AUTOMATIC HEURISTIC CUES OR SLOW AND DELIBERATE REASONING: USING COGNITIVE-EXPERIENTIAL SELF-THEORY TO EXPLAIN VARIABILITY IN UNDERSTANDING DECISION AID PROGNOSTIC ESTIMATES

Purpose:

Cognitive-Experiential Self-Theory (CEST) proposes two processing systems: intuitive-experiential (IE) and analytical-rational (AR). The IE operates preconsciously, relying on heuristic cues and emotions. The AR operates consciously, is slow, deliberate, and relies on logical reasoning. In this study we used CEST as a conceptual framework to understand the diversity of responses about a web-based decision aid (DA) for choosing mechanical ventilation in severe obstructive pulmonary disease (COPD).

Methods:

Participants were asked to interact with a DA that compared estimated outcomes for severe COPD patients choosing whether or not to receive invasive mechanical ventilation (Full Code versus Do Not Intubate). Subsequently, subjects were individually interviewed and participated in a focus group. Twenty-seven semi-structured interviews and 4 focus groups were conducted: 7 physicians, 11 COPD patients, 4 surrogates and 5 mixed. Discussion involved experience with DA and decision-making for COPD patients. Triangulation (methods and analyst) was conducted to facilitate deeper understanding of the interview and focus group data. Transcriptions were coded based on the IE and AR systems, for example, if a respondent indicated that their decision-making process involved personal experience, emotion or gut feeling, the text was coded as ‘IE’. Transcriptions were analyzed using thematic analysis using NVIVO 10.

Results:

Participants whose comments reflected the AR system when describing their experience with the DA, better understood the contents of the tool, scored higher on knowledge questions about estimated outcomes within the DA, and had more specific feedback and suggestions for DA improvement.  Conversely, participants whose comments reflected the IE system scored lower on these knowledge questions and relied on their personal experience (witnessing a family member/friend with COPD at the end of their life) as well as images of an intubated patient depicted within the DA, when describing their decision-making process.

Conclusion:

CEST is useful for gaining insights into how individuals process information about prognostic estimates presented within a DA. Participants using the AR processing system better understood DA content as compared to participants using the IE system. However, the direction of this relationship cannot be assessed in our study. For example, it is not clear whether those with lower understanding relied instead on their IE system, or if their heuristic cues and personal experience led them to bypass engagement with and understanding of the content.