Development of an instrument to elicit patient preferences for Person‐Centered Dementia Care – insights from a formative pilot study to design a dementia‐friendly Analytic Hierarchy Process‐Survey

Abstract

AbstractBackgroundPerson‐Centered Care (PCC) requires knowledge about patient preferences for multiple components (criteria) of care. The Analytic Hierarchy Process (AHP) is one approach to measure the importance of individual criteria from the patient perspective and to identify quantitative preferences for PCC. The AHP is expected to be suitable for People living with Dementia (PlwD) due to its simple pairwise comparisons with individual criteria of a complex decision problem. The aim of this pilot study was to design a dementia‐friendly assisted AHP‐survey with PlwD.MethodWe conducted two expert panels with n = 4 Dementia Care Managers and n = 4 physicians to ensure content validity, as well as individual “thinking aloud”‐interviews with n = 11 PlwD and n = 3 informal caregivers (CGs) to ensure face validity. Interviews followed a semi‐structured interview guide. Participants were asked to assess the survey’s appropriateness, coherence, and to provide information about the motivation for their answers. Main source of data were field notes. Interviews were recorded and partially transcribed. To follow an iterative constant comparative approach, data were analyzed and patient’s feedback incorporated on a continuous basis by two reviewers until the patients had no further comments and no struggles with completion of the survey. Survey data were analyzed with Microsoft®Excel and ExpertChoice Comparion®, to i.a. assess consistency ratios (CRs).ResultThree main categories with several sub‐categories emerged from data analysis: 1) content including clear introduction, description of (sub)criteria, homogeneity of criteria, appropriateness of (sub)criteria, inclusion of re‐test questions, sociodemographic questions to consider heterogeneity of respondents, 2) format including structure of an AHP‐survey for PlwD, sequence of pairwise‐comparison questions, length of survey, visual formatting (incl. AHP‐scale), explanation of survey procedure, assistance during survey, perspective during completion of survey, and 3) layout including easy wording, avoidance of long sentences, and use of ICONs as visual aids. Individual CRs ranged from 0.08 to 0.859, whilst the consolidated CR was 0.038.ConclusionOur study provides insights from a formative pilot study to design a dementia‐friendly AHP‐Survey. Consideration of our findings may contribute to content‐ and face validity, acceptability, feasibility and internal consistency, as well as reproducibility in future dementia‐focused quantitative preference research.