The cost and effectiveness of achieving universal HIV treatment coverage in Africa: a modeling analysis of scaling up “treat all” in Zambia

Background: To achieve “90-90-90” targets, many countries have adopted “treat all” with no eligibility threshold for ART. We modeled the impact of adopting treat all in Zambia on total cost and HIV incidence and mortality, to inform policy choices and resource mobilization.
Methods: We used existing Zambian Spectrum AIM and GOALS models and new data on HIV program coverage, ART effectiveness, and current local unit costs to estimate the impact and costs of implementing treat all under three scale-up scenarios: (1) “baseline” (500 CD4 threshold for ART eligibility, treatment coverage of 70% by 2020 and 82% by 2030); (2) “90-90-90” (treat all, reaching 81% in 2020 and 90% in 2030); (3) “Fast Track” (treat all, reaching 90% in 2020 and 95% in 2030, with scaled-up prevention interventions).
Results: Figure 1 presents resulting changes in (i) ART patient numbers, (ii) costs, (iii) HIV-related deaths, and (iv) HIV infections. Adult patient numbers increase by 45% and 60% by 2020 in scenarios 2 and 3, respectively. New infections and deaths both decline under scenarios 2 and 3, averting 3,000-5,000 more deaths per year and 20,000-30,000 more new infections per year than scenario 1 by 2030. Treatment costs rise from $261 million/year in 2015 to $398 million, $411 million or $443 million by 2020 in the three scenarios, respectively. Annual costs for scenarios 2 and 3 plateau around 2025 and then fall below baseline due to reduced transmission. Compared to baseline, the incremental cost per infection averted over the period was $281 for Fast Track and (-$466) (cost saving) for 90-90-90.


Zambian Cost & Effectiveness of Treat All scale-up
[Zambian Cost & Effectiveness of Treat All scale-up]


Conclusions: Scaling up treat all in Zambia has the potential to save thousands of lives and reduce new infections, but costs will increase dramatically. Domestic and international financing initiatives will need to be explored, with greater efforts to ensure sustainability.

T. Guthrie1, C. Moyo2, A. Kinghorn3, C. van Rensberg4, J. Kuenhle5, W. Kaonga2, L. Hehman Soares6, M. Kamanga7, G. Sinyangwe5, J. Stover8, L. Long4, S. Rosen9
1Health Economics and Epidemiology Research Office, WITS University, Johannesburg, South Africa, 2Ministry of Health, Zambia, Clinical Services, Lusaka, Zambia, 3WITS University, Perianatal HIV Research Unit, Johannesburg, South Africa, 4WITS University, Health Economics & Epidemiology Research Office, Johannesburg, South Africa, 5USAID, Lusaka, Zambia, 6CHAI, Lusaka, Zambia, 7EQUIP, Lusaka, Zambia, 8Avenir Health, Washington, United States, 9Boston University, Boston, United States