 |
|
|
| |
Achieving the WHO/UNAIDS antiretroviral treatment 3 by 5 goal: what will it cost?
|
| |
| |
Juan Pablo Gutierrez, Benjamin Johns, Taghreed Adam, Stefano M Bertozzi, Tessa Tan-Torres Edejer, Robert Greener, Catherine Hankins, David B Evans
Lancet 2004; 364: 63-64
Division of Health Economics and Policy, National Institute of Public Health, Cuernavaca, Mexico (J P Gutierrez MSc, S Bertozzi PhD); Health System Financing, Expenditure, and Resource Allocation, WHO, CH-1211 Geneva 27, Switzerland (B Johns MPA, T Adam MD, T Tan-Torres Edejer MD, D B Evans PhD); Centre for Research and Education in Economics (CIDE), Mexico City, Mexico (S Bertozzi); Strategic Information, Social Mobilisation, and Information, UNAIDS, Geneva, Switzerland (R Greener PhD, C Hankins MD)
Correspondence to: Benjamin Johns johnsb@who.int
The "3 by 5" goal to have 3 million people in low and middle income countries on antiretroviral therapy (ART) by the end of 2005 is ambitious. Estimates of the necessary resources are needed to facilitate resource mobilisation and rapid channelling of funds to where they are required. We estimated the financial costs needed to implement treatment protocols, by use of country-specific estimates for 34 countries that account for 90% of the need for ART in resource-poor settings. We first estimated the number of people needing ART and supporting programmes for each country. We then estimated the cost per patient for each programme by country to derive total costs. We estimate that between US$5·1 billion and US$5·9 billion will be needed by the end of 2005 to provide ART, support programmes, and cover country-level administrative and logistic costs for 3 by 5.
In September, 2003, at the second UN General Assembly Special Session on HIV/AIDS, WHO and UNAIDS declared the lack of treatment in low and middle income countries to be a global public health emergency and launched the "3 by 5" initiative, which aimed to enrol 3 million people on antiretroviral therapy (ART) by the end of 2005. For this ambitious goal to be reached--starting from a base of fewer than 200000 patients on treatment--countries, donors, and multilateral agencies must know what resources need to be rapidly mobilised.
Since previous estimates of the cost of scaling up interventions against HIV and AIDS were made,1 further WHO-recommended treatment protocols for resource limited settings have been published.2 We aimed to estimate the financial costs needed to implement these protocols, by use of country-specific estimates for 34 countries that account for 90% of the need for ART in resource-poor settings. The 3 by 5 strategy includes standardised treatment protocols, simplified clinical monitoring and record keeping, best use of the existing health system resources, active involvement of communities and people living with HIV, and efforts to minimise the cost of drugs and diagnostics.2,3 No major changes to the health system infrastructure, the numbers of available personnel, or transmission of HIV as a result of ART are deemed likely in view of the short time frame.
We defined the number of people needing treatment as those expected to die within 2 years in the absence of ART. People needing treatment in 2004, for example, would be those expected to die before the end of 2006 without treatment. People stopping treatment or dying while on treatment would be replaced, so that 3 million people would be receiving ART at the end of 2005 (see http://image.thelancet.com/extras/04let5139 webappendix.pdf for more details).
The three main entry points for recruitment of eligible patients are tuberculosis clinics, inpatient and outpatient health facilities, and mother-to-child-transmission prevention programmes in antenatal care clinics. Drug regimens and testing procedures vary by entry point (see http://image.thelancet.com/extras/ 04let5139webappendix.pdf). We used two assumptions of the growth rates in coverage to reach the target: slower scale-up, reaching 10% of the target in 2004 and 90% in 2005; and more rapid scale-up, meeting 20% of the target in 2004.
Estimates of treatment costs took several factors into account. Enrolment of patients requires confirmation of positive HIV status by rapid testing and counselling. A doctor or nurse confirms clinical eligibility. To stabilise patients starting ART and ensure continued well-being while on therapy, patients are diagnosed and treated for opportunistic infections. Medicines for prevention of opportunistic infections and laboratory tests for suspected toxicity help to ensure successful ART, while patients who do not respond to treatment need palliative care.
Support costs at country level include training of existing health personnel, supervision of ART delivery, and remuneration of volunteers providing adherence support to patients. Universal precautions and post-exposure prophylaxis were included, as were costs for limited upgrades to laboratories and drug storage and distribution systems.
The number of facility-based visits for ART initiation and monitoring, frequency of adherence counselling and monitoring, laboratory tests for toxicity and estimated times associated with each of these activities were based on reports from a consensus meeting.2 Quantities of inputs used in other interventions were based on recommendations in published guidelines supplemented by expert opinion.
Data on the costs per patient of interventions and activities were obtained from studies done in sub-Saharan Africa or Asia or, where no such data were available, extrapolated using methods described elsewhere.4,5 The web appendix includes details of assumptions, and estimated costs per patient of interventions and activities for Botswana (http://image.thelancet.com/ extras/04let5139webappendix.pdf).
We used two assumptions for the cost of drugs. The high cost option assumes6,7 US$304 for first line therapy (fixed-dose combination), $1108 for second line therapy, $706 for a switch in treatment due to toxicity, $505 for patients with tuberculosis ($353 for 6 months then first line for the remaining 6 months), and $831 for pregnant patients with tuberculosis ($679 for treatment switch of 6 months). The low cost option uses $140 per patient per year, the price negotiated by the Clinton Foundation for standard first-line treatment in selected countries, which might become more generally available. Drug costs for other categories are assumed to undergo a similar cost reduction ($510 for second line therapy; $325 for toxicity; $423 for tuberculosis; and $749 for pregnant patients with tuberculosis. See the web appendix at http://image.thelancet.com/extras/ 04let5139webappendix.pdf for further details).
The estimated costs for 2004-05 are summarised in the table for four combinations of assumptions of the rate of scale-up and pharmaceutical costs. The total estimated cost was between US$5·1 billion and US$5·9 billion for the 2 years. The figure shows the breakdown of total 2-year costs by category for scenario 2A in the table. Costs relating to patients account for more than 77% of the total costs, with purchase and provision of ART responsible for more than 43%. Universal precautions, post-exposure prophylaxis, and other programme costs account for less than 23%.
SCENARIO
| | 2004 | 2005 | Total | | 1A: | 10%/90% with higher drug costs | $2.0 | $3.8 | $5.7 | | 1B: | 10%/90% with lower drug costs | $1.9 | $3.2 | $5.7 | | 2A: | 20%/80% with higher drug costs | $2.2 | $3.7 | $5.9 | | 2B: | 20%/80% with lower drug costs | $2.0 | $3.1 | $5.1 |
10%/90% =10% coverage in 2004 and 90% in 2005; 20%/80%=20% coverage in 2004 and 80% in 2005 (more rapid scale-up). Table: Estimated costs of 3 by 5, 2004-05 (US$ billions)
Support activities include post-exposure prophylaxis and scale-up of preventing mother-to-child transmission of HIV. OI=opportunistic infections.
The primary purpose of these estimates is to provide a global aggregate target to guide short-run resource mobilisation. Precise country estimates are being developed to help local planning. The values presented here show how crucial it is to intensify resource mobilisation activities. This action is even more important for the long term, because the need for additional resources will continue to grow after 2005, as more patients survive and others are recruited.
Contributors
J P Gutierrez constructed the costing model including estimation of people receiving treatment each year, and contributed to the calculation of unit costs of interventions per country per year and total costs per intervention per year. B Johns estimated the programme-level costs, contributed to unit price estimates, and wrote the first draft of the paper. T Adam contributed to the construction of the costing model and the selection and analysis of the interventions, researched the treatment and care model, did entry point analysis, and contributed to unit price estimates. S Bertozzi contributed to the development of the costing model and estimates of number of people under treatment for each intervention and year and unit and total costs per intervention and year; he also helped research the treatment and care model. T Tan-Torres Edejer helped estimate unit prices and research the treatment and care model and validated the outputs. R Greener contributed to the estimation of intervention cost per country and total cost per intervention per year. C Hankins contributed to the estimation of people in need and receiving treatment for each intervention and year, and researched the treatment and care model. D B Evans helped develop the unit price and programme costs model and estimates, contributed to the selection and analysis of the interventions, and helped to validate the outputs. All authors contributed to the writing of the letter and read and approved the final draft.
|
| |
|
|
|
|
|
