Eliciting preferences for alternative cancer drug treatments. The influence of framing, medium, and rater variables.

In oncology there is increasing interest in the development of techniques to help patients choose between alternative therapies in ways that are consistent with their preferences. The purpose of this study was to examine some methodological problems associated with the elicitation of preferences. Preferences for alternative drug therapies were sought from 208 visitors to an open house at the Ontario Cancer Institute and from 216 university nursing students. Preferences were not significantly dependent on the sex, age or professional status of the respondents, nor on the medium used for elicitation of preferences (computer terminal versus pencil-and-paper questionnaire). In contrast, the importance of the way decision problems are framed was confirmed. Comparison of a positive frame (outcomes expressed as the probability of surviving) with a negative frame (outcomes expressed as the probability of dying) and a mixed frame (probabilities of surviving and dying were both given) pointed to the presence or absence of the word "survive" in the outcome description as the main source of framing bias. The framing effect was in the opposite direction to that hypothesized.