The following eye-opening letter on Google’s side-door entry into medicine is from the November 10th issue of the New England Journal of Medicine:
To the Editor: At a recent case conference with a distinguished visiting professor, a fellow in allergy and immunology presented the case of an infant with diarrhea; an unusual rash (“alligator skin”); multiple immunologic abnormalities, including low T-cell function; tissue eosinophilia (of the gastric mucosa) as well as peripheral eosinophilia; and an apparent X-linked genetic pattern (several male relatives died in infancy). The attending physicians and house staff discussed several diagnostic possibilities, but no consensus was reached. Finally, the visiting professor asked the fellow if she had made a diagnosis, and she reported that she had indeed and mentioned a rare syndrome known as IPEX (immunodeficiency, polyendocrinopathy, enteropathy, X-linked). It appeared to fit the case, and everyone seemed satisfied. (Several weeks later, genetic testing on the baby revealed a mutation in the FOXP3 gene, confirming the diagnosis.)
“How did you make that diagnosis?” asked the professor. Came the reply, “Well, I had the skin-biopsy report, and I had a chart of the immunologic tests. So I entered the salient features into Google, and it popped right up.”
“William Osler,” I offered, “must be turning over in his grave. You googled the diagnosis?”
Where does this lead us? Are we physicians no longer needed? Is an observer who can accurately select the findings to be entered in a Google search all we need for a diagnosis to appear, as if by magic? The cases presented at clinicopathological conferences can be solved easily; no longer must the discussant talk at length about the differential diagnosis of fever with bradycardia. Even worse, the Google diagnostician might be linked to an evidence-based medicine database, so a computer could e-mail the prescription to the e-druggist with no human involvement needed. The education of house staff is morphing into computer-search techniques. Surely this is a trend to watch.
Robert Greenwald, M.D.
North Shore–Long Island Jewish Health System
Lake Success, NY 11042
Keyword-driven medical disagnosis, created around baskets of easily-described symptoms, however esoteric: How much more sense could it make?
Related posts:
This is nothing new. I know I’ve used web resources like webmd.com to look up symptoms when I’ve been sick. My girlfriend’s doctor has used it in her presence to find information about a rare condition that she has. It seems like this is just more Google fanboi-ness.
dr K, when I search GREED, i got here. you are no 1 on the ranking. surely you did something neat. but with what you have done, you might well ruin the rankings with many other words or phrases.
dont you think? this is something new!
Whoa, a cynic after my own mind! More seriously, while I generally agree that being surprised that Google is being used widely in Market X is much less surprising than finding out that Google isn’t being used in Market X, I still think this piece is interesting in that you have a practitioner conceding guiltily in a prominent journal that their model is broken. That’s interesting.
“Are we physicians no longer needed?”
“The education of house staff is morphing into computer-search techniques. Surely this is a trend to watch.”
Is there a conflict inherent in these two statements? The Googler in the example was a “fellow”, a post-doc, post-resident physician. That the “peripheral brain” we carried around when I was a resident in the 1980s is now a WiFi connected PDA, Treo, or laptop is just progress, not creative destruction.
I am all for empowered consumers (patients) but I’m also for empowered doctors. Now if we can just fix the healthcare delivery system.
The education of house staff is morphing into computer-search techniques. Surely this is a trend to watch.
Watch, sure. But not out of fear, embrace the trend. This is automation coming to help you diagnose and treat, not take your jobs away.
I think the key here is “easily-described symptoms, however esoteric.” Many (most?) medical conditions consist of a range of symptoms that are neither sufficiently specific nor sensitive for simple computer-based diagnosis.
Medical diagnosis is not an area that has suffered for lack of attention from the AI community, with all the attendant handwringing over the coming obsolescence of doctors. One of the earliest expert systems, MYCIN, was developed in the 1970s to diagnose blood infections.
Which is not to say that Dr. Google won’t happen eventually. But when the revolution comes, I doubt it will be keyword-driven.
why do people continue to think medicine is some arcane art, that mere mortal cannot penetrate or divine? its understanding repeatable principles. its cause and effect. i am not trvializing the field – the diagnoses found on google may have been entered originally by a very wise physician. or a fool. you definitely want a second opinion – so you should yahoo! it too!!! haha.
Mr. Kedrosky,
I’m not sure how it is done in other parts of the country, but in California, all of the medical students that I know are required to get a PDA (personal digital assistant) and put the PDR (Physician’s Desk Reference) on it. There is a program that allows you to put in patient symptoms one by one and likely candidates for illness or disease are displayed. Physicians often do this to “check their work” in essence. Some do it while they make you wait for them between when the nurse sees you and they enter or just after an exam.
There has been much research into this, so suffice it to say that there are more comprehensive databases being searched other than google’s own every day. So while we may get to the day when google can do one more thing for us, this is hardly new in the medical field. There is just too much information for someone to remember everything, so it is helpful to augment a physicians memory and diagnostic ability with some compting power.
This is an interesting letter, and tells you something about the world that most doctors live in.
Google was never designed to act as a medical diagnosis tool, but they are getting beat by it. Imagine what would happen if people actually invested in hypothesis generation from all of the existing literature using scientific heurisitics, performed extensive tests of the automation procedure with existing correct diagnoses, and fed back mistakes.
I think I’m dreaming. Most doctors don’t know what evidence based medicine is.
Doctors who fear this have been obsolete for a decade. NEXT!
surely the use of expert systems in medicine has been going on for some time.
my feeling – from whenever I go to the doctor – is that doctors seem to know the human questions to ask to coax out the information they need to get a correct diagnosis. based on visual, audio and physical cues etc.
although every time I have been to a doctor I can’t help but think that there is some finite number of symptoms etc and if it were all in an ontology (forget Google) then your off to the races. i am surely missing something though since it must be in an ontology somewhere … then again there are also things like Gestalt perception maybe …
It’s here,
http://www.tecn.upf.es/scope/showcase/ontology.htm
Google is exactly what we need to slice through decades of bureaucratic protocol and the moated castles of “health care.”
The healing business is broken, hospitals charge hundreds of dollars to apply band-aids and hundreds more to remove them. Fraud and waste are the rule not the exception…
There are special sites and forums for medicine people. Those who are in tune of course.
The government should organize easy access to Medline and Health topics, medical dictionaries, directories and publications. WBR LeoP
Please remember that the medical community and the scientific community are VERY POLITICAL. WBR LeoP