Most health decision aids (HDAs) do not formally incorporate patient health preferences. To address this limitation, we developed software technology (“Health e-Decision”) to build HDAs designed to support healthcare decisions faced by older adults. The Health e-Decision design consists of a domain knowledgebase, decision-analytic model, data-access interface to electronic health records, and an interactive Web-based interface. For the preference assessment module, we evaluated FLAIR, a multimedia utility assessment, for usability among older adults and its adaptability within a HDA for the decision of taking an anticoagulant following atrial fibrillation.

552 older adults (mean age=75, range 65-92; 61% female; 31% non-white) reported Standard Gamble (SG) utilities for hypothetical health states of dependency in six Activities of Daily Living using FLAIR. Participant reactions to the program were also recorded using a scale from 1 (strongly agree) to 5 (strongly disagree). Length of time to complete the survey was also recorded. For further description of FLAIR design features, see Sims et al., Proc AMIA 2005. We mapped FLAIR health states into our decision-analytic model for atrial fibrillation and analyzed FLAIR's software code, to determine whether the health preference assessment could be integrated into e Preference.

87% (480) of respondents agreed that the computer was easy to use, 74% (408) enjoyed using the FLAIR program, and 84% (464) thought the questions were clear and easily understood. Only 17% (94) thought the questions were confusing and just 14% (77) preferred something other than the computer to record their responses. Mean time spent on explanation of the SG and rating of health states was 11 minutes. We were able to encode the health states modeled within FLAIR into e-Preference and reuse the related interfaces for utility assessment to create the HDA for atrial fibrillation.

Reactions to the computerized assessment were positive overall and the average time spent on rating health states was minimal, indicating that use of the FLAIR program to elicit health utilities is feasible among the older adult population. Our results indicate that FLAIR is a viable foundation for an automated method for formally incorporating patient preferences into Health e-Decision. Such an approach to HDAs is needed to ensure that patient decision-support methods can be tailored to individual preferences, therefore optimizing health care choices made by patients and physicians.