« Questions About DPC?... | Main | Fewer, Not More, Vac... »

Tuesday Oct 06, 2015

Cloudy With a Chance of Misdiagnosis

"As an adolescent, I aspired to lasting fame, I craved factual certainty and I thirsted for a meaningful vision of human life -- so I became a scientist. This is like becoming an archbishop so you can meet girls."

-- Matt Cartmill, Ph.D.
"Seventy-five Reasons to Become a Scientist(www.jstor.org)
American Scientist
(Sep/Oct 1988)

"Predictions show a steady low.
You're feeling just the same.
But seasons come and seasons go.
I'll make you smile again.
If you don't believe me,
Take me by the hand.
Can't you feel you're warming up?
Yeah, I'm your weatherman."
-- Delbert McClinton, "Weatherman"

As the sum of our knowledge grows, it becomes more apparent how much more there is to learn. What we think of as fact can be easily and rapidly overturned by new observations and understanding. During the past month, I've heard more than one person discuss the "old days" of medicine, often lamenting that their doctor now gives them choices in their care.

"I miss when they would just tell me what was wrong and give me a medicine to fix it," I heard one person say. My first instinct was to defend the rights of the patient and to expound on the greatness of patient autonomy. I queued up long discussions of the antiquated paternalistic model of medicine and why the informative and deliberative models are far superior. My internal anarchist railed against blindly following those who have confidence and who assume they know what is best.

But then I stopped and looked at the situation from the perspective of these patients. What would someone -- raised among the miracles of technology, without any training in pathology or diagnostics -- see from the outside in the paternalistic model? The doctor listens, examines, and then pronounces with absolute certainty a diagnosis and treatment plan. It’s the magical black box. Information goes in and apparent cure comes out. As physicians, we know the truth. The reality is far more mundane and far less certain.

This fundamental misunderstanding of the actual amount of knowledge we as physicians possess and how we apply it leads to a mismatch in expectations and reality. Most patients don’t understand that a large proportion of what we do is based on statistical probability. The informative and deliberative models of medicine we were taught in medical school encourage us to share the thought processes and basic science behind our differential, but the process underlying our diagnostics is fundamentally no different from the paternalistic model. We observe, collect data, calculate probability and choose the most likely etiology.

It strikes me, as I contemplate this process, how closely we resemble another group of scientists -- meteorologists. We deal with complex systems. Both climate and human pathology involve hundreds, if not thousands, of variables. We collect data and select the most pertinent facts for inclusion in our models. We move from the general to the specific and make predictions based on collected data. We choose the models with the highest probability and share those with our constituents. As time passes, we collect more data, which often improves the probability analysis. We compile a differential diagnosis or a forecast. And, unfortunately, hindsight often makes fools of us all.

The more pertinent information we collect, the better our predictive models will be. Whether a function of more time or better tests, higher volumes of data improve predictions. To that end, both medicine and meteorology now outsource data collection for stronger models. Not only do we use measurements from inside the clinic, but patient-collected data from health trackers and fitness tools expand the pool of available data. We also use stronger computing tools to crunch the higher volumes of numbers. Advanced computing solutions such as IBM’s Watson can assess facts and context and ultimately output predictive models on par with those of many physicians and meteorologists. And with the growth of precision medicine and molecular genetics, we now have more specific data about the molecular underpinnings of our biology.

Although there are many facets of medicine -- and meteorology -- that carry high levels of certainty, few outcomes approach 100 percent probability. There is always a chance we are wrong. We describe most diagnostics in terms of sensitivity and specificity, or what’s the probability that a positive test result is truly positive or that a negative test result is truly negative. If there are too many false-positives or false-negatives, the test doesn’t help us with accurate prediction.

Which brings us back to the initial discussion of patient perception. Under the paternalistic model of medicine, the process was the same, but the internal mathematics remained hidden from the patient. This gave the mistaken appearance that progression from simple discussion to diagnosis followed an absolute and direct causal pathway. Tell the doctor the symptoms, and the doctor tells you what caused them. No mention of process. No peeks behind the curtain. From the lay perspective, it looked like magic. Multiple generations of patients grew up thinking many diseases and disorders were easily and rapidly cured with the magic prescription pad. The trouble with that belief isn't just that it fed skewed understanding of causality and correlation, but more that if the disease wasn't immediately cured or the problem resolved completely, the patient assumed the treatment was inadequate or incorrect and the doctor incompetent, even if both treatment and physician were medically correct.

Much like incorrect forecasts have left us skeptical of the abilities of meteorologists, incorrect or incomplete diagnoses do little to endear us to our patients. If we wish to combat this phenomenon, further explanation and discussion of the probabilities and process involved in differential diagnosis will go a long way toward improving patient compliance.

Perhaps the focus should not be on misdiagnosis, but rather the process of diagnosis as a whole and how we can improve it over time.

Gerry Tolbert, M.D., is a board-certified family physician who practices in northern Kentucky. A lifelong technophile, his interests include the intersection of medicine and technology. You can follow him on Twitter @DrTolbert(twitter.com).

Posted at 10:44AM Oct 06, 2015 by Gerry Tolbert, M.D.

« Questions About DPC?... | Main | Fewer, Not More, Vac... »


Subscribe to receive e-mail notifications when the blog is updated.



Leader Voices Blog - A Forum for AAFP Leaders and Members

FPs on the Front Lines


The opinions and views expressed here are those of the authors and do not necessarily represent or reflect the opinions and views of the American Academy of Family Physicians. This blog is not intended to provide medical, financial, or legal advice. All comments are moderated and will be removed if they violate our Terms of Use.