Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
“Imagine If Doctors Relied on Google as Much as Programmers Do” (farai.xyz)
29 points by xojoc on Dec 17, 2021 | hide | past | favorite | 80 comments


My father was a GP, which obviously meant he didn't just see patients with symptoms related to a specific specialism he had studied intensely. He saw patients with a broad variety of symptoms that needed a diagnosis to decide if he could just administer treatment himself, or might require a referral to a consultant.

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).


Edit: hospital dev if you are reading this, know that UpToDate.com must be 1 CLICK AWAY! :-)

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.


UpToDate looks great.

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) ?


In all honesty and with the best of intentions from an academic doc, I think that trying to validate your doc's work other than from what you think of the result is a bad idea. Modern medicine is actually not that modern and still requires a massive amount of know-how that doesn't show in literary resources. For a trivial example, consider that UpToDate is really only valid for mainland US: an infection will be treated differently in Europe because of different resistance patterns. The list of such examples is infinite.

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.


Thank you for the recommendation to go to a specialized medical center.

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 ?


No, I don't do research myself on issues that are not my specialty. Either the next step is obvious, or it's beyond my reach. Precisely because I lack the necessary experience and know-how to apply research results to the clinical problem at hand. For my family, I refer them to someone I believe is good because I've seen them in action and judged them to be good from my limited perspective, if possible.

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.


A doctor doesn't need Uptodate to manage your run of the mill asthma and diabetes :). Best tip? Do your own research and take it to your GP. Either they appreciate it, or they dismiss it. That's the best way to differentiate between good+bad doctors.


As a doctor myself, I can assure you that it's possibly the worst way to differentiate capable from incapable docs. It's a good way to differentiate socially skilled docs from socially inept ones. This I can agree with.


Most hospitals provide UpToDate for all providers with privileges at the hospital.


To add to this, which at least in the systems I'm aware of, being a GP is a specialisation. General doesn't mean non specialist. It means being a specialist at general medicine.


This is country-dependent. In practice though, it's absolutely true that GP is specialist work.


potato potato, the meaning is obvious. GP differs from a medical specialist in that they focus on diagnostics and broad knowledge as opposed to deep narrow focus. Whether you call focusing on that a specialization is irrelevant.


Calling focusing on that a specialization means a 6y education difference in some countries.


As a programmer one of the skills I value the most is being able to find and use information from external sources. These days that primarily means the internet.

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.


That shows that there's a need for a specialized database of cases where physicians can search for symptoms and add new information. If it was only used by medical personnel it would filter out various blogs and forum posts that probably are getting worse and worse and more time consuming to sift through.


> with personalisation of search results

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


It also might prefer to tell you the patient has scurvy instead of cancer if you’ve been searching for and buying pirate toys for your niece recently.


Did the first few commenters not read the article at all? It’s literally about how Doctors look up information all the time. This makes perfect sense because not only is there so much information to remember best practices and other information change all the time.

In programmers avoiding looking up references is far more an ego driven ascetic approach than it is anything practical. Same with eschewing modern tooling.


Did the first few commenters not read the article at all?

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.


That's fine in itself, particularly as these threads get longer. I was just super confused that everyone assumed the opposite to what is actually true and patently clear after a cursory read of the article. I don't think those comments were actually about conversation or ultimately very curious or intelligent which is what HN is supposed to be about.


Me too.

Now, what were we talking about?


You read the openers for sophisticated small talk and more or less funny banter at the club?


go back to Reddit.


No, thank you. Too rude over there.


There's no knowledge job that doesn't rely on reference material. If a field is a knowledge field, the material cannot be anything but too big and detailed to remember the entirety of.

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.


What amazes me about medicine is how incredibly much students have to cram into their heads in a short amount of time, and that they are capable of that. I don't think I could. What is even more amazing is that they still start almost from scratch when faced with the reality of practicing medicine. Because there is still so much to learn and so much to keep up with.

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.


it's like the joke about the plumber who just whacks a pipe with a hammer to fix a clog and charges $200. the client protests, saying he should only charge $5 for such a simple fix. the plumber itemizes the bill - $5 for hitting the pipe, $195 for knowing where to hit it.

I think doctors have to cram so much in because they need to have all that knowledge available to pick up what to google.


This actually also happens I our (tech) field: Several years ago I was working with 2 other very smart developers. We were the tech team in a small startup, in charge of keeping the systems running.

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.


They have lots of reference material for the non usual stuff, I remember looking for something that came wrong on a blood test and landing on the NHS GP manuals and it was full of flow charts for diagnosis and when to refer to an specialist etc


The big resource for doctors (or was back in my day) is UpToDate.com - which includes up-to-date clinical summaries of diseases, treatments, with references to clinical studies that back them.

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".


Lol, I remember reading a story about how you quit everything to run macrumors. Why were you on uptodate way back then? (I'm also a daily reader of macrumors!).


Former physician, so very familiar.


Using Google to augment memory is fine for both engineers and doctors. Using it to replace comprehension is acceptable for neither. If a doctor administered insulin to a diabetic without any understanding of the pancreas or the body's blood glucose regulation pathways, that would be bad (yes, nurses do that, but under the supervision of a doctor). If a doctor uses a program to calculate the dose, there is nothing wrong with that.


I remember going to my doctor years ago - pre google. She dragged out this big book and had a look through it, I forget the name. She said at the time she used to go out the back and have a look through it so patients wouldn't see her, but stopped worrying about it.


It's because doctors know that they're so well respected, they have to keep the appearances up too. At least a big book looks more official. Wondering where I can find one of these big books so it doesn't look like I'm using google :P


No, it's because many patients don't trust you anymore when they see you looking things up instead of answering immediately without thinking. #-(


I'd rather have a doctor looking things up than one that thinks they can remember everything from school/training.


I work with surgeons on a product for skill development and lately I’ve had the opportunity to go to a few conferences that specialized in endoscopic surgery.

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.


I don't know how much the average doctor uses Google, but here's an anecdote that may be slightly relevant. I was visiting my GP for something or other a couple of years ago, and we were sitting in his office going over some stuff about some medicines and stuff, and he turns to his computer, Googles something, winds up on a Wikipedia page (in addition to some other pages), and after a minute or two turns back to me and goes "OK, yeah, so you should be taking <blah, blah, whatever>".

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.


There's a bunch of doctors who realized that doctors were ending up on Wikipedia a lot, and who try to keep the medical pages on Wikipedia somewhat sane.

https://en.wikipedia.org/wiki/Wikipedia:WikiProject_Medicine...


I find that as the scope of information grows that meta-information becomes far more important. I don't need to know how to solve a problem, but I need to know the terminology for how to get to information about it.

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.


They don't program humans, they maintain or repair them, help them to cope and soothe them. The system they're working on doesn't change that fast biological, albeit a bit faster social. The main components are millions of years old. Methods and tools do change over their career a lot, but not that fast.

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.


I always wondered why there is no software to aid diagnosing that is used by all doctors.

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 then suggest a set of tests to got to XX% where the diagnosis is "good enough".

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)


This exists. Current implementations work badly enough even in the trivial case that they are completely useless. Making a better implementation would involve technology that we don't have yet. Also, https://news.ycombinator.com/item?id=27428671


Thank you for the information.

Do you know if this is an active branch of research where there is enough interest to put good devs? Or just a curiosity?


Automated diagnosis and clinical recommendation engines are seeing a lot of research, yes. As a clinician unfortunately, I can already tell you that all that will lead pretty much nowhere. The first problem to solve is data collection/information retrieval, and the solution is not "see, with this clean data my model achieves 101% accuracy, that's better than a human" but will be a systemic effort to enforce policies about data collection and storage. Only then can we hope to use this data consistently.


I did research in physics - something much simpler than the human body so please take this as a disclaimer :)

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 hear you. Like I said, all attempts using decision trees have up to now failed miserably. The main problem is that for any given set of symptoms, the ~5 most likely diagnoses will be very obvious to most docs, but there will be a gigantically long distribution tail of less likely diagnoses not even worth testing for. This is especially problematic for rare diagnoses, and I'm sorry to say that your wife unfortunately fell into that category. Then come all the other UX issues, such as what search terms to use (different users will use different names), etc.

I know a decision tree seems attractive to someone coming from science. But in clinical practice, it's totally useless.


More targeted towards the patient herself, but pretty much what you described: https://ada.com/


That was one of the first attempted applications for GOFAI in the 80s. Check out CADUCEUS:

https://en.wikipedia.org/wiki/CADUCEUS_(expert_system)


If the medical industry worked like modern web development then I would expect them to use Google more.

“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. ”


We do. I use Google images heavily. You just use the right words to filter out the lay stuff ("conjunctival injection and chemosis differential" instead of "red swollen eye") and then you skim thru images until you find the exact diagnostic situation you're looking for. It's fast and specific. Havent used uptodate since the 2000s because it is not free.


The only reason other professions aren't relying as much on Google (with emphasis being on the relative intensity of use) is that their problems are harder to search (they don't get standardized error messages) and that they have levels upon levels of leeches asking for money in exchange for information.


I very much hope my doctor relies a great deal on Google.

Any expert in any field should hone their skills via google.


I wonder what the equivalent for a doctor is for me with the "complete guide to flexbox" is that I seem to Google every time I forget the difference between align items, justify items, align content.. just can't seem to get that into my head ..


UpToDate. Paid subscription.


Imagine if doctors read books instead of just winging it.

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 phrase is just flamebait and that's probably why it's framed in quotation marks.

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.


Beyond a few well-curated sites like DuodenumOverflow, I think they could get in just as much trouble with shallow and incomplete websites as programmers do. There are many poor results for many programming queries, only a few good or deep authorities.


Why imagine? If I'm not mistaken there are reference guides for symptoms in different specialities

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)


I've worked with healthcare in many different roles. I can tell you conclusively that doctors and nurses are absolutely googling the issue. Though often not actually google but something equivalent.


The difference with non doctors is 1- knowing how to do a clinical exam 2- more often than not when looking at a disease, knowing it’s epidemiology or deducing it 3- knowing which questions to ask :)


To become a doctor you complete 5+ years university training then internship/residency before you are able to practice privately.

Anybody can call themselves a programmer and get hired by a credulous manager.


And yet, as the article says, doctors still rely on Google a ton!


The condition between those two things is an "AND" not an "OR".


As a programmer, I have a trivial way of finding out if a solution I found on the internet really works: run it. What is the equivalent procedure for doctors?


We search the web for remembering things we already know. We are not looking for insight on Google. If we really have a tough nut to crack, we consult with experts. If I am the expert on the topic, then I review recent research papers.


Human anatomy doesn't change; Computers and the software that run them change constantly.


Programmers can afford to mistake. Doctors, pilots, train mechanics, firefighters don't.


Read about Therac-25* before making such a proclamation.

* - not the only time a software bug was directly responsible for death(s), but likely the most famous.


What does the color gamut picker at the bottom of the page do?


I noticed there was some Twitter discourse over something DHH had written about not having looking things up and copy/pasting being your main skill. There was some backlash and a lot of people responded emphatically that they look things up all the time and discouraging people from looking things up is bad, etc.

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.


Agreed with everything you've just said. I'm a programmer and if I googled "how to do X in ruby on rails" I could probably dollow the instructions and have the output out the other side. On the other hand my partner is not technical at all and would likely get very stuck on some assumed knowledge about a text editor, environment variables, command prompts or some mysterious error message!


Well, imagine if engineers did 12+ years in university before coming to a codebase.

In reality, someone with NO school and 12 years on the job is worth a lot more to my team and I, ymmv.


which part of the article is this a response to?


News flash: they basically do.


We do.


(they do)


Which vaccine was it?


Maybe if doctors had to actually build a human they would. When I'm debugging I reach for Google a lot less than when I'm building something.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: