IN THE HIGH COURT OF SOUTH AFRICA (WITWATERSRAND LOCAL DIVISION) In the matter between TREATMENT ACTION CAMPAIGN AND OTHERS Applicants and MINISTER OF HEALTH AND OTHERS Respondents AFFIDAVIT: PROF N NATTRASS I, the undersigned NICOLI JEAN NATTRASS hereby make oath and state: 1 I have previously submitted an affidavit in this matter. 2 I wish to respond to the affidavits filed in reply on behalf of the Respondents, to the extent that they address matters relevant to my previous affidavit. I deal principally with the affidavit of Dr Ntsaluba, and refer where appropriate to the affidavits of other deponents. THE AFFIDAVIT OF DR NTSALUBA Ad para 128 3 The cost-effectiveness of reducing mother to child transmission (MTCT) of HIV/AIDS in South Africa and other African countries has been shown in a number of studies, all of which draw on empirical data. 4 My original affidavit cited these studies. 5 It is therefore not correct that the Applicants rely on only three studies, as stated by Ntsaluba in para 128). 6 I know of no study which concludes that reducing MTCT (through the AZT 'Thai' drug regimen or the Nevirapine 'HIVNET' regimen – with or without breastfeeding) is not cost-effective. 7 Since I wrote my previous affidavit, I have noticed that the basic costing data I have used is similar to that presented in a recent paper describing the implementation of a MTCT prevention site at Hlabisa in rural KwaZulu-Natal. The Hhlabisa study includes more cost categories (such as insurance, education materials, ongoing training, technical assistance, petrol, maintenance, packaging costs etc) than I used in my study. 8 If my estimates are adjusted to take into account such additional costs, my model still concludes that it would be cost-saving to introduce MTCT reduction programmes – even in outlying rural areas where problems of transport and co-ordination are particularly challenging. Ad para 133 9 Dr Ntsaluba states that none of the cost-effectiveness studies take into account the costs to mothers (such as travelling to health facilities, purchasing and sterilising bottles etc). 10 However, he fails to consider the costs to mothers of travelling to and from hospitals with their HIV+ children, and of missing work to look after sick HIV+ children. Such costs would be substantially in excess of the costs associated with participating in an MTCT programme. 11 My original affidavit was limited to the question of direct costs to the public health sector. If it had included private/social costs, my conclusions would have been even stronger in favour of the cost-saving impact of this method of reducing MTCT. 12 Such an analysis would not be able to quantify the emotional trauma which is experienced by mothers of HIV+ children. Ad para 137 13 As I stated in my previous affidavit, my conclusions are robust even when some of the underlying cost estimates are varied by 25 percent simultaneously. The effect of changing these estimates is dealt with specifically in that affidavit. 14 My conclusions remain robust even if the period of substitute feeding is extended from six months to twelve months. 15 In para 137.2 of his affidavit, Dr Ntsaluba suggests that the period of substitute feeding in a MTCT reduction programme should be extended beyond six months. 16 If we extend the period to 12 months (and include a further charge of R50 per mother to cover the cost of bottles), then the model estimates that the government still saves substantial amounts of money. Using the same costing model for 1000 pregnant women (of whom 245 are assumed to be HIV+) that was presented in my first affidavit, the following best estimate results are obtained: • R201 000 would be saved if the AZT (Thai) MTCT regime with 12 months of breastfeeding was introduced; and • R227 000 would be saved if the Nevirapine (HIVNET) MTCT regime with 12 months of breastfeeding was introduced. 17 In other words, the conclusion that the government will save resources by introducing MTCT reduction programmes is robust even when some of the input assumptions are changed. Ad paragraph 227 18 It may be that Dr Ntsaluba has not fully understood my calculations, the basis on which I made them, and my conclusions. 19 Dr Ntsaluba states that my model 'only estimates the cost of treatment'. I do not know whether by this he implies that I have excluded other relevant operating costs. If so, this is not correct. 20 In my calculations, I have used South African data drawn primarily from the Western Cape Department of Health and the Western Cape HIV/AIDS Directorate. The references were set out in my previous affidavit. 21 In my calculation of the costs, I have included estimates for: ? site costs (transport, stationery, phones, photocopies); ? project manager costs; ? the cost of mentoring nurses; ? the cost of drugs; ? the cost of tubes, needles, the Rapid test and the Smartcheck confirmatory test; ? the cost of paying counselors; and ? in the case of two of the calculations, the cost of six months of infant formula (substitute feeding) 22 Because of the sources on which I have drawn, these costs are broadly in line with international evidence from developing countries, as well as costs incurred in pilot and research sites in South Africa (including from rural areas). And, as noted in paragraph 8 above, if more recent and more comprehensive costing data is used in my model, the overall policy conclusion remains unchanged. 23 Where South African evidence and data is not available, I have made use of research from other developing countries – particularly Southern Africa. 24 Dr Ntsaluba's apparent doubts, expressed in para 227.2.2, about my assumptions regarding infant formula feed, are not well-founded. My study draws on evidence (available on www.unaids.org) collected in developing countries on the risk of infant mortality due to substitute feeding. 25 The risks associated with substitute feeding will obviously vary across South Africa. However, given South Africa's international development ranking, it is reasonable to use the information I had at my disposal when estimating the average increased risk of infant mortality due to substitute feeding. More detailed and specific data would certainly have been preferable. However, it should be noted that even if we were to multiply the number of children who can be expected to die as a result of substitute feeding by ten (i.e. a most unlikely scenario), my model predicts a marginal increase in total costs, and the cost per child saved rises by less than 20 percent in both costing scenarios involving substitute feeding. In other words, MTCT reduction programmes with substitute feeding are cost-effective and cost-saving even when infant mortality is assumed to be extraordinarily high as a result of substitute feeding. While ensuring safe water for all is a desirable objective, it is inappropriate to include in my calculation the capital costs of ensuring safe water for all (as apparently suggested by Dr Ntsaluba in para 132). 26 I of course accept that more detailed and up-to-date information about costs and benefits of MTCT reduction programmes (and of the costs of treating HIV+ children) would improve my calculations. However, as the key data is in line with international and local experience – and given that the cost sensitivity analysis indicates a very robust finding – I am confident that my policy conclusions will not be over-turned by new data from the research sites. 27 My projections regarding cost-savings are therefore not 'speculative'. They are the result of a robust calculation (complete with a cost-sensitivity analysis) based on what Dr Ntsaluba appears to accept is the best available information. 28 As I have pointed out above, I have used detailed South African data on the costs of MTCT reduction – and hence have a 'good sense of the totality of the costs necessary to avert those (HIV) infections'. Ad para 145 29 Dr Ntsaluba states (para 145.2) that because the full cost implications are not yet known, the government cannot expand the existing programme of reducing MTCT beyond the research sites. A similar position is advanced by Dr Simelela. 30 This position is untenable when one considers the enormous amount of research which has been undertaken into the costs and benefits of reducing MTCT of HIV/AIDS. I know of no other single health intervention in South Africa that has has been more researched from the perspective of cost- effectiveness. Are resources available? 31 Dr Ntsaluba states repeatedly that there are inadequate resources for a more comprehensive MTCT reduction programme. Dr Simelela and the Provincial deponents express the same view in their affidavits. They point in particular to the difference between their conditional allocation for HIV/AIDS and estimated costs of starting and running a MTCT reduction programme. 32 However, no counter-evidence has been offered to the conclusion in my original affidavit – namely that not only is a MTCT reduction programme cost- effective (as is accepted by Dr Ntsaluba in para 131) but it is also probably cost-saving. What this means, in effect, is that the government wastes resources by not introducing a MTCT reduction programme. 33 The start-up and running costs of most MTCT reduction programmes will probably be recovered over a reasonably short period. This is because of the reduced costs incurred for treating children born HIV positive. 34 Dr Ntsaluba implies at para 164 that an AZT-based programme would be unaffordable. An AZT-based programme would be more expensive than a Nevirapine-based programme. This is so even if one disregards the fact that Nevirapine is available free, and includes a cost for the Nevirapine in the calculations. 35 However, my study clearly estimates that an AZT-based programme, too, is not only affordable but cost-saving. This is the case for all four programmes that I costed (the AZT Thai regime with and without breastfeeding; and the Nevirapine HIVNET regime with and without breastfeeding). I understand that no evidence has been offered to contradict this conclusion. 36 Dr Ntsaluba draws attention at various points to the fact that there are competing health concerns. Again, this ignores the cost-saving which would result from a comprehensive programme to reduce MTCT of HIV/AIDS. 37 My analysis demonstrates that if the government introduced a more comprehensive MTCT reduction programme, there would be fewer HIV+ children in the paediatric wards and clinics, and hence more resources (human and financial) available for other health needs. 38 In para 127.3 Dr Ntsaluba states 'Clearly additional resources will be needed in the short-term. In the longer term, of course, this may be compensated by the cost of care averted, but this does not deal with the immediate problems'. 39 Clearly training and organisational costs will need to be covered in the short term – and in some cases, additional space may need to be provided. In my previous affidavit, my best estimate calculations indicated that the government could save between R171000 and R341000 (depending on what MTCT programme was introduced) for every 1000 pregnant women presenting at public health facilities. 40 These cost savings are substantial. They should be able to cover the cost of training the counselor (one counselor can counsel 1000 women in a year) and providing a small consulting room (even if only in a prefabricated outbuilding – as is currently being used for such purposes in Khayelitsha). 41 It is thus reasonable to conclude that the start-up costs of most MTCT sub- programmes are likely to be covered within the first two years of operation (and probably within the first year of operation). Detailed data on start-up costs from the Hlabisa study supports this contention. Including this data in my model does not change my conclusions: in all four MTCT programmes, start-up costs are recovered in the short-term. 42 A common statement in the responding affidavits is that the supporting infrastructure and services necessary for a MTCT reduction programme are simply not available and not affordable in the short term. Dr Ntsaluba states (para 116 and 117) that, like specialised cardiac and neuro-surgical care, MTCT programmes should thus only be available at specific (limited) sites. This argument lacks any economic foundation. A cost-saving MTCT reduction programme, which attends to the needs of very large numbers of people, can not be compared with a far more complex intervention such as highly expensive neuro-surgical care, which attends to the needs of relatively few people. Providing highly expensive care to a relatively small number of people obviously does not produce the same cost-benefits as providing more basic care to a very large number of people, many of whom may require extended treatment by the health care system if this preventative measure is not adopted. 43 The Respondents point to shortages of personnel in the health sector. In this regard it is worth noting that provincial governments reduced employment in the public health sector by an average annual rate of 3,3% between 1998/1999 and 2000/01. This has put pressure on the delivery of health services – pressure that could have been alleviated from provincial government surpluses. Provincial finances were in substantial surplus in 1999/2000 and 2000/01. The decision to prioritise surpluses over (for example) health services was not forced on governments by any absolute shortage of resources. PROF NICOLI NATTRASS SIGNED AND SWORN TO BEFORE ME AT ON NOVEMBER 2001, THE DEPONENT HAVING TAKEN THE OATH IN THE PRESCRIBED MANNER COMMISSIONER OF OATHS Geffen, N. 2001, Cost and Cost-Effectiveness of Mother-to-Child Transmission Prevention of HIV (TAC Briefing Paper), www.tac.org.za/mtctcost.rtf; Skordis, J. and N. Nattrass. 2001. What is Affordable: The Political Economy of Policy on the Transmission of HIV/AIDS from Mother to Child in South Africa. Paper presented to the AIDS in Context Conference, University of the Witwatersrand, April 2001; Wilkinson, D., Floyd, K., Gilks, C. F., 1999, A National Programme to Reduce Mother-To-Child HIV Transmission is Potentially Cost Saving: Evidence from South Africa., Medical Research Council; Wilkinson, D., Floyd, K., Gilks, C. F., 2000, National and Provincial Estimated Costs and Cost Effectiveness of a Programme to Reduce Mother-To-Child HIV Transmission in South Africa., SAMJ 90 (8), pp. 794-797; .Marseille, E., Kahn, J. G, Saba, J., 1998., Cost-Effectiveness of Antiviral Drug Therapy to Reduce Mother-To- Child HIV Transmission in Sub-Saharan Africa. AIDS 1998, 12, pp. 939-948; Marseille, E., Kahn, J. G., Mmiro, F., Guay, L., Musoke, P., Fowler, M. G., Jackson, J. B., 1999, Cost Effectiveness of Single Dose NVP for Mothers and Babies to Decrease Vertical HIV-1 Transmission in Sub-Saharan Africa., The Lancet 354 (9181), pp. 803-809; Soderlund, N., Zwi, K., Kinghorn, A., Gray, G., 1999, Prevention of Vertical Transmission of HIV: Analysis of Cost Effectiveness of Options Available in South Africa., British Medical Journal 318, pp. 1650-1656 (19 June); Hensher, M., 2000, Confidential Briefing: The costs and effectiveness of using NVP or AZT for the prevention of mother-to-child transmission of HIV – current best estimates for South Africa., Health Financing & Economics. Galbraith, D., Bennish, M., The Child Health Team, August 2001, Reducing Mother-to- Child Transmission in Hlabisa District: Incremental Program Costs. 13 13