METHODS: Subjects were recruited in a hospital endoscopy waiting room, where they were waiting to escort a patient home after an endoscopic procedure. Data on individual preferences were collected using a questionnaire administered using a laptop computer. Standard risk communication techniques were implemented to develop a web-based preference-threshold elicitation interface. Participants were presented summary profiles outlining the benefits and harms of celecoxib and aspirin (presented anonymously as Drug A and B), using absolute lifetime risks. Participants were then asked if they would be willing to take each drug. Depending upon the original choice, subjects responded to a series of binary, follow-up questions where the risk-benefit ratios were altered in a ‘ping-pong' fashion. The exercise was designed to elicit personal threshold values for each risk and benefit of the two drugs and to determine conditions under which patients would be willing to take them.

RESULTS: Of 229 persons asked to participate, 202 (88%) completed the survey. Under base conditions, 15 (7.35%) would take celecoxib and 89 (43.6%) would take aspirin. Of the participants, 63.2% refused celecoxib even if it eliminated all future CRC risk. Of those who would take celecoxib (36.8%), they would do so if the CRC risk threshold was reduced by 55% on average. 64.9% refused celecoxib even if it did not increase lifetime risk of heart attack (HA). The remaining (35.1%) would take it with an average HA risk increase of 31%. 30.7% refused aspirin even if it eliminated all future CRC risk. 69.3% would take it with an average CRC risk reduction of 32%. Of men surveyed, 14% refused aspirin even if it eliminated future risk of HA. However, 29% would take aspirin without a HA benefit. 57% would take it at an average decreased HA risk of 37%. 34% of women refused aspirin even if it eliminated future risk of stroke. 22% would take aspirin even if it had no stroke benefit. 44% would take it with an average decreased stroke risk of 44%.

CONCLUSIONS: Patients are six times more willing to take aspirin than celecoxib for CRC chemoprevention. Threshold values for the risks and benefits of the drugs are presented.