Cognitive overload? An Exploration of the Potential Impact of Cognitive Functioning in Discrete Choice Experiments with Older People in Health Care. Running Head: Impact of Cognition in Discrete Choice Experiments

Objectives: This exploratory study sought to investigate the impact of cognitive functioning on the consistency of individual responses to a DCE study conducted exclusively with older people. Methods: A DCE to investigate preferences for multidisciplinary rehabilitation was administered to a consenting sample of older patients (aged 65 years and over) following surgery to repair a fractured hip (N=84). Conditional logit, mixed logit, heteroscedastic conditional logit and generalised multinomial logit regression models were utilised to analyse the DCE data and to explore the relationship between the level of cognitive functioning (specifically the absence or presence of mild cognitive impairment as assessed by the minimental state, MMSE) with preference and scale heterogeneity. Results: Both the heteroscedastic conditional logit and generalised multinomial logit models indicated that the presence of mild cognitive impairment did not impact significantly upon the consistency of the responses to the DCE. Conclusions: This study provides important preliminary evidence relating to the impact of mild cognitive impairment upon DCE responses for older people. It is important that further research is conducted in larger samples and more diverse populations to further substantiate the findings from this exploratory study and to assess the practicality and validity of the DCE approach with populations of older people.