Purpose: To develop a simple and robust relative performance measurement framework for comparison of multiple alternative strategies in health technology assessment that can be consistently extended to relative efficiency measurement of providers in practice.

Methods: Currently, when multiple treatment strategies are compared in the incremental cost-effectiveness plane, dominance is identified by the ability to move ‘south-east' to an ‘efficiency frontier', which represents linear combinations of strategies for which there is no such movement possible. However, with such south-east movement, performance does not improve in expanding from, or contracting to a vertex, and hence the incremental cost effectiveness plane does not allow efficiency (ratio) measures with radial properties. A linear transformation where effects are framed as disutility events and performance measured in the cost disutility plane is demonstrated to allow radial properties while retaining an underlying objective of maximising net benefit, under a correspondence theorem. Costs and effect data for six alternative treatment strategies for treating gastro-oesophageal reflux disease (GORD) are used to illustrate advantages of constructing efficiency frontiers and comparing net benefit in the cost-disutility plane.

Results: Radial properties in measuring performance in the cost-disutility plane but absent in the incremental-cost effectiveness plane allow technically simpler construction of efficiency frontiers and identification of dominance (reduction in costs and disutility possible).These properties also allow estimation of degree of dominance (analogous to technical efficiency) as equi-proportional reduction in costs and disutility possible, 0 for strategies on the frontier and positive for extended dominated strategies (e.g. 0.04 and 0.68 in the case of GORD). Differences in net benefit at a given willingness to pay are represented by distances between isocost curves, with net benefit maximised at the point of tangency to the efficiency frontier (analogous to economic efficiency).

Conclusions: Performance measurement in the cost-disutility plane provides technical and conceptual advantages in comparing multiple strategies in HTA, related to radial properties which are absent in the incremental cost effectiveness plane. These radial properties in the cost-disutility pane allow the framework to be naturally extended to efficiency measurement of health care providers consistent with evidence-based medicine.