Including quality attributes in efficiency measures consistent with net benefit: creating incentives for evidence based medicine in practice.

Evidence based medicine supports net benefit maximising therapies and strategies in processes of health technology assessment (HTA) for reimbursement and subsidy decisions internationally. However, translation of evidence based medicine to practice is impeded by efficiency measures such as cost per case-mix adjusted separation in hospitals, which ignore health effects of care. In this paper we identify a correspondence method that allows quality variables under control of providers to be incorporated in efficiency measures consistent with maximising net benefit. Including effects framed from a disutility bearing (utility reducing) perspective (e.g. mortality, morbidity or reduction in life years) as inputs and minimising quality inclusive costs on the cost-disutility plane is shown to enable efficiency measures consistent with maximising net benefit under a one to one correspondence. The method combines advantages of radial properties with an appropriate objective of maximising net benefit to overcome problems of inappropriate objectives implicit with alternative methods, whether specifying quality variables with utility bearing output (e.g. survival, reduction in morbidity or life years), hyperbolic or exogenous variables. This correspondence approach is illustrated in undertaking efficiency comparison at a clinical activity level for 45 Australian hospitals allowing for their costs and mortality rates per admission. Explicit coverage and comparability conditions of the underlying correspondence method are also shown to provide a robust framework for preventing cost-shifting and cream-skimming incentives, with appropriate qualification of analysis and support for data linkage and risk adjustment where these conditions are not satisfied. Comparison on the cost-disutility plane has previously been shown to have distinct advantages in comparing multiple strategies in HTA, which this paper naturally extends to a robust method and framework for comparing efficiency of health care providers in practice. Consequently, the proposed approach provides a missing link between HTA and practice, to allow active incentives for evidence based net benefit maximisation in practice.