Purpose: Previous research has shown the importance of considering both the format and type of information presented to patients in decision aids (Fagerlin et al., 2005; Ubel et al., 2001). This research was designed to evaluate the impact of testimonials in a patient decision aid for early stage breast cancer.
Method: 200 women were paid $100 to participate in a randomized controlled trial. Participants were asked to imagine they had been diagnosed with early stage breast cancer and must choose between two treatments: lumpectomy with radiation and mastectomy. The participants were randomly assigned to view one of two DVDs that contained facts about the treatments and their relative effectiveness. One of the DVDs included testimonials from women who had made this treatment choice (statistics + testimonials condition); the other contained no personal testimonials (statistics only condition). After watching the DVD, participants made a treatment choice and answered additional questions about their decision making process. Participants also provided demographic data and completed the Subjective Numeracy Scale (Fagerlin et al., 2007), the Decision Making Styles Inventory (Nygren, 2000), the Test of Functional Health Literacy-short form (Parker et al., 1995), and the e-Health Literacy Scale (Norman & Skinner, 2006).
Result: Preliminary results from 84 participants showed marginally significant group differences in treatment choice. Participants in the statistics only condition were more likely to choose mastectomy than lumpectomy; however, participants in the statistics + testimonials condition were equally likely to choose either treatment. Additionally, participant education, numeracy, and health literacy interacted with condition to affect treatment choice. In the statistics + testimonials condition, higher levels of education and health literacy were associated with a preference for lumpectomy, while higher levels of numeracy were associated with a preference for mastectomy.
Conclusion: The inclusion of testimonials in a patient decision aid for breast cancer altered treatment choices made by participants. In addition, the presence of testimonials evoked different treatment choices in participants with less education and low levels of numeracy and health literacy. This research calls into question the current practice of creating a single decision aid for all patients. Furthermore, patient characteristics, such as education, numeracy and literacy, should be considered when constructing and using patient decision aids. Additional implications for the development of patient decision aids will be discussed.
Candidate for the Lee B. Lusted Student Prize Competition