Dr Harry weighs in on the dangers of diagnosing with Google
Every so often an ad pops up on my Facebook feed - selling a coffee mug with the words “Your Dr. Google is no match for my medical degree”. I always have a chuckle when I see the ad (nope, I still haven’t been tempted to buy the mug!) and my mind flashes back to the hundreds of patients I’ve seen for a second opinion….the first opinion coming from a Google search followed by trawling through an online conversation between strangers and self-proclaimed experts, supposedly with precisely the same problem. I get it, of course. It’s completely normal to turn to the internet when you want information and I confess to doing the same. Well, kinda….and I’ll get to that later.
One of the reasons that I studied medicine was to understand the human body in health and disease - and to help my family and friends to navigate the healthcare system. Medicine is really (really) complicated and well over half of the battle is in knowing where to find good information and advice, let alone treatment. Medicine is a discipline that has evolved over centuries and is based on a curious mixture of anecdotes, traditions, trial and error and yes, science. Science is the most desirable of the foundations but of the 1000 piece jigsaw that is medical practice, I would think than less than 50 pieces are truly based upon large-scale, robust and reliable scientific studies.
Still, we are certainly getting there with better and better studies being published daily. Over the course of many years, these studies make their way into mainstream practice via medical educators, clinical guidelines or both. If I’m giving you the impression that the medical system isn’t particularly trustworthy, I’d encourage you to read on a little more before turning to naturopathy or worse still, Dr Google.
Medical training is training like NOTHING else I’ve ever heard of
As a headline, you should understand that it takes about 16 years of full-time study and training to become a junior specialist. And while there are different formats across the world, the most common path to becoming a doctor is to take the smartest kids from high school and put them through 6 years at university (with placements in the hospitals). At the conclusion of the pressure cooker of medical school, junior doctors are employed in hospitals and basically work as apprentices.
But unlike a trade apprentice, junior doctors are directly responsible for people’s lives. All the while, there is fierce competition for the precious few training places in the post-graduate training programs (yep, including general practice). It’s not uncommon for a doctor to need to undertake a Ph.D. just to be considered for a place in some of the most hotly contested specialties. This means becoming a scientist, in order to become a doctor!! In the year I graduated, among the new entrants into ophthalmology there were several scientists, a movie producer, an Olympian and a Rhodes scholar. Specialist training takes between 3 and 6 years, with many taking additional ‘fellowship’ years overseas.
As for my own journey, at university, we had about 30 hours each week in lectures, practicals and tutorials not to mention another 20 to 30 studying the ocean of information that’d been dumped on us. It was relentless and all-consuming. The exams were terrifying, but if nothing else, they taught us how to cope under pressure.
At the end of it all, there was a 7-day period during which I sat my last exam, registered with the health board and was handed a hospital ID and a pager and told to hit the wards! Within just weeks of graduating, I was working in a country hospital on a surgical rotation - staffing the emergency department on my own overnight. I was the only doctor awake within a 100km radius! Within 2 years in the hospitals, I’d looked after the sickest of the sick, resuscitated patients that had suffered cardiac arrest, pulled tic tacs out of ears, watched over unconscious patients in the ICU, seen multiple people in florid psychosis (that were convinced they were Jesus) and delivered half a dozen babies.
My experience is not the exception - it’s the norm. And it’s this experience that neither Dr. Google nor the dozens of pseudo-experts don’t have. The experience is based on lots and lots of look-alikes, where pattern recognition does the trick just nicely with the exceptions. All doctors have heard the expression “common things are common” but those same doctors (as well as the people that taught them) know full-well that masquerades and clinical mysteries are a part of life in health care. In fact, the healthcare system is set up to work in just that way.
For instance, consider the patient that visits their GP to complain of chest pain. The patient, having looked up Dr Google is worried about heart attack - which of course, it COULD be. Within a minute or two, the doctor will be 99% sure that the person has inflammation of their ribs, a muscle strain, a chest infection, or has been having panic attacks. Like I said, it could be a heart attack - and the GP will ponder that - flicking through their mental file of what heart attacks look like, as well as asking themselves if they’ve ever seen a heart attack look like THIS as well as considering the clinical guidelines for the evaluation of a patient presenting with chest pain.
You can think of sensitivity as the disease detector. Say you have an ear infection - what is the chance that your doctor will actually detect that and treat you appropriately. No system is perfect so of course, the odds of any one doctor detecting your ear infection are less than 100%. On the other hand, if you type your symptoms into Google, there is a very good chance, possibly closer to 100% than the doctor, that you’ll come across ear infection as a probable cause. For argument’s sake, let’s say Google is slightly more sensitive than the doctor at detecting a disease.
Specificity is the proportion of cases that the detector got it right when it detected a disease. Using our ear infection example, again, if a doctor diagnoses ear infection in 100 patients, I can virtually guarantee you that not all 100 had an ear infection. It may be 80-90% but of course varies enormously for different doctors, conditions etc. But specificity is where Google fails miserably. And the more alarming the symptoms, the worse things are likely to be. Back to our chest pain example, it’s almost certainly true that the person searching Google for chest pain will start to worry about heart attack. In any case, the doctor is far better at making such a diagnosis because they can look at the patient! And it is the ability to look at the patient - along with knowledge of “red flags” that reduces the real doctor’s risk of error. Indeed, my biggest fear is not that Dr. Google unnecessarily scares the shit out of people: it’s that it may lull someone into a false sense of security. Yes, tinnitus (ear ringing) is almost always benign, but every now and then, it’s caused by a tumour...
Having said all of this, and I do believe that reading strangers threads online remains next to useless, machines are getting better at detecting health problems accurately. This field of medicine is called Clinical Decision Support and is a field of medicine that I am personally involved in on a large scale (I can’t give any more details than that I’m afraid!). Decision support combines things like artificial intelligence with clinical guidelines and diagnostic “rules” to help clinicians handle patients and their presentations in a field that is growing increasingly complex. One of the key aims of decision support systems is to improve both the sensitivity and specificity of a medical service without increasing the cost to deliver that service. I believe that decision support systems will become a necessary and ubiquitous part of all modern health care systems within a decade.
Earlier I alluded to the fact that I too go online to solve medical problems. That’s because, unfortunately, my medical degree hasn’t conferred immunity from health problems for me, my family or my close friends! The difference is, though, that I never just search Google and I especially never ever read threads (unless I want to cause myself undue panic).
So where to look?
I personally use heavily researched and expertly curated medical information sites such as uptodate.com. A subscription to this isn’t cheap but it’s not like I have a choice. The other, though markedly less useful, place to look is PubMed. Pubmed provides a searchable listing of all modern published research studies. The key to using PubMed is knowing how to search as well as what to search for. It’s incredibly powerful but I’m afraid it simply isn’t much good to anyone other than scientists and doctors for the most part. Why is that? Refer to the above statement about medical training.
I could definitely go on and on about this subject but instead, I’m going to now make just one simple recommendation. I make this recommendation in the steadfast belief that all of us (adults at least) are responsible for our own health and the decisions around health care.
And here it is: if you are sufficiently worried about a symptom that you are tempted to ask Dr Google, don’t. Make an appointment to see your local doctor.