CHALLENGES OF COMMUNICATING PROGNOSTIC ESTIMATES FOR DECISIONS ABOUT LIFE SUPPORT – THE EVOLUTION OF A DECISION AID TO FACILITATE SHARED DECISION MAKING ABOUT MECHANICAL VENTILATION

Purpose: We report the results of usability testing conducted as part of the iterative development of InformedTogether - a decision aid designed to facilitate shared decision making between clinicians and their patients with Chronic Obstructive Lung Disease (COPD) about invasive mechanical ventilation. 

Method(s): We conducted 4 usability testing sessions using data triangulation, including ‘think-out-loud’ task completion, followed by individual interviews, followed by focus groups to elicit further feedback.  Audio-recordings were transcribed for analysis and mixed methods were used to analyze the data. 

Result(s): We enrolled 7 clinicians, 11 COPD patients and 5 surrogate decision makers for patients with COPD. Overall, a majority of the patient and surrogate participants were able to correctly answer questions assessing gist understanding, as opposed to verbatim understanding. As expected, in the patient and surrogate groups, both low numeracy and low education levels were associated with difficulty understanding the decision aid, particularly the icon arrays and terminology. Most participants stated that they would be very likely to extremely likely to recommend that actual patients use the decision aid with their doctors. However, we saw a great deal of variety in people’s emotional reactions to the information being presented, with some stating there were elements of the decision aid which made them uncomfortable (“a slap in the face”), while others stated they liked that the information was realistic (“tell it like it is”).  We also noticed that although the decision aid was designed to achieve shared mind between clinician and patient via the exchange of biomedical information and values/preferences for outcomes, an additional layer of information we had not considered seemed to influence participants’ understanding/uptake of the data presented in the decision aid. We termed this additional layer ‘alternative knowledge’ which included personal experiences with life support shaping whether and how they were able to understand the data.

Conclusion(s): clinicians need to be aware of the influence of alternative knowledge on patients’ understanding and uptake of prognostic data. Decision aids should include tools to elicit not only values and preferences for outcomes but also patients’ experiences, pre-conceived notions, fears and beliefs which shape this alternative knowledge and which may be critical to understanding data and informed decision making.