Online Symptom Checkers

A long review of online symptom checkers (a Google system of searching medical texts) appeared on this Dutch web site WebWereld that made for interesting reading

The downside to these tools is the problem medical students, doctors, nurses and most other healthcare professionals suffer from – the recurrent fear they are suffering from <insert latest disease you read/studied here> (as a balancing side bar to this I went to medical school with someone who suffered a ruptured brain aneurysm and recognized his symptoms and walked himself to the ED/Casualty)

As the writer points out in many cases getting an appointment is tough proposition in the best of circumstances no matter what country you are in and when you do it requires getting there, waiting in a room full of lots of other people who are sick (increased risk of contracting additional infection!) and unless you are really lucky waiting interminably to be seen. Add to that the distinct probability that you receive treatment that may not be warranted (antibiotics for a viral infection for instance) simply because the expectation has been set that you took the day off, made the trip the least you should receive is a prescription for your trouble.

The sites reviewed

WebMD – which interesting did not produce different results even if you answered the questions it asked differently
Revolution Health – basic mapping to the body and symptoms and really more about educating the user than diagnosing (to include the usual disclaimer of calling your healthcare professional no matter what)
Mayo Clinic – still very broad with little to narrow searches to anything useful
About.com – yikes a general site for clinical diagnosis seems like a bad idea and sure enough no ability to get to useful information quickly enough
WrongDiagnosis.com – now you would expect this given the name to be a better performer but with 10,000 medical conditions and enormous lists of symptoms to choose form unwieldy

Their summary (not all scientific but worth listing)

• WebMD: 12 clicks to get a list of 15 possible causes
• Revolution Health: 2 clicks to get a list of 13 possible causes
• MayoClinic.com: 5 clicks to get a list of 7 possible causes
• About.com: 9 clicks to get a list of 3 possible causes
• WrongDiagnosis.com: 1 click to get a list of 6 possible causes

The New York times feature today covered the struggles of young adults balancing limited income with large insurance premiums and high cost of living. As one participant put it

There was no way that I could pay my rent, buy insurance and eat

So despite the limits of these systems they are being used and for many probably with reasonable success but this is tempered with the negative effect that include antibiotic misuse and resistance or masking of problems with pain medication:

We see people with urinary tract infections taking meds better suited for ear infections or pneumonia — the problem is, they haven’t really treated their illness, and they’re breeding resistance….Or they take pain medicine that masks the symptoms. And this allows the underlying problem to get worse and worse

There is no easy answer to this but as discussed in the NY times article coverage is cost prohibitive for many and the bills out of line with people’s financial circumstances.As with all caveats – Your Mileage May Vary (YMMV). It is hard to assess the usefulness of these sites as a fully trained physician. Years in medical   and clinical practice mean that you intuitively apply filters to this information and can navigate these sites as aide memoir’s to help as Zebra Hunters. So for those of you not clinically trained I’d be interested in your thoughts on any of the above sites – how did you find them

For the purposes of complete information these sites are long step away from real diagnostic support tools which base their clinical decision support on a more detailed history that is detailed enough and with specifics to help narrow the choices for causes for specific symptoms based on knowledge of disease symptoms and signs. The rule of thumb is that 80% of the diagnosis comes from the history, 15% from the physical examination and the remaining 5% from additional investigation. That mix has changed with the array of diagnostic tests and imaging choices but this is not necessarily a better more cost effective service. Case in point some years back when a relative of mine injured themselves skiing and was transported to the first aid station. They waited for their spouse to show up before allowing any kind of investigations to be carried out (in this case X-rays of the foot). Their spouse happened to be a clinicians who examined and diagnosed the specific fracture much to the amazement of the local clinicians who said that was impossible to tell without an X-ray….end result correct diagnosis and the treatment remained the same despite the X-ray. Sure it could have been more complicated but did it require an urgent (and expensive X-ray) there and then…..no way!

But there is real scope in the clinical world to use Diagnostic support tools and one I saw some years back showed tremendous promise – First Opinion. There are others Isabelle is probably the most well known and the concept is sound but the challenge is seeing the wood for the trees and getting to relevant information quickly. Integrated into an electronic medical record, driven by clinical data there is tremendous scope help support clinicians arriving at the right diagnosis quicker and more accurately and overcoming the tsunami of medical knowledge. But to get to that data requires the data to be entered somehow……forms, drop down boxes and hunt and click systems just don’t meet the need and turning clinicians into data entry clerks is not a viable or sound strategy. I made this point recently in a presentation and this was commented on in  the MTExchange forum regarding the concern over transcription jobs and the fear that this these jobs are disappearing. EMR vendors are using transcription as part of their ROI justification suggesting that the “elimination of transcription costs is as a sales selling point”. As Julie Weight said in her posting:

The fallacy in this is that highly-skilled, highly-paid professionals then perform the task of a medical transcriptionist – which makes absolutely no sense. Physicians struggle enough with the financial justification for an EMR; coupled with resistance to change, it has pretty much guaranteed slow adoption of EMRs, especially in smaller practices.

And went on to hammer this point

you wouldn’t find Jack Walsh typing his own annual reports or Warren Buffet manually entering stock values into the computer (well, maybe Warren does – but I’ll bet Jack doesn’t)

Exactly! Stop the madness, allow clinicians to capture information without creating a burden of data entry for the sake of billing and administrivia. Clinical documentation is supposed to support clinical care, so lets give clinical documentation back to the clinician, provide them with the means to capture all the fine detail necessary in a detailed narrative description that contains this detail but include the data elements that are necessary to feed the EMR’s. Allow clinicians to capture the full Healthstory that contains both these elements and satisfies the clinical need and computer’s insatiable demand for structured data.