I think much of generalist (and to a lesser extent sub-specialist) medicine hinges on management of uncertainty as experienced by physicians and patients, which has driven me to look into this fairly deeply. My first tweetorial. 1/18
Think of the patients or clinical situations that have most made you
-uncomfortable or uneasy
-intelletually stimulated
-grow as a physicain
For most physicians, these situations all involve uncertainty
2/18
-uncomfortable or uneasy
-intelletually stimulated
-grow as a physicain
For most physicians, these situations all involve uncertainty
2/18
Tolerance and management of uncertainty drives much of diagnostic reasoning behaviours and susceptibility to cognitive biases and traps, test ordering, and patients’ confidence in their providers
3/18
3/18
We tend to lionize and praise completeness and thoroughness in medical education- less so the tests appropriately not ordered. @ADetsky and @AmolAVerma have described this phenomenon. This has implications for Choosing Wisely. 4/18
pubmed.ncbi.nlm.nih.gov
pubmed.ncbi.nlm.nih.gov
Hospitals, often driven by physicains, give an image to the public that we will always find the answer. Foundations and the public like these slogans and sentiments. Unfortunately they often don't reflect reality 5/18
We emphasize Occam’s razor- the most parsimonious explanation. This is completely reasonable, but if applied too rigidly, makes us prone to early closure, framing bias, confirmation bias, motivated reasoning, and diagnostic momentum, all well known sources of error 6/18
There is also an asymmetry between feedback from tests ordered and not ordered. When you decide to pursue a given diagnostic path, you get feedback (the results), which helps inform next time that scenario is encountered 7/18
Harder to know if you made the correct decision by not pursuing a diagnostic path, particularly for those providing episodic care. I think this makes it harder to "learn" tolerance of uncertainty 8/18
Obtaining more data does not necessarily equate to having more certainty, though if often does.
-Incidental findigns on imaging studies
-nonspecific results that lead to more confusion and risk over-valuing information (e.g. a moderately high ESR or CRP). 9/18
-Incidental findigns on imaging studies
-nonspecific results that lead to more confusion and risk over-valuing information (e.g. a moderately high ESR or CRP). 9/18
I’ve always liked this pragmatic approach JP Karriser described in NEJM way back in 1989
“Our task is not to attain certainty, but rather to reduce the level of diagnostic uncertainty enough to make optimal therapeutic decisions”
NEJM 1989; 320:1489-1491
10/18
“Our task is not to attain certainty, but rather to reduce the level of diagnostic uncertainty enough to make optimal therapeutic decisions”
NEJM 1989; 320:1489-1491
10/18
Uncertainty may therefore be considered as a threshold sufficient to permit action. I’ve illustrated this here in the form of a matrix, with diagnostic certainty plotted against risk of therapy. Of course, the risk of not treating the condition in question also matters 11/18
What about patients' experience of uncertainty?
Many of us have seen that disclosure of even a devastating diagnosis, can be met with a sense of relief. Perhaps a label, and the associated action plan, is psychologically more tolerable than the black hole of uncertainty 12/18
Many of us have seen that disclosure of even a devastating diagnosis, can be met with a sense of relief. Perhaps a label, and the associated action plan, is psychologically more tolerable than the black hole of uncertainty 12/18
This was examined in a study of patients receiving a diagnosis of melanoma. Discussing life expectancy, severity of illness, and using the word “cancer” were associated with less anxiety and depression 13/18
sciencedirect.com
sciencedirect.com
The way we express uncertainty to patients and families also matters.
A scenario-based study assessed perceived competence and confidence in physician treating a hypothetical child with abdominal pain NYD 14/18
A scenario-based study assessed perceived competence and confidence in physician treating a hypothetical child with abdominal pain NYD 14/18
Explicit declaration of uncertainty was associated with lower perception of competence and lower degree of trust. Expressing uncertainty as a differential dx, or as a probability may be preferable to patients. I think this makes intuitive sense 15/18
pubmed.ncbi.nlm.nih.gov
pubmed.ncbi.nlm.nih.gov
Conclusion- Embrace uncertainty. This is the paradox- uncertainty makes us uncomfortable, yet is where the most opportunity for growth lies 16/18
Optimal state is proper balance between the 2 extremes; obtain enough certainty to initiate necessary therapy, but not so much as to over-investigate or be excessively rigid in approach and more prone to cognitive biases/errors. 17/18
I hope this was as useful to read as it was for me to research and prepare 18/18
Loading suggestions...