A critical issue in using these cost curves into the future is to what extent they might be stable. We know from experience that these general shapes by age and gender are persistent, not only over different schemes but also over time in the same funder (changing benefits are more important than time) and across countries (with some differences). Subsequent sections in this policy brief explore some of the philosophical assumptions that underlie whether the curves are in fact stable into the future.
Ideally, we would want to study the same curves in the public sector in South Africa but despite several attempts over a seven year period it has not yet been feasible to estimate any public sector curve. It is of course very difficult to use data from an under-resourced public service to predict cost in a future better-resourced system. This brings us back to attempting to use the excellent private sector data and adjust it to the total cost likely in a well-resourced public system.
In all likelihood the public sector curves have a similar shape to those shown in Figure 3, but with at least the following differences: lower costs for Under 1s (fewer ICU admissions and very low birth-weight babies are not resuscitated in the public sector); lower maternity costs (far fewer Caesarean sections); and higher HIV admissions and costs in the HIV years (a first attempt at this has been tried). The cost of chronic disease from age 40 onwards may also be affected by different patterns of disease and treatment regimens.
While we know that the mix of chronic diseases amongst the poor is different to that of the higher income groups in medical schemes, the total burden of disease (excluding HIV) might possibly be of about the same magnitude. This is a research question that needs further work and the results will have important implications for the use of private sector data to project public sector costs.
