Bias control in shared decision making: still too many loose endsBMJ 2012; 345 doi: http://dx.doi.org/10.1136/bmj.e8291 (Published 20 December 2012) Cite this as: BMJ 2012;345:e8291
- Aquiles R Henriquez, PhD student in medical sciences1
- 1Department of Primary and Interdisciplinary Care, University of Antwerp, Universiteitsplein 1, Wilrijk 2610, Belgium
The relevance of shared decision making to quality of care is not under debate,1 but it is unclear how we can minimise the noise and biases that influence the complexity of the task while respecting patient preferences and values.2
In contexts like primary care the probability of facing undifferentiated problems is higher. The evaluation of a patient’s preferences is challenging because the diagnostic landscape is broader and multiple trade-offs between benefits and risks need to be considered. The biases caused by how the information is framed are always at play,3 and uncertainty, error, and regret are inherent in every decision made.4 What is an acceptable burden of regret when the outcomes of our decisions do not match our expectations? To avoid an unnecessary increase in our decision thresholds, doctors and patients must learn how to deal with uncertainty and regret.
How can we discern “preference misdiagnosis,” hindsight bias, attribution error, and regret on the basis of the decision outcome? Elwyn and colleagues said “decisions cannot be measured by reference to their outcomes” and proposed that we should emphasise “the deliberation process rather than the decision’s end results.”5
Instead of a straightforward step-by-step process, shared decision making must be seen as a constant comparison process. A patient’s ideas, concerns, and expectations should be constantly weighted and compared against the professional’s own ideas, concerns, and expectations within the boundaries imposed by the published evidence, the context, and our biased nature.
Cite this as: BMJ 2012;345:e8291
Competing interests: None declared.