Background: Randomized trials have shown that patient decision aids can modify users' preferred healthcare options, but research has yet to identify the attributes embedded in these tools that cause preferences to shift.
Objectives: The aim of this study was to investigate people's preferences as they used decision aids for 5 health decisions and, for each of the following: 1) determine if using the interactive Option Grid led to a pre-post shift in preferences; 2) determine which frequently asked questions (FAQs) led to preference shifts; 3) determine the FAQs that were rated as the most important as users compared options.
Methods: Interactive Option Grid decision aids enable users to view attributes of available treatment or screening options, rate their importance, and specify their preferred options before and after decision aid use. The McNemar-Bowker paired test was used to compare stated pre-post preferences. Multinomial logistic regressions were conducted to investigate possible associations between covariates and preference shifts.
Results: Overall, 626 users completed the 5 most-used tools: 1) Amniocentesis test: yes or no? ( n = 73); 2) Angina: treatment options ( n = 88); 3) Breast cancer: surgical options ( n = 265); 4) Prostate Specific Antigen (PSA) test: yes or no? ( n = 82); 5) Statins for heart disease risk: yes or no? ( n = 118). The breast cancer, PSA, and statins Option Grid decision aids generated significant preference shifts. Generally, users shifted their preference when presented with the description of the available treatment options, and the risk associated with each option.
Conclusion: The use of decision aids for some, but not all health decisions, was accompanied by a shift in user preferences. Users typically valued information associated with risks, and chose more risk averse options after completing the interactive tool.
Keywords: decision support intervention; option grid patient decision aids; preference shift.