"Is 28% good or bad?" Evaluability and preference reversals in health care decisions.

Choices of health care providers can become inconsistent when people lack sufficient context to assess the value of available information. In a series of surveys, general population samples were randomized to read descriptions of either 2 possible health care providers or a single provider. Some information about providers was easy to consider (e.g., travel time), but some was difficult to interpret without additional context (e.g., success rates). Ratings of the described health care providers varied significantly by whether options were evaluated independently or concurrently. For example, one fertility clinic (33% success rate, 15 min away) was rated higher than a 2nd (40% success rate, 45 min away) when each clinic was considered separately (7.1 v. 6.2, P = 0.046), but preferences reversed in joint evaluation (5.9 v. 6.7, P = 0.051). The results suggest that clinicians and developers of patient information materials alike should consider information evaluability when deciding how to present health care options to patients.

Duke Scholars

Author Peter A Ubel Fuqua School of Business