Multidimensional performance assessment using dominance criteria

Reaching judgements about hospital performance is difficult for two main reasons. First, there is no single measure of performance. This is because hospitals have different objectives that they are expected to achieve. These objectives cover access, safety and affordability, meaning hospitals will be judged on such things as waiting times, the safety and quality of care and their costs.

Second, people might value these objectives differently. This makes it difficult to construct an overall measure of performance that everybody would be happy with.

In this paper we propose a way to get around these problems. Our approach avoids constructing an overall performance measure. Instead, performance is judged against each objective separately. But we allow for the possibility that objectives may be related: if a hospital performs well against one objective, they may also perform well against another.

We also overcome the problem that people value objectives differently. For our approach all we need to know is that everyone agrees that either “more is better” or “less is better” for each objective. This means that we can say that a hospital performs well overall if it performs better than others across all objectives.

We apply our approach to study the performance of providers of hip replacement surgery in the English NHS. We consider four objectives: waiting times between referral and treatment; length of stay; unplanned readmission within 30-days of discharge; and post-operative health status as reported by patients themselves.

We find that if providers perform well against one objective, they tend to perform well against another. Indeed, some providers excel with respect to all measures of performance. In contrast, a small number of providers perform poorly across the board. Those providers that perform well tend to be privately operated independent sector treatment centres, which focus almost exclusively on providing just hip (and, sometimes knee) replacement.

Our approach can be applied to measure the performance of other organisations that pursue multiple objectives and to consideration of more than just four objectives.

Full Report: CHE Research Paper 115 (PDF , 1,533kb)

Other papers in the CHE Research paper series can be found at: In house publications