Objective: To contribute to the debate about whether numeric statements of risk ought to be included in risk communications.
Design: Subjects (n = 115) completed a questionnaire involving a physician risk communication and 4 scenarios, each of which described a patient with symptoms and signs potentially suggestive of cancer. Each scenario was presented in 3 risk communication versions (a verbal version and 2 numeric versions) in a within-subject 4 x 3 design. Subjects rated their trust in and comfort with the information and their belief that the physician distorted their risk level.
Results: Subjects were significantly more trusting of (t =4.0, P < 0.001) and comfortable with (t =3.4, P = 0.001) the risk information, less likely to believe that the physician minimized the risk in the numeric versions than verbal versions (t =4.3, P < 0.0001), and just as likely to believe that the physician exaggerated the risk in the 2 versions (P = 0.588).
Conclusions: Including a numeric statement of risk in a risk communication can increase trust and belief in and comfort with the risk information.