Improving the Adoption of Advance Directives in Primary Care Practices

Background: Oklahoma's Advance Directive completion rate is less than 10%. We compared the implementation performance of 2 advance directive forms to determine which form could be more successfully disseminated.

Methods: The implementation of the Oklahoma Advance Directive (OKAD) and the Five Wishes form were compared in an 8-month pair-matched cluster randomized study in 6 primary care practices. The outcomes measured during the 22-week implementation included form offering rate, acceptance/completion rate by patients, and documentation in the chart. Twenty semistructured interviews with patients and clinicians were conducted to assess intervention experience.

Results: A total of 2748 patient encounters were evaluated. OKAD was offered in 33% of eligible patient visits (493/1494) and accepted 54% of the time (266/493). Five Wishes was offered in 36% of eligible patient visits (450/1254) and accepted 82% of the time (369/450). Unadjusted analyses found no significant difference in offering of advance directive forms between groups. However, the odds of accepting Five Wishes were 3.89 times that of OKAD (95% CI, 2.88 to 5.24; P < .0001). Logistic regression models controlling for several confounders indicated that the acceptance of Five Wishes was favored significantly over OKAD (OR = 1.52; 95% CI, 1.27 to 1.81; P < .0001). Qualitative analyses indicated a clear clinician and patient preference for Five Wishes.

Conclusions: Results suggest that Five Wishes was more readable, understandable, appealing, and usable. It seemed to capture patient preferences for end-of-life care more effectively and it more readily facilitated patient-clinician conversations.