Complex Valuation: Applying Ideas from the Complex Intervention Framework to Valuation of a New Measure for End-of-Life Care

The UK Medical Research Council approach to evaluating complex interventions moves through development, feasibility, piloting, evaluation and implementation in an iterative manner. This approach might be useful as a conceptual process underlying complex valuation tasks. The objective of the study was to explore the applicability of such a framework using a single case study (valuing the ICECAP-Supportive Care Measure) and considering three key uncertainties: the number of response categories for the measure; experimental design; and the potential for using slightly different variants of the measure with the same value set. Three on-line pilot studies (n = 204, n = 100, n = 102) were undertaken during 2012 and 2013 with adults from the UK general population. Each used variants of discrete choice and best-worst scaling tasks; respondents were randomly allocated to different groups to allow exploration of the number of levels for the instrument (four or five), optimal experimental design and the values for alternative wording around prognosis. Conditional logit regression models were used in the analysis and variance scale factors were explored. The five-level version of the measure seemed to result in simplifying heuristics. Plotting the variance scale factors suggested that best-worst scaling answers were approximately four times more consistent than the discrete choice answers. The likelihood ratio test indicated there was virtually no difference in values between the differently worded versions. Rigorous piloting can improve the design of valuation studies. Thinking in terms of a 'complex valuation framework' may emphasise the importance of conducting and funding such rigorous pilots.