Preferences cannot be treated as epidemiological characteristics in person-centred care: a riposte

Abstract

We are delighted that Pedersen and colleagues have responded [1] to our implicit challenge [2] and welcome the opportunity to clarify our position in this Riposte. Essentially, it will reaffirm our position, but we will also identify important underlying sources of disagreement that surface in this Discrete Choice Experiment (DCE) debate. We will also raise issues concerning person-centred decision support, especially in the coming era of so-called ’precision medicine’, given that the requirements for this support was the prime motivation for the original paper.

We remain unconvinced that the quantitative group average results from a DCE have any serious role to play in Person-Centred Care (PCC). In our definition, PCC requires the individual's preferences to be formally and explicitly elicited at or near the point of the clinical decision in the most effective and practical way possible. Doing so is obviously the simple and direct means to address any informational asymmetry between doctor and patient in relation to the latter's preferences. Not doing so and endorsing the use of DCE group average results as ‘a starting point’ in a ‘clinician-guided’ procedure represents a threat to implementing fully Person-Centred Shared Decision-Making (SDM). We repeat our belief, in the light of the commentators’ contrary one, that this will constitute a major deterrent to the development and adoption of practical clinical decision support tools; ones that include the personalised preference elicitation at the point of decision that is an essential aspect of preference-sensitive, multi-criterial decision-making.