OBJECTIVES: Individual preferences with respect to the possible outcomes of an intervention and its attributes may vary. Some patients may have a strong preference for maximising the probability of a response to treatment, whereas others may have a strong preference to avoid certain adverse events. Some patients may have a strong preference to avoid injections whereas others may wish to avoid frequent oral medication. However, the standard decision rules for cost-effectiveness analysis identify a single technology as maximising expected health benefit. Decision made on this basis - failing to recognise variation in preferences and circumstance - may be economically inefficient, fail to maximise patient utilities, and may worsen health inequalities.

METHODS: Using a hypothetical case study with three treatments A, B, and C, where C must always be offered regardless of cost effectiveness (Standard of Care) and A is most cost-effective option (highest NMB) and the Decision Maker’s objective is to maximise expected population (health related) utility we consider the impact of heterogeneity in individual preferences where individual utility is maximized when (fully informed) patients select treatment based on personal preference. We analysis outcomes allowing for all patient subpopulations in terms of preference rankings across a set of cases in terms of ranking of treatments in terms of costs in order to identify where should preference heterogeneity be accommodated by also allowing access to treatment B, alongside A and C.

RESULTS: We describe cases where accommodation of preference utility unequivocally increases population utility and where the results of ‘average’ cost effectiveness analysis vs. standard of care and where deliberation may be required

CONCLUSIONS: We should consider modifying the standard rules of cost-effectiveness analysis to accommodate preference heterogeneity