Purpose: Given that cost-effectiveness analysis is based on an efficiency motive it makes major assumptions on peoples preferences for the allocation of resources and health outcomes. This study examines distributive preferences for health, especially the role of fairness, in order to counter the weight often given to cost-effectiveness data.

Methods: A systematic search and review of the empirical literature on equity and distributive preferences in health care resource allocation decisions has been undertaken. This has been supplemented with a general review of the guidance issued by the UK National Institute for Health and Clinical Excellence (NICE). Studies reporting attitudinal, experimental and analytical contributions have been considered. A detailed narrative review, with an examination and interpretation of current empirical findings in the context of fairness, has been undertaken.

Results: Empirical findings indicate that efficiency should be considered alongside other important distributive concerns. Whilst it is difficult to find general support for the role of characteristics such as age, gender, social status, and lifestyle, there is some support from the evidence base for a distributive preference to favour the most severely affected groups, and a preference to favour life-saving versus life-enhancing interventions. However, the literature is difficult to disentangle. The findings from attitudinal studies are difficult to weigh up in the absence of a choice scenario or opportunity cost, and much of the experimental work using choice experiments can be characterised as theoretical and/or context specific often presenting findings from simple surveys on convenience samples. This expanding literature is helpful but it may leave policy makers wondering what it is these studies are telling them about distributive preferences.

Conclusions: This research places cost-effectiveness and health gain as important inputs to many health care decisions, especially funding decisions for new and existing treatments. However, the maximisation of health gains is not supported empirically as the sole decision making criterion. The evidence base strongly supports a view that the distribution of health outcomes matters. What is needed is some idea of when/how and to whom it matters most. A preference for fairness is suggested here as a focus for attention in the allocation of health care resources. This study points to empirical work that may offer some help to those policy makers faced with difficult choices.