Background: The UNAIDS has proposed achieving 90% diagnosis of HIV-positive persons, 90% antiretroviral therapy (ART) coverage among those diagnosed, and 90% viral suppression among those on ART. ART initiation guidelines have primarily depended on CD4 cell count, with the World Health Organization (WHO) currently recommending ART initiation for persons with CD4≤500 cells/µL. However, HIV transmission, morbidity, and mortality are more closely linked with the level of virus—viral load (VL)—than CD4 count. The impact of CD4- versus VL-based ART criteria on HIV-associated outcomes has not previously been studied.
Methods: We reviewed the literature for studies assessing the association between clinical outcomes—HIV transmission, time to AIDS, HIV-associated mortality, and ART initiation—and CD4 count and VL. These measures were used to parameterize a compartmental mathematical model of HIV transmission in KwaZulu-Natal, South Africa, that was stratified by gender, CD4 count, VL, and ART status. The model was used to estimate the proportion of ART-ineligible persons with high viral load, and the proportion of HIV infections, deaths, and quality-adjusted life-years (QALYs) that could have been saved if ART had been initiated for those with high VL. These values were estimated from 2004 to 2014 and from 2015 to 2025. From 2015 to 2025, we also estimated the incremental cost-effectiveness ratio (ICER) of ART initiation at VL>10,000 copies/mL to CD4≤500 cells/µL.
Results: We estimate that in KwaZulu-Natal from 2004 to 2014, 35% of ART-ineligible patients by CD4 count had VL >10,000 copies/mL, and 12% had VL>50,000 copies/mL. With 30% of ART-ineligible persons with VL>10,000 copies/mL initiating ART, an additional 72% of HIV infections and 46% of HIV-associated deaths could have been averted, and 44% of QALYs gained. From 2015 to 2025, ART initiation at CD4≤500 cells/µL provides 5% more individuals with ART than VL>10,000 copies/mL. Using the VL criterion results in an ICER of $5,709 per HIV infection averted and $537 per QALY gained, whereas using the CD4 criterion results in an ICER of $7,851 per HIV infection averted and $612 per QALY gained.
Conclusions: By focusing on CD4 measurements rather than HIV viral load, past ART policies missed opportunities to fully utilize the benefits of ART to improve quality-of-life and prevent HIV transmissions. Future HIV prevention strategies should consider focusing on viral load rather than CD4 criteria in order to cost-effectively maximize public health impact.