“Imagine If Doctors Relied on Google as Much as Programmers Do”
codelab.farai.xyz
codelab.farai.xyz
He was an excellent diagnostician, but when he was unsure or stumped by a set of symptoms he would outright just use Google. All the partners at the practice did. The skill was then sorting the wheat from the chaff in the search results, and then making the correct decisions for treatment.
Doctors don't just use specialist software to look things up, they often just use Google like the rest of us, it's bloody fast and with personalisation of search results, it often returns answers most relevant to you (or your profession).
The problem is more that:
1. speed is crucial in clinical medicine
2. UpToDate access is usually buried under several layers of corporate links or simply unavailable because too expensive
3. Google is inferior to UpToDate, but it's only a Ctrl+L away
Although really, we mostly rely on memory because looking things up isn't fast enough for the throughput we are required to keep up with. Web search is reserved for things where we really don't want to mess up.
How are we, as patients, can choose a doctor which has UpToDate and uses it ?
And how a patient, which can read the medical literature ,and use Google/Pubmed/Medical books, benefit from a subscription to your system to help validating a doctor's work(because they're human too, they sometimes make mistakes) ?
Of course, assessing results only has its limits. You can't know as a patient if you are treated optimally for a cancer that causes no symptom but that may kill you in 5 years. I think the only thing you can do to act positively on that is go to a big center specializing in your pathology. Notice that I say 'center' and not 'professor' or 'reputed specialist'. Modern medicine is a system, not an individual (as brilliant as he/she may be).
tl;dr: if you're not happy with the results of your current doc, find another one that suits you better.
That's the ideal.
It's possible in some cases(if you have a big medical issue diagnosed), maybe less possible in others(unexplained symptoms for example that your GP may not consider as important).
You. yourself, when someone in your family faces a medical issue outside of your specialty , do you sometimes do research on the subject ? do you talk with the doc about that research ?
Some docs have bigger egos and will do what you say and research and act outside of their specialty for themselves or their family. In my experience, this very often leads to subpar care (and even quite horrifying failures, sometimes).
Now, one can of course look things up to discuss them with their doc and learn why said doc chose whichever option they chose. That shows you're interested and your doc will usually happily oblige if he/she is not already overworked and completely burned out. If your doc's answer to that doesn't satisfy you, then maybe it's better to change.
There is not much personalization on Google except on your location, which is not going to help you have more relevant results as a doctor
I'm not very good with recalling specifics, but I'm very good at recalling how or where to find the specifics.
As with your father, expertise is necessary to interpret and determine which information is relevant, and to notice discrepancies between the found information and the concrete issue at hand.
In programmers avoiding looking up references is far more an ego driven ascetic approach than it is anything practical. Same with eschewing modern tooling.
Can't speak for anybody else, but I did not. Nor do I particularly intend to. I'm here for the conversation, not the article itself.
Now, what were we talking about?
What you're hiring when you hire a doctor, lawyer, accountant, or software engineer is judgement. Either X is simple and don't worry about it, or this will have a simple solution and here it is, or this will take more investigation and let me look stuff up, or this is a specialist problem and I know who you need.
Of course it's often the case that the professional will just give you an answer, because there's certain bread and butter things that everyone asks about. But that shouldn't trick you into thinking they've memorized their whole field, you just haven't asked an interesting question.
Fun episode at the doctor's office: was in for a routine checkup. Joked with doctor about patients that google their symptoms and come up with all manner of wild diagnoses. Remember to ask about a minor symptom I'd been experiencing for about a week. Doctor Googles symptom. Eyes meet. She blushes, I almost fall off my chair laughing.
I think doctors have to cram so much in because they need to have all that knowledge available to pick up what to google.
One of the 2 guys and I had quite some experience with "real life" systems in general. The other guy, really smart but his experience was in competitive programming (leetcode like things, super complex algorithms, etc).
At some point something broke in production. While looking at the first logs, the algorithms guy gave his diagnostic... but the other guy and I knew it couldn't be such thing due to some very subtle things in the error that was happening.
That kind of experience makes the difference of the time it takes to find the root of a problem.
It was an indirect inspiration for MacRumors' roundups. Where we summarize the current information on a particular product rumor, so you don't have to keep up with the "literature".
There is a lot work being done on augmenting the surgeon - solving the problem of access to information at the right time in the right place would dramatically improve work in that field.
Currently, if you’re making a decision in the OR, all you have is your knowledge/experience, and maybe when/if you’re lucky, of the other surgeon/s who happens to be physically available to consult the situation. This is something I witnessed personally when visiting an OR - a surgeon ran in from another OR to get a quick consult on an unfamiliar situation.
There are a few products and ideas I’ve seen for headsets which allow others outside of the room to have tele-presence.
It’s not hard to imagine how much having an instantly searchable and accessible global knowledge base for an operator would be. The newest methodologies, know-how, research, reference media etc.
And this guy wasn't some fly-by-night quack either... he was a very experienced and highly regarded local doctor who had been in his practice a long time.
So yeah, let's not kid ourselves into thinking that doctors don't Google for the specifics on occasion just like we do.
https://en.wikipedia.org/wiki/Wikipedia:WikiProject_Medicine...
Sometimes you can only guess at the meta though, for example I'm building an RTS type game at the moment and I needed to figure out how to get units to move together and pack around the destination point. Even though the scene is 3d they are only moving on a 2d plane, so I searched for a circle packing algorithm and found usable results.
Sadly we're not looking for this skill when interviewing, there they'd have asked me to display skill in different algorithms and structures, language questions etc.
Administrative regulations, now that changes really fast. At least in my jurisdiction. But thank God, the problems stay the same.
Beside, they google all the time. They don't rely on it, the way programmers may do, but it's much faster than consulting their professional networks and data sources.
Of course it depends a lot on their professional area.
As an input you would provide the typical information doctors look for otr get from the user (cough, the elbow hurts, ...) and then suggest
99.0% common flu
0.5% ankle is broken
0.5% <neurological scary illness>
It would help to avoid missing a broken ankle when this is flu season and the MD assumes that this is "like everyone".
(the examples are extreme for some lightness)
Do you know if this is an active branch of research where there is enough interest to put good devs? Or just a curiosity?
More than an AI kind of system, I was thinking about something more predictable, intended to supplement the MD memory.
I imagine that if you input the symptoms you see, some basic measurements (temperature, blood pressure, O2 saturation, ..) + some input from the general news (flu season, allergy season, ...) you end up with the typical diagnosis.
But then, there could be other pathologies th eMD does not think about immediately but when displayed would ring a bell.
My wife for instance had several mis-diagnoses after our second child, only to find with a doctor that she may have a neurological issue. Then came the neurologist. Then came MS.
In her case time was not that of an issue but maybe in other cases that would help.
Soless AI and more "decision tree" kind of software.
I know a decision tree seems attractive to someone coming from science. But in clinical practice, it's totally useless.
“Sorry, we can’t do the heart surgery today, turns out our scalpels are made by one guy as an unpaid side project and he’s on vacation and it also turns out that for the last 5 years they’ve been mistakenly made of lead but no one noticed.
Also, we stopped doing heart open heart surgeries like 2 years ago. We now do micro-surgeries where we just stab the patient over and over again until we hit the right spot. ”
Any expert in any field should hone their skills via google.
This is the good old "you wouldn't download a car". Everyone would download a car if they could. And doctors would consult Doctor Overflow, MD all the time if such a thing existed. Perhaps it should?
That said, I'm not surprised to learn that doctors use Google. We go to doctors to benefit from their judgment and experience, not because we need to learn facts from them.
In fact differential diagnosis are just decision trees with a different name
But you need to know the usual stuff and the day to day stuff by memory (same with programming)
Anybody can call themselves a programmer and get hired by a credulous manager.
* - not the only time a software bug was directly responsible for death(s), but likely the most famous.
I happen to agree with both takes and frankly I think if the people involved sat down and had a conversation, they probably would agree too. Looking things up is part of the job, but it is not the only part of the job and distilling programming down to “professional googling” is a disservice. I can look up what Bell’s Palsy is and read all I want, but it’s not the same as a doctor looking it up and reading about it. The same is true with programming.
In reality, someone with NO school and 12 years on the job is worth a lot more to my team and I, ymmv.