Nathan Geffen, TAC
In an oral presentation, Leigh Johnson of the Centre for Actuarial Research at the University of Cape Town presented an analysis of South Africas HAART requirements.  Johnson is one of the developers of the Actuarial Society of South Africas AIDS models, including ASSA2003. 
Albeit that HAART has been available at a few public sector research and pilot sites for about a decade, the implementation of treatment in the public sector began in 2004. The number of people on treatment in both the private and public sector has risen from about 50,000 at the start of the programme in mid-2004 to about 550,000 in mid-2008. 
This is based on Department of Health statistics as well as Johnsons research of private sector and NGO treatment numbers. The Department of Health data is subject to limitations. The most glaring is that five of the countrys nine provinces (Eastern Cape, Gauteng, Kwazulu-Natal, Limpopo and North West) only track the number of people who initiated treatment, not the number currently on treatment. Patients lost to follow up and deaths are therefore included in its count. To correct this, Johnson calculated a rate of retention in these provinces based on data from the Western Cape. He also checked the quality of his adjusted estimates against sales data from the pharmaceutical company supplying the bulk of the state tender.
Johnson has calculated that in mid-2004, the public sector treated less than 20% of HAART patients (the remainder were treated by the private sector and NGOs). This had increased to nearly 80% by mid-2008. He also calculated unmet need.
- The number of adults with untreated clinical AIDS as at mid-2008 was 430,000. Using this criterion, HAART coverage is 54%.
- The number of adults with untreated clinical AIDS or CD4 counts < 200, i.e. the Department of Health criteria, was 760,000, in which case coverage is only 40%.
- If the CD4 count criterion was changed to less than 350, i.e. according to the Southern African HIV Clinicians Society guidelines, then 1.8 million people were untreated. In this case coverage is a mere 22%.
- Using the Departments criteria, the province with the lowest coverage is the Free State (26%). The Western Cape, at 72% has the best coverage.
To determine future need, Johnson ran various scenarios through the ASSA2003 model. He used the Department of Healths estimates of the number of people on HAART, the District Health Barometers data on PMTCT and results from the Western Capes programme to calculate HAART effectiveness.
If the target of placing 80% of newly eligible patients on HAART is met by 2010 (80% is the target of the states National Strategic Plan) and current HIV incidence trends continue, then by 2011 (the end of the target period for the plan), just under 1.5 million people will be on treatment. More than 2 million people will be on treatment in 2014 and more than 3 million in 2020.
If instead, HIV incidence is halved (also a target of the states plan), then the consequences of this become more apparent the further into the future the estimates are projected. By 2020, half-a-million fewer people will be on treatment in this scenario.
Johnson demonstrated the substantial benefits of the HAART rollout. It conferred 24% fewer AIDS deaths in 2008, than if the programme had not been rolled out. This benefit is becoming more pronounced with time resulting in approximately 200,000 fewer deaths per year for the next decade. In 2008, there were 8% fewer maternal orphans under the age of 18 due to HAART. This becomes even more beneficial over the next 15 years, with nearly a million fewer orphans by the middle of the next decade.
Despite the actions of the Mbeki regime, particularly former Health Minister Tshabalala-Msimang, South Africa has rapidly scaled up its HAART rollout. This has been achieved because of the efforts of health workers, researchers activists and some civil servants. Nevertheless, as Johnson has demonstrated, the unmet need is substantial and growing. It will be challenging to meet the prevention and treatment targets of the National Strategic Plan.
Johnsons painstaking analysis of the number of people on HAART as of mid-2008 across all sectors must be considered the definitive estimate of coverage for South Africa. Yet the excellent work of the Centre for Actuarial Research continues to be limited by the quality of the data from the Department of Health on numbers of patients initiated, lost-to-follow-up, currently active and died on the PMTCT and HAART programmes. The Department must prioritise improving the monitoring and evaluation of the HAART and PMTCT programmes.
References 1. Johnson L. How big is the need for antiretroviral treatment? 4th South African AIDS Conference, Durban, 31 March – 3 April 2009. 2. See: http://www.tac.org.za/community/keystatistics 3. Department of Health. National Strategic Plan for HIV and AIDS Statistics. October 2008.