Subgroup analysis and heterogeneity
| Information: | Borislava Mihaylova |
|---|
An important area of methodological research concerns subgroup analysis and heterogeneity. It is common practice in economic evaluations alongside clinical trials to present cost-effectiveness for a comparison of the treatment and control arms of a trial. However, where there is substantial heterogeneity among patients it may be appropriate to model that heterogeneity directly in order to identify subgroups of patients for whom treatment is or is not cost-effective.
In the cost-effectiveness analyses of the Heart Protection Study, we have demonstrated an approach in which the trial population was stratified into multivariate risk groups. The relative risk change in vascular events and costs due to treatment was shown to be similar across the risk groups, and therefore the overall trial estimates of relative treatment effects were applied to the absolute risk in each risk group, resulting in a clear gradient in cost-effectiveness across the vascular disease risks studied (published in Lancet by the Heart Protection Study Collaborative Group 2005).
A decision analytic cost-effectiveness model was then developed based on the individual participant data, that permits the extrapolation of the participants survival and healthcare costs beyond the end of the study. The parametric survival models of cardiovascular events that were in the basis of the decision model allowed for the propagation of the impact of heterogeneity and uncertainty into cost-effectiveness parameters (i.e. quality adjusted survival and lifetime healthcare costs) and overall cost-effectiveness. This is reported in a paper on the lifetime cost-effectiveness of simvastatin published in the British Medical Journal (Heart Protection Study Collaborative Group 2006).
Publications
Mihaylova, B, Briggs, A, Armitage, J, Parish, S, Gray, A, and Collins, R (2006). Lifetime cost-effectiveness of simvastatin in a range of risk groups and age groups derived from a randomised trial of 20,536 people. BMJ 333(7579):1145-8.
Mihaylova, B, Briggs, A, Armitage, J, Parish, S, Gray, A, and Collins, R (2005). Cost-effectiveness of simvastatin in people at different levels of vascular disease risk: economic analysis of a randomised trial in 20,536 individuals. Lancet 365(9473):1779-85.
