What to do with future unrelated costs in health economic analysis?

Presenter: Andreas Gerber

Abstract

“... if a treatment causes a woman to live long enough to give birth to children who would not otherwise would have been born, the causality rule would dictate that their consumption be included, as well as their children’s consumption costs, and their children’s children’s consumption costs.” (Nyman 2004). There is a broad consensus that all treatment-related or disease-specific future costs should be included in such a way that neither alternative should be biased. However, the obsession with the inclusion of all so called unrelated future costs in the health economic evaluation of a treatment can lead to absurdities and is much more controversial. Unrelated future health care costs are all non-treatment or non-disease specific costs which arise from the fact that an individual lives longer due to an intervention. These costs in later years of life arise from keeping individuals alive longer. The question is, how much of unrelated future costs should be assigned to an intervention? In order to avoid “unduly favorable cost-effectiveness ratios,” it must be carefully considered which additional costs and productivity gains should be included for interventions in the first years of life, as even some negative consequences of immunizations were only recently detected such as unfavorable effects of varicella immunization on the elderly (Brisson and Edmunds).

Therefore, we analyzed all current health economic evaluations of immunizations and assessed them for inclusion of future disease-specific as well as unrelated costs and productivity gains and time horizons. Also, we scanned the studies for whether and how they took into account uncertainty in terms of effects on populations not immunized.

On the grounds of this review, some of our suggestions for the future of health economic analysis are:

1) We should not use average costs (cp Gandjour 2008), but percentiles in order to duly assess which subgroups would benefit from some interventions.

2) Correct mathematics may not solve the ethical question of what costs to include in case of very early interventions for chronically ill or disabled children.

3) We should ask whether health economic evaluation is ancillary to health policy or a means in itself. If it should be regarded ancillary, the health economic analysis of interventions in children and adolescents and especially for chronically ill or impaired children should be done independently from a comparison to all other interventions in adults.

Brisson M, Edmunds WJ (2003) Varicella vaccination in England and Wales: cost-utility analysis. Arch Dis Child 88: 862-69.

Gandjour A (2006) Survivor costs in cost-effectiveness analysis. Ann Internal Med 144(7): 534-35.

Nyman JA (2004) Should the consumption of survivors be included as cost in cost-utility analysis? Health Economics 13: 417-27.