Thanks for this thoughtful post. My professional communications focus on autoimmune diseases made me find your choice of examples very ironic, since it is far more common for autoimmune to be missed rather than overdiagnosed. It still takes on average 3.5 years & 3-5 physicians for patients to get an autoimmune diagnosis. Far too many patients are sent to mental health care (aka gaslighting, likely because depression/anxiety is comorbid with AIIDs) rather than rheumatology or other appropriate specialists.
What kind of confirmation bias is in play here? Vague symptoms that could suggest any of a large number of diseases? Ignorance about autoimmune? Reliance on lab tests that are not very sensitive or specific?
I frequently see the misdiagnosis of lupus in two scenarios: 1) the false positive ANA in the absence of symptoms of SLE, and 2) rosacea, which is mistaken for the "butterfly" rash of SLE. Both of these are quite common.
Closely related is the urge to narrow a diagnosis prematurely...a habit often seen in new practitioners...the assessment should strive to encapsulate just the right amount of uncertainty so as to keep future thinking open to new possibilities
The physicians cited in the attribution bias article I am certain are exceptional, diagnosticians, as well as generalists. Unfortunately, the same cannot be said of the younger generalists trained in the last 15 years.
Critical thinking is a lost Art. Technocrats are the norm. Abdominal pain now equals CT scan. In the past good training, taught that a good history would generate a mental differential diagnosis. Followed by a good physical exam the differential would be further narrowed. Then,
the one or two tests, with the highest probability of confirming (with the least harm), the diagnosis would be ordered. The majority of generalist today are punters not quarterbacks.
In my humble opinion, virtual office visits are just as good as in person visits because more often than not doctors don’t even examine the patients when given the opportunity.
As Einstein said, in reference to the general population: “when technology overtakes human interaction, we will have a generation of idiots“ Medicine, is no exception😢 B Hourani
Sure, cognitive errors are common and we're all vulnerable to making them, but the core error in this vignette was the lack of a competent physical exam by the referring physician. The a fib could and should have been detected initially in a patient with fatigue and dyspnea had the physician performed a competent physical examination, palpating the pulse and auscultating the heart, applying the stethoscope to the chest wall rather than through layers of clothing and then listening. This isn't complicated, it's fundamental and sadly uncommon these days. With a confirming ECG, the diagnosis would be made, the patient started on anticoagulation and rate control if needed and referred to cardiology, all in one visit.
With the explosion in the number of CT scans and other advanced imaging modalities so-called incidentalomas have become a huge problem. However, even back in the days of plain film radiology (yes, on actual film!) there were radiologists who overcalled every tiny shadow on a film making mountains out of molehills.
When I was a radiology resident my mentor, the late Dr. Harry Mellins, taught us that there were some “ditzels” that could be ignored and observed that, “The reason radiologists are paid well is to keep some of our doubts to ourselves.”
Good book to read ""How We Do Harm: A Doctor Breaks Ranks About Being Sick in America" is a non-fictional account written by Otis Webb Brawley, the chief medical officer and vice president of the American Cancer Society."
Here is a tool to use to see if you REALLY need screening.
Last but not LEAST.. why do Oncologist get to sell the therapy that they RX? No one has been able to give me an answer. Seems like a conflict of interest, no?
Google is the WORST place for attribution bias and I have family members and friends who still search “Dr. Google” for the cause of a problem. And of course I have been guilty of an occasional dip many years ago...the more I know the more I don’t know, and in my own case it’s best to leave the hard work to my PCP and specialists!
Wrong. The more you know (and you know YOU best) the MORE you can ADVOCATE for YOU. Doctors miss a lot of things (to be fair patients often don't tell them important things either) - 15 minutes in and out and here's the damn pill. Like BP and Statins. Oh, take that Statin and that BP...for example please do. Statins are useless or next to and come with some pretty bad side effects. Want to reduce hardening of arteries? Test for Insulin Resistance, and if that's positive put them on a Keto reduced sugar/carb diet, check the autoimmune diseases and the thyroid/parathyroid - look for the root causes, get them on the natural BP reducers from foods (Beets, tart cherries, hibiscus tea. Prune out the bad fats, incorporate the good ones. Check for sleep apneas and the like. Incorporate those changes into the diet. Exercise, get them to walk 5 times a week 30 minutes a session. Stop smoking, reduce alcohol. .. Do that and follow up in 6 months. See the difference in those folks if they take the prescription for diet, exercise ...balancing of their system's needs....But that takes work on the doctor's part. And effort on the patient's part. And we are a very LAZY nation that doesn't want to make the changes necessary for good health. And doctors who only believe in pills...
I said that tongue in cheek - I truly am my best advocate! I remember my medical director (I was a program director for addiction and pain) telling a patient once he could be off of at least 5 drugs if he changed his life habits - some of which you describe above. The patient said, “Nah, I’ll just keep eating the way I want and I like smoking so u just keep giving me those pills.” CRAZY true about Americans.
This is a great point. Unfortunately there is no consensus definition of diagnostic excellence and no way to measure it. To answer your question, think diagnostic humility is the place to start. You’re pointing to Dunning Kruger effect and the solution to lack of insight is always highlighting the scope of the problem first. I haven’t gotten a chance to read the article you sent but look forward to it for this reason.
As individual diagnosticians, we would all benefit from a personal dashboard of missed diagnoses - it would be VERY humbling I would imagine. I think AI could help track it for us. If taking a 1% better approach, my guess is that improvement comes at the cost of deep, individual reflection as to what could’ve been done differently. I say ‘deep’ because in many cases you could learn the ‘wrong lesson’ (eg head CT for every headache). Again, no quick, easy answers here.
Certainly diagnostic humility is a dose needed by all though.
I met the great blues musician, Duke Tumatoe, (https://en.m.wikipedia.org/wiki/Duke_Tumatoe) in the 80s when he played a show at Notre Dame. I helped load and unload his equipment. As we were hanging together after everything was packed, a friend told him I played and asked what advice he'd give me. Duke offered the following: "Try not to suck!" We had a good laugh.
Decades later the 2016 Chicago Cubs adopted the same rallying cry: “Try Not to Suck” on their way to a long awaited triumph. (I watched my first game at Wrigley in 1967. About time we won a WS that eluded my grandfather who saw them lose in 35 and 45. Although, I often regret the tear in time-space that it created leading to 🍊🤡, pandemic, January 6, Canadian wild fire smoke, etc.)
But what are we to make of Meyer, et al. Physicians’ Diagnostic Accuracy, Confidence, and Resource Requests A Vignette Study (2013)
"Results A total of 118 physicians with broad geographical representation within the United States correctly diagnosed 55.3% of easier and 5.8% of more difficult cases (P < .001). Despite a large difference in diagnostic accuracy between easier and more difficult cases, the difference in confidence was relatively small (7.2 vs 6.4 out of 10, for easier and more difficult cases, respectively) (P < .001) and likely clinically insignificant. Overall, diagnostic calibration was worse for more difficult cases (P < .001) and characterized by overconfidence in accuracy. Higher confidence was related to decreased requests for additional diagnostic tests (P = .01); higher case difficulty was related to more requests for additional reference materials (P = .01)."
"Conclusions and Relevance Our study suggests that physicians’ level of confidence may be relatively insensitive to both diagnostic accuracy and case difficulty. This mismatch might prevent physicians from reexamining difficult cases where their diagnosis may be incorrect."
Dr. Rohlfsen suggests, not incorrectly, to avoid "Attribution Bias". But as the statistician and quality expert, W. E. Deming would often ask: "By what method?"
The pathetic success results reported by Meyer, et al (55% on easy cases?) are even more troubling given physician confidence. Maybe Meyer et al is not reproducible. Maybe physicians a decade later are even more arrogant and sucky. Maybe they are better?
Telling physicians to avoid attribution bias without a proposed method is no better than the slogan "try not to suck". (Maybe all doctors should take a placebo before making a diagnosis: we might get a 20% improvement!?)
Good advice and great slogan but no one ever dies from a poor musician and certainly not from the Cubs losing (Chicago would be uninhabited if the Cubs' suckiness were a life or death proposition.)
How shall physicians not suck so much? Maybe you are forgetting to sacrifice a rooster to Asklepios. I don't know. Saint Luke, Saints Cosmas and Damian, Saint Camillus de Lellis and Saint Gianna Beretta Molla pray for physicians and all of us.
Interesting intro to a fabulous response - Duke and I did many shows together back in the “Bob and Tom” days - “try not to suck” is a classic Duke line!
Thanks for writing this article. I am now a retired academic surgeon who was a faculty member at a major medical school; now aged 80, I am reminded often about the Huge Challenges faced by Family Practice or Internal Medicine doctors all day long every week. I have believed since medical school days that both of those specialties are much more difficult than any one of the various surgical subspecialty areas -- the breadth of knowledge required is stunning, the potential for harm is sometimes great, and the ability to think analytically is crucial. And the pay is lousy.
Dope piece!
What do you mean by “equitable” care?
Thanks for this thoughtful post. My professional communications focus on autoimmune diseases made me find your choice of examples very ironic, since it is far more common for autoimmune to be missed rather than overdiagnosed. It still takes on average 3.5 years & 3-5 physicians for patients to get an autoimmune diagnosis. Far too many patients are sent to mental health care (aka gaslighting, likely because depression/anxiety is comorbid with AIIDs) rather than rheumatology or other appropriate specialists.
What kind of confirmation bias is in play here? Vague symptoms that could suggest any of a large number of diseases? Ignorance about autoimmune? Reliance on lab tests that are not very sensitive or specific?
I frequently see the misdiagnosis of lupus in two scenarios: 1) the false positive ANA in the absence of symptoms of SLE, and 2) rosacea, which is mistaken for the "butterfly" rash of SLE. Both of these are quite common.
Closely related is the urge to narrow a diagnosis prematurely...a habit often seen in new practitioners...the assessment should strive to encapsulate just the right amount of uncertainty so as to keep future thinking open to new possibilities
The physicians cited in the attribution bias article I am certain are exceptional, diagnosticians, as well as generalists. Unfortunately, the same cannot be said of the younger generalists trained in the last 15 years.
Critical thinking is a lost Art. Technocrats are the norm. Abdominal pain now equals CT scan. In the past good training, taught that a good history would generate a mental differential diagnosis. Followed by a good physical exam the differential would be further narrowed. Then,
the one or two tests, with the highest probability of confirming (with the least harm), the diagnosis would be ordered. The majority of generalist today are punters not quarterbacks.
In my humble opinion, virtual office visits are just as good as in person visits because more often than not doctors don’t even examine the patients when given the opportunity.
As Einstein said, in reference to the general population: “when technology overtakes human interaction, we will have a generation of idiots“ Medicine, is no exception😢 B Hourani
Paging Dr. House...
Sure, cognitive errors are common and we're all vulnerable to making them, but the core error in this vignette was the lack of a competent physical exam by the referring physician. The a fib could and should have been detected initially in a patient with fatigue and dyspnea had the physician performed a competent physical examination, palpating the pulse and auscultating the heart, applying the stethoscope to the chest wall rather than through layers of clothing and then listening. This isn't complicated, it's fundamental and sadly uncommon these days. With a confirming ECG, the diagnosis would be made, the patient started on anticoagulation and rate control if needed and referred to cardiology, all in one visit.
With the explosion in the number of CT scans and other advanced imaging modalities so-called incidentalomas have become a huge problem. However, even back in the days of plain film radiology (yes, on actual film!) there were radiologists who overcalled every tiny shadow on a film making mountains out of molehills.
When I was a radiology resident my mentor, the late Dr. Harry Mellins, taught us that there were some “ditzels” that could be ignored and observed that, “The reason radiologists are paid well is to keep some of our doubts to ourselves.”
Good book to read ""How We Do Harm: A Doctor Breaks Ranks About Being Sick in America" is a non-fictional account written by Otis Webb Brawley, the chief medical officer and vice president of the American Cancer Society."
Here is a tool to use to see if you REALLY need screening.
https://www.uspreventiveservicestaskforce.org/uspstf/
Last but not LEAST.. why do Oncologist get to sell the therapy that they RX? No one has been able to give me an answer. Seems like a conflict of interest, no?
Google is the WORST place for attribution bias and I have family members and friends who still search “Dr. Google” for the cause of a problem. And of course I have been guilty of an occasional dip many years ago...the more I know the more I don’t know, and in my own case it’s best to leave the hard work to my PCP and specialists!
Wrong. The more you know (and you know YOU best) the MORE you can ADVOCATE for YOU. Doctors miss a lot of things (to be fair patients often don't tell them important things either) - 15 minutes in and out and here's the damn pill. Like BP and Statins. Oh, take that Statin and that BP...for example please do. Statins are useless or next to and come with some pretty bad side effects. Want to reduce hardening of arteries? Test for Insulin Resistance, and if that's positive put them on a Keto reduced sugar/carb diet, check the autoimmune diseases and the thyroid/parathyroid - look for the root causes, get them on the natural BP reducers from foods (Beets, tart cherries, hibiscus tea. Prune out the bad fats, incorporate the good ones. Check for sleep apneas and the like. Incorporate those changes into the diet. Exercise, get them to walk 5 times a week 30 minutes a session. Stop smoking, reduce alcohol. .. Do that and follow up in 6 months. See the difference in those folks if they take the prescription for diet, exercise ...balancing of their system's needs....But that takes work on the doctor's part. And effort on the patient's part. And we are a very LAZY nation that doesn't want to make the changes necessary for good health. And doctors who only believe in pills...
I said that tongue in cheek - I truly am my best advocate! I remember my medical director (I was a program director for addiction and pain) telling a patient once he could be off of at least 5 drugs if he changed his life habits - some of which you describe above. The patient said, “Nah, I’ll just keep eating the way I want and I like smoking so u just keep giving me those pills.” CRAZY true about Americans.
This is a great point. Unfortunately there is no consensus definition of diagnostic excellence and no way to measure it. To answer your question, think diagnostic humility is the place to start. You’re pointing to Dunning Kruger effect and the solution to lack of insight is always highlighting the scope of the problem first. I haven’t gotten a chance to read the article you sent but look forward to it for this reason.
As individual diagnosticians, we would all benefit from a personal dashboard of missed diagnoses - it would be VERY humbling I would imagine. I think AI could help track it for us. If taking a 1% better approach, my guess is that improvement comes at the cost of deep, individual reflection as to what could’ve been done differently. I say ‘deep’ because in many cases you could learn the ‘wrong lesson’ (eg head CT for every headache). Again, no quick, easy answers here.
Certainly diagnostic humility is a dose needed by all though.
The likely attribution bias of specialists is summed up in the well known aphorism, “When you’re a hammer everything is a nail.”
I met the great blues musician, Duke Tumatoe, (https://en.m.wikipedia.org/wiki/Duke_Tumatoe) in the 80s when he played a show at Notre Dame. I helped load and unload his equipment. As we were hanging together after everything was packed, a friend told him I played and asked what advice he'd give me. Duke offered the following: "Try not to suck!" We had a good laugh.
Decades later the 2016 Chicago Cubs adopted the same rallying cry: “Try Not to Suck” on their way to a long awaited triumph. (I watched my first game at Wrigley in 1967. About time we won a WS that eluded my grandfather who saw them lose in 35 and 45. Although, I often regret the tear in time-space that it created leading to 🍊🤡, pandemic, January 6, Canadian wild fire smoke, etc.)
But what are we to make of Meyer, et al. Physicians’ Diagnostic Accuracy, Confidence, and Resource Requests A Vignette Study (2013)
https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/1731967
"Results A total of 118 physicians with broad geographical representation within the United States correctly diagnosed 55.3% of easier and 5.8% of more difficult cases (P < .001). Despite a large difference in diagnostic accuracy between easier and more difficult cases, the difference in confidence was relatively small (7.2 vs 6.4 out of 10, for easier and more difficult cases, respectively) (P < .001) and likely clinically insignificant. Overall, diagnostic calibration was worse for more difficult cases (P < .001) and characterized by overconfidence in accuracy. Higher confidence was related to decreased requests for additional diagnostic tests (P = .01); higher case difficulty was related to more requests for additional reference materials (P = .01)."
"Conclusions and Relevance Our study suggests that physicians’ level of confidence may be relatively insensitive to both diagnostic accuracy and case difficulty. This mismatch might prevent physicians from reexamining difficult cases where their diagnosis may be incorrect."
Dr. Rohlfsen suggests, not incorrectly, to avoid "Attribution Bias". But as the statistician and quality expert, W. E. Deming would often ask: "By what method?"
The pathetic success results reported by Meyer, et al (55% on easy cases?) are even more troubling given physician confidence. Maybe Meyer et al is not reproducible. Maybe physicians a decade later are even more arrogant and sucky. Maybe they are better?
Telling physicians to avoid attribution bias without a proposed method is no better than the slogan "try not to suck". (Maybe all doctors should take a placebo before making a diagnosis: we might get a 20% improvement!?)
Good advice and great slogan but no one ever dies from a poor musician and certainly not from the Cubs losing (Chicago would be uninhabited if the Cubs' suckiness were a life or death proposition.)
How shall physicians not suck so much? Maybe you are forgetting to sacrifice a rooster to Asklepios. I don't know. Saint Luke, Saints Cosmas and Damian, Saint Camillus de Lellis and Saint Gianna Beretta Molla pray for physicians and all of us.
Interesting intro to a fabulous response - Duke and I did many shows together back in the “Bob and Tom” days - “try not to suck” is a classic Duke line!
Dr. Duke definitely a different kind of doctor.
There would be no reason for him to even remember college kids helping him with equipment at La Fortune Center at ND in 83 or 84.
Thanks for writing this article. I am now a retired academic surgeon who was a faculty member at a major medical school; now aged 80, I am reminded often about the Huge Challenges faced by Family Practice or Internal Medicine doctors all day long every week. I have believed since medical school days that both of those specialties are much more difficult than any one of the various surgical subspecialty areas -- the breadth of knowledge required is stunning, the potential for harm is sometimes great, and the ability to think analytically is crucial. And the pay is lousy.