Death of the medical riddle and why Google is responsible

May 25, 2009

Stanford’s Abraham Verghese challenges readers to solve his medical riddle, without using Google.

These exercises, when thought through, offer the student the opportunity to “formulate hypotheses, go to the book, research and eliminate possibilities . . . and come to the answer,” and can be a valuable learning experience.

But with the dawn of Google, many arcane answers can simply be looked up, often at the expense of thinking through a problem.

So, as Dr. Verghese poignantly asks whether “Google killed the riddle,” try pondering his latest offering, without search engine assistance.

Can a medical condition really give someone the ability to hold their breath for twenty minutes? And no, a tracheostomy is not involved.

KevinMD.com on Facebook

Follow on Twitter Subscribe

{ 10 comments }

1 VendorMD May 25, 2009 at 6:37 pm: Cool Riddle. I saw the answer at Verghese’s blog.
2 Richard G. Lanzara, Ph.D. May 25, 2009 at 6:53 pm: Odine’s curse.
3 Dr. Grumpy May 25, 2009 at 9:19 pm: Google is more a pain because of what cyberchondriacs will find when they type in symptoms. Such as numbness, tingling, weakness, fatigue, etc.

In my field (neurology) all searches lead to a self-made diagnosis of MS.
4 Matt S. May 25, 2009 at 9:29 pm: Ahhh yes, Google HAS killed not only the riddle, but more practically the challenge of the difficult diagnosis. I’m currently a medical student and although I adore a tricky patient more than anything else in medicine, I’ve ruled out internal because of this looming change. I feel that the much of the job satisfaction of being an internist will be history as soon as computer software can out-diagnose and out-manage doctors. And it WILL happen.

[sniff]
5 Anonymous May 25, 2009 at 9:33 pm: death?
6 Tom May 25, 2009 at 9:48 pm: Odine’s curse? That’s just the elimination of unconcious respiratory drive. Try hypothermia or extreme emphysema (if one were able to take out all the CO2, then it may be possible for a severe COPDer to survive on the stored oxygen).
7 Marky May 26, 2009 at 3:58 am: Anyone know where I can find more riddles like this one, and the one about the glass eye and the big liver? It was kind of fun trying to find the answer without Google as an upcoming M2 student.
8 ArkyDoc May 26, 2009 at 11:56 am: I’m a general internist working at a critical access hospital with no specialists on staff. I LOVE Google/UpToDate/PubMed, etc. I love internal medicine for the challenge, but it is wonderful to be able to “check my answers”. These are people we work on – not crossword puzzles! Anything that helps us grow in our fund of knowledge and improves patient care is a real benefit. And, even with all the resources available, every day still has challenges. No two patients are alike in their presentation of illness. Matt S., don’t rule out internal medicine just yet – there are real people (patients and future colleagues) who need young physicians who love a diagnostic challenge. And, computers can’t appreciate the “bio-psycho-social” nuance presenting with every interesting (or common) disease.
9 Joseph F. Sucher, MD FACS May 31, 2009 at 9:23 am: Matt S.

Computers are so far away from out-diagnosing and out-managing doctors that you can go an enjoy internal medicine and retire before this ever happens. The fact is that we have been trying to implement computer systems for diagnoses since the 1950s (Dr. Ledley), and with great effort in the 1970s – 80s (Dr. Shortliffe – SUMEX-AIM). Much effort has been put into advanced AI systems and Bayesian logic with very little real practical return. Until we can build true neural networks that can accept data as natural language, sound, sight and smell we will not have computers be able to do what an expert can. There is much to be said for the subconscious thinking that occurs as an expert comes upon a patient and is able to “thin-slice” otherwise imperceptible clues that can’t be picked up by the crude datapoints we currently use to enter into a computer system. Additionally, it has been folly to make a computer for diagnosing everything including the common cold, or cholecystitis or diabetes. How many of these things do you have to see to make these kind of common diagnoses? But that is exactly what has been attempted for 40 years.

Let me add now that I am an expert in computerized clinical decision support and a fervent advocate for implementation of computers in medicine. But I have a significant difference in belief in how we should leverage computers now versus how we have tried in the past. That is, these systems can be used effectively to aid the clinician in making the best decisions at the right time for individual patients. The key here is using computers as an “aid”. They can augment our capabilities. They are poor replacements for us.

So, will computers do a better job that an expert clinician? Maybe someday. I won’t rule it out. But its more likely that they will continue to help us be better clinicians for a long long time to come.

JFS
10 Brad K May 31, 2009 at 7:07 pm: While Google can help every doctor (and patient) access a much broader array of useful information, the process of diagnosis is as much art as science and medical conditions are so complex that many can never be consistently distilled to an algorithm. Look at Clinical Problem Solving in NEJM or Diagnosis in NYTimes – plenty of cases evade diagnosis until the right MD sees them. There are still an abundance of medical riddles out there – though the means to solve them are changing and improving. What can’t be replaced, though, are smart and committed MDs (maybe with a bit of computerized assistance). Extensive discussion of this phenomenon at http://hyoumanity.blogspot.com/.