Purpose: In past research that presented treatments for colon cancer, most people rated death as worse than any possible complication after surgery (specifically: colostomy, chronic diarrhea, intermittent bowel obstruction, wound infection); however, people then go against these ratings by choosing a treatment that puts them at greater risk of death in order to avoid the chance of complications associated with other treatment options. We tested several interventions designed to counteract this inconsistency by helping people to recognize that even if they got a complication, they would still be alive.

Methods: We recruited subjects from a demographically balanced panel to participate in an online survey experiment. Participants were asked to imagine they have colon cancer and that there are two surgeries available to treat their illness. Surgery 1 has a lower chance of death with some chance of complications whereas Surgery 2 has a higher chance of death with no chance of complications. In addition to a no-intervention version, we tested two versions designed to focus people's attention on the fact that even if they got complications, they would still be alive. One version included a pictograph which used color and graphics to represent the various outcomes of treatment. The other version asked subjects to rate how good each possible outcome would be. We tested whether these interventions would reduce the proportion of individuals choosing Surgery 2.

Results: 1264 people completed the online survey. In the no-intervention version, 65.6% chose the surgery with the higher chance of death. Of the two interventions, only the pictographs had a significant effect, reducing the number who chose Surgery 2 to 55.3% (p=0.002). Nonetheless, the majority of people in this version still chose the surgery with the higher chance of death instead of the surgery that could potentially cause complications.

Conclusions: Neither of our interventions came close to eliminating the inconsistency, suggesting this inconsistency is extremely resilient. Our interventions were both based on a common theme: helping individuals to be consistent with their value judgments (i.e. belief that complications are better than death) and their treatment choice. Because, at the very best, pictographs reduced the inconsistency by 10%, we have only limited evidence that this theme underlies the inconsistency.