Patient Narratives and Medical Decision Making: The Role of Transportation in Treatment Selection Bias

Abstract

Patient decision aids (PDAs) are tools that help patients select a treatment option for a medical condition. PDAs often include narratives about other patients’ experiences, but this inclusion may lead to biased decisions due to psychological transportation into the narrative world. Transportation has been shown to facilitate attitude changes, but it has not been studied in a medical decision context. Patients may be less susceptible to transportation if they are focused on extracting answers to specific questions that they have about the treatment options.

Winterbottom’s (2012) dialysis selection protocol was adapted to determine whether transportation could underlie decision bias in four experiments. Participants were asked to choose between two forms of dialysis to treat a hypothetical kidney failure. Information segments about others’ treatment experiences were presented in a narrative format or a non-narrative question and answer (Q&A) format, and transportation was assessed after the presentation of each information segment. In the first experiment, information about the dialysis options was presented in different formats (narrative versus Q&A) whereas the second experiment presented the two options consecutively in the same format. Overall, the results confirm that transportation can occur in a medical decision context. However, both formats were equally transporting when presented first, and participants were not biased toward the option presented in the narrative. When presented second, both formats were less transporting if preceded by a narrative, suggesting that transportation into one narrative world may hinder immediate transportation elsewhere. The last two experiments explored whether transportation can account for the biasing effect of patient narratives over doctor narratives reported in Winterbottom’s original study. A dialysis option was more likely to be selected if it was presented in a patient narrative than a doctor narrative and participants favoured the option described in the more transporting narrative. This effect of narrator identity was completely accounted for by a mediated path through transportation, however, there may be other mediators and/or moderators along this indirect path. Given this incomplete understanding of narrative bias and the aforementioned order effects, decision aid developers are advised to exercise caution when considering the inclusion of narratives in PDAs.