Fifty Cognitive and Affective biases.

Pat Croskerry croskerry at EASTLINK.CA
Sun Sep 8 11:40:51 UTC 2013

It is unreasonable to expect people to remember the list of biases and their
descriptors, but bias is such an important feature of everyday brain
functioning that all of us need to be aware, at least, of what is out there.

Wikipedia's list is steadily growing and currently stands at 95 cognitive
biases as well as lists of social biases and memory errors/biases. They also
have the beginnings of a classification under 'Common theoretical causes' Jenicek in his book
(Medical error and harm: understanding, prevention and
Control) lists over a hundred. There clearly is a need for a taxonomic
grouping. We suggested one in an earlier paper (Campbell et al. Profiles in
Patient Safety: A ''Perfect Storm'' in the Emergency Department. Acad Emerg
Med 2007; 14: 743-749). 

We are probably a long way from developing lists of prevalence. Unlike
medication errors, they do not leave any tangible trail of evidence, and
which particular bias has been involved often depends on who is looking at
it. One can imagine different prevalence profiles for different disciplines
- those in the visual specialties (radiology, dermatology, anatomic
pathology)  might look quite different from those in internal medicine.  In
several papers, depending on the discipline, people have drawn up lists of
the most common ones that they see (e.g. Crumlish and Kelly, Advances in
psychiatric treatment (2009), vol. 15, 72-79; Crowley RS, et al. Automated
detection of heuristics and biases among pathologists in a computer-based
system. Adv Health Sci Educ 2013;18:343-63.) and this is probably the
easiest way to go right now. We drew up the list of 50 for our Critical
Thinking Program at Dalhousie to make things more manageable, but give
clinical case examples that focus mostly on the common ones. We stress to
medical students that they rarely occur in isolation and usually several can
be identified in clinical cases that have adverse outcomes. Often, they are
also mixed in with various system errors. As for studies to pick bias up,
Silvia Mamede's group in Rotterdam  appears to be well on the way with this.
This appears to be a very relevant area in clinical decision making and
several medical undergraduate curricula now review the issue. They are not
the only source of cognitive failure in diagnostic reasoning, however, and
what appears to be more broadly needed is coaching in critical thinking.
Arguably, how we think is the most important thing that we do, yet this is
usually not formally acknowledged (or coached) in the undergraduate
curricula of most medical schools.  
It seems to be well recognized in the critical thinking literature that
awareness of cognitive and affective biases is an important critical
thinking skill. There is now a growing awareness of the importance of bias
in clinical reasoning and most disciplines in medicine have now recognized

Now that the problem has been recognized, the next logical step appears to
be serious research effort aimed at cognitive debiasing. It doesn't look as
though it will be easy but it does look doable. Several papers  have been
published recently that attempt to get this ball rolling  (Graber et al,
Cognitive interventions to reduce diagnostic error: a narrative review, BMJ
Q&S 2012; Croskerry, Singhal, Mamede, Cognitive debiasing 1 and 2 BMJ Q&S


-----Original Message-----
From: robert bell [mailto:rmsbell at ESEDONA.NET] 
Sent: Thursday, September 05, 2013 7:17 PM
Subject: [IMPROVEDX] Fifty Cognitive and Affective biases.

Dear all,

I have really appreciated Pat Croskerry's list of 50 biases that physicians
can have when making diagnoses. And have sent to others by e-mail to create
discussion and thought.

I wondered if this list could be made more useful?  Personally, I would
never remember the whole 50 biases on the list.

Could the descriptors more accurately describe the biases to perhaps
remember more easily; could similar issues be grouped together; perhaps
could there be an attempt to rank the biases in terms of which are more
likely to to be present in the average practice situation? 

Which brings up the issue of would the types of biases be different for PCPs
versus Specialists? And yet again how would you know which are the most
important without research? If the biases were ranked would you need
different lists for specialists and PCPs or even different lists for
different specialities?

And how would you design a study to pick the biases up? Is there bias
research in other areas that could perhaps be extrapolated the the medical
diagnostic arena?  e.g. Airline industry?

And finally, would the list be more useful if there were 5 to 10 main

Rob Bell, M.D., Ph.C.

Moderator: Lorri Zipperer Lorri at, Communication co-chair, Society
for Improving Diagnosis in Medicine

To unsubscribe from the IMPROVEDX list, click the following link:<br> <a

To unsubscribe from the IMPROVEDX list, click the following link:<br>
<a href="" target="_blank"></a>

HTML Version:
URL: <../attachments/20130908/dcee3f35/attachment.html>

More information about the Test mailing list