobi, Kenya. 4 Community Health Sciences Unit, Ministry of Health, Lilongwe, Malawi. 5 Médecins Sans Frontières, Nairobi, Kenya. 6 Médecins Sans Frontières, Paris, France. 7 Médecins Sans Frontières, Lilongwe, Malawi.

Abstract Body: 

Background: Viral load (VL) is one of the key factors for HIV transmission. However, at population level, little is known about viral load distribution, especially among those undiagnosed or diagnosed but not on treatment. Critical data to better identify those most at risk of transmitting HIV are needed, particularly in sub-Saharan Africa where most of the transmission occurs.

Methods: Three population-based surveys of persons age 15 to 59 were conducted in Ndhiwa (Nyanza, Kenya), Chiradzulu (Malawi) and Eshowe (Kwazulu-Natal, South Africa) between September 2012 and November 2013 to assess HIV incidence and cascade of care. Each individual who agreed to participate was interviewed and tested for HIV at home. All HIV-positive were tested for VL and CD4, regardless of their ART status. T-test was used to compare means VL (in log10 copies/ml). Multivariate linear regression models were fitted to evaluate factors associated with VL among individuals not on ART.

Results: In total 9,802 houses were visited and among 21,782 individuals eligible, 19,006 (87.5%) were included and tested for HIV. Of the 4,117 individuals who tested positive, 3,938 (95.7%) had their viral load assessed. Population viral suppression (VL<1,000copies/ml) was higher in Malawi (61.9%, 95%CI 58.9-64.5) and South Africa (57.1%, 95%CI 54.5-59.6) than in Kenya (40.0%, 95%CI 37.5-42.6).
Of individuals not receiving ART, overall mean VL was higher among men compare to women (4.61 vs 4.17log cp/mLvs, p<0.01) and among those with CD4 between 500 and 749 cells/µL compared to CD4>750cells/µL (4.23 vs 3.72log cp/mL, p<0.01) but was similar across age groups (4.27, 4.37 and 4.22 log cp/mL for age 15-29, 30-44 and 45-59, respectively, p=0.18). Using the multivariate model, men had a VL higher than women (+0.33 log10 cp/mL, 95%CI 0.22-0.43, p<0.01). VL increased with decreasing CD4. CD4 between 500 and 750 CD4 cells/µL was associated with higher VL compared to CD4>750cells/µL (+0.47 log10 cp/mL, 95%CI 0.33-0.61, p<0.01).

Conclusions: Among individuals not receiving ART, those with CD4 500-750 cells/µL and men had higher viral load and could be at higher risk of transmitting HIV compared to those with CD4>750 cells/µL and women. Targeting men for HIV testing and treatment and ART initiation at higher CD4 thresholds (750 cells/µL) could contribute to decrease HIV transmission.


Session Number: 
Session Title: 
Reaching Populations: Demonstrating Impact
Presenting Author: 
Maman, David
Presenter Institution: 
Epicentre/Médecins Sans Frontières