What makes a doctor come to a conclusion about a diagnosis and a treatment? What determines the need for a surgery or whether some other modality of treatment is an option?
Is surgical decision making binary?
The internet is a scary quagmire. It is wondrous, often hilarious, and most recently has become the bane of medical practice. Not because individuals should not know about their illnesses – of course they should, it may help them make wise choices regarding their lifestyle that could benefit them. But because there is such a thing as too much information. And we’ll get to that.
Medicine is a field of study that is dependent on acquiring a sizeable body of knowledge before it becomes an art. Like any musician, it’s years of study and practice to make performance seem effortless, graceful and almost magical. Acquiring said sizeable body takes years and at least as of 25 years ago the only way to do it was to pore over volumes of texts.
I lost my will to read almost anything after I finished my neurosurgical training largely because just looking at print would generate the mildest form of PTSD. A discomfort at the least and frank aversion if it meant looking through another textbook. During exams, we spoke of how much we had left to study or revise in inches of books and not pages because that just seemed most apt.
That’s one way to gain knowledge.
As time passed along my medical career I discovered more ways. Corridor knowledge was an easily available drug. It lurked skulking at corners of the hospital, peddled by fellow residents and like any cheap, illicit substance, often had the worst side effects. The main one being it was often wrong, and there was no respectable citation to back it up. Moreover this corridor knowledge wasn’t created in research facilities and universities, it was cooked up in the exhausted, sleep deprived minds of surgical trainees. Often mixing up ingredients from different recipes.
“Hey what’s the incidence of tumor?”
“Oh that’s about 17%”
“17? Seems high, where did you read that?”
Silence. “I don’t remember where I read it but I’m sure it’s 17%.”
It wasn’t. I found out when I was being quizzed by a professor that the number was maybe closer to 8. Glaring at the erroneous source didn’t help, he was alseep in the back of the room.
The other source of knowledge that became more prevalent as I progressed along the way, was the often-sarcastic, “In my vast experience”. The exact quantity and quality of the experience is often unknown but the confidence with which this phrase has been used has been no less than authoritative.
But, jokes apart, textbooks are the first source of knowledge. They collect and distill what is considered facts and present them as is. They acknowledge that some information may still be evolving as studies continue to be done, and they accept that there’s enough we know absolutely nothing about. Textbooks are great for learning established facts that haven’t changed since at least 3 or 4 editions, to be safe. Anatomy, for instance, a large part of what is considered normal anatomy has been a constant over time. Surgical technique has a fairly immovable foundation. But there are topics that evolve rapidly, where nothing is written in stone and our incredibly incomplete knowledge changes ever so often. Neuroscience is a great example. Research into how and why neurons work or don’t work, chemical and electrical signalling, the pathology and genetics of tumors and other diseases, neurodegeneration, neuromodulation, brain-machine interfacing, neuronal plasticity, and many more, it frankly put the tip of a 1.5kg mass of gooey tissue.
How then are we to decide what test would clinch a diagnosis and what treatment would cure? In about maybe 70% of the time the foundation of knowledge from all the studying comes to our, and our patient’s rescue.
In 30% we depend on the latest research. And Statistics.
There are lies, damn lies and there are statistics.Mark Twain
The highest court of appeal in medical research was a prospective double blinded randomised control study with crossovers. A statistically significant result from any such study was once considered gospel.
And then someone took all randomised studies in a particular area and meta-analysed the results. And destroyed the reputation of Prozac. Showing that while individual studies showed the landmark antidepressant was the best thing since sliced bread, an in-depth look at all studies showed that while Prozac worked it wasn’t much better than placebo. Or at least significantly better.
There in lies one of the many issues with evidence based medicine. Statistics are a tool. And a powerful tool and with great power…
If you won’t give a 10th grader a craniotome drill to make a hole in the skull, you certainly shouldn’t allow just anyone to download SPSS. It has been said.
It is possible to, within certain limits, to push the dough of data in to a cookie cutter of any shape and throw away the discards. Often times, governments, government agencies, industry has been accused of using studies to push an agenda.
That scary fact aside, there are many situations that a randomised controlled study is just not possible, feasible or just illogical.
The British Medical Journal often publishes articles that aren’t cutting edge research but do prove some point. In 2003 a paper was published that said there was no evidence that using a parachute while jumping off a plane would save ones life. Simply because there was not enough data to prove that NOT using one was likely fatal. Barring common sense.
In 2018 another study managed to randomise 2 populations, one with and one without parachutes and proved that after jumping out of a plane there was no difference in the outcome of the 2 groups. The plane was a small single propellor. And stationary on a runway.
The statistical analysis was flawless, except for a small number of study subjects.
So ‘evidence’ cannot be followed blindly.
It goes back then to what we think will work, “in our vast experience” and that’s honestly where experience makes a difference. The quantum of disease that is seen in many years of a medical or surgical career sometimes can outweigh statistics. A fairly judicious use of knowledge gained in training, and every single day and from every single patient and surgery, can make things work almost a 100% of time. What remains is what Rene Leriche said.
Rene Leriche (1879-1955)
Every surgeon carries within himself a small cemetery, where from time to time he goes to pray – a place of bitterness and regret, where he must look for an explanation for his failures.
Before we proceed lets lighten our spirits with this gem of a paper that looks at all head injuries from the Asterix and Obelix series. Hit me up if you can’t get the full text.
Following that intermission, can we speak of my dear friend Google. We learn early that hoof prints don’t mean zebras. That outliers exist in every situation and nothing is binary. And that there’s a lot we don’t know about the human body and a lot less about the human brain.
I see 3 main issues with the available information on the internet. It’s unregulated and it’s disorganised, but those aren’t the issues, they are the problem in itself. Googling symptoms can pull out a laundry list of diseases that are often irrelevant. We used to joke about medical student syndrome where with each passing week we imagined we had been suffering from a new disease, as we extrapolated symptoms, imagined events and convinced ourselves to exhibit signs of obscure illnesses. The result of unregulated information.
The other is in treatment. I’ve had patients google their medications and not take them for various reasons. Fear of every side effect listed, deciding that that wasn’t the correct medication. I can argue that there’s a list of dos and don’ts and warnings and disclaimers for every thing we buy and use but that still doesn’t stop someone from deciding that the simultaneous usage of a cellular device and an internal combustion engine is the way to demonstrate the supremacy of human endeavour.
The third is the inability to often distinguish between anecdotal instances and what happens in the majority of situations. Yes, cancers can sometimes show spontaneous remission. These instances are rare. Rare enough that they get written up as case reports in scientific literature. That doesn’t mean every one with that disease should hope for spontaneous remissions and not take chemo. We’ve lost patients to that delay. The occasional report of a phone spontaneously combusting doesn’t stop people from buying the phone if 99% of the time it works well.
This post started as a way to explain how we, as doctors think in a sequential process with history and examination and investigations and come to a conclusion as to the diagnosis. And then to look at the patient, their expectations, the reality of the situation, in as holistic a way as possible to advise a treatment. Sometimes it’s surgery and sometimes it’s medication. Often it’s reassurance that things will get better or that nothing is seriously amiss. Our route to coming to these conclusions is long and arduous and filled with ups and downs. The internet, in it’s infinite wisdom throws down everything from alternate systems of medicine, to natural healing, to.. (I’ll plead the 5th and shut up). It’s bad enough we have to avoid everyone who’s falling out of planes without a parachute.