Comments on: Death of the medical riddle and why Google is responsible http://www.kevinmd.com/blog/2009/05/death-of-the-medical-riddle-and-why-google-is-responsible.html medical blog Wed, 17 Mar 2010 23:46:10 -0400 http://wordpress.org/?v=2.8.4 hourly 1 By: Brad K http://www.kevinmd.com/blog/2009/05/death-of-the-medical-riddle-and-why-google-is-responsible.html/comment-page-1#comment-91682 Brad K Sun, 31 May 2009 23:07:54 +0000 http://www.kevinmd.com/blog/?p=29907#comment-91682 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/. 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/.

]]> By: Joseph F. Sucher, MD FACS http://www.kevinmd.com/blog/2009/05/death-of-the-medical-riddle-and-why-google-is-responsible.html/comment-page-1#comment-91666 Joseph F. Sucher, MD FACS Sun, 31 May 2009 13:23:11 +0000 http://www.kevinmd.com/blog/?p=29907#comment-91666 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 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

]]> By: ArkyDoc http://www.kevinmd.com/blog/2009/05/death-of-the-medical-riddle-and-why-google-is-responsible.html/comment-page-1#comment-91507 ArkyDoc Tue, 26 May 2009 15:56:09 +0000 http://www.kevinmd.com/blog/?p=29907#comment-91507 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. 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.

]]> By: Marky http://www.kevinmd.com/blog/2009/05/death-of-the-medical-riddle-and-why-google-is-responsible.html/comment-page-1#comment-91496 Marky Tue, 26 May 2009 07:58:46 +0000 http://www.kevinmd.com/blog/?p=29907#comment-91496 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. 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.

]]> By: Tom http://www.kevinmd.com/blog/2009/05/death-of-the-medical-riddle-and-why-google-is-responsible.html/comment-page-1#comment-91486 Tom Tue, 26 May 2009 01:48:34 +0000 http://www.kevinmd.com/blog/?p=29907#comment-91486 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). 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).

]]> By: Anonymous http://www.kevinmd.com/blog/2009/05/death-of-the-medical-riddle-and-why-google-is-responsible.html/comment-page-1#comment-91485 Anonymous Tue, 26 May 2009 01:33:35 +0000 http://www.kevinmd.com/blog/?p=29907#comment-91485 death? death?

]]> By: Matt S. http://www.kevinmd.com/blog/2009/05/death-of-the-medical-riddle-and-why-google-is-responsible.html/comment-page-1#comment-91484 Matt S. Tue, 26 May 2009 01:29:36 +0000 http://www.kevinmd.com/blog/?p=29907#comment-91484 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] 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]

]]> By: Dr. Grumpy http://www.kevinmd.com/blog/2009/05/death-of-the-medical-riddle-and-why-google-is-responsible.html/comment-page-1#comment-91483 Dr. Grumpy Tue, 26 May 2009 01:19:29 +0000 http://www.kevinmd.com/blog/?p=29907#comment-91483 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. 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.

]]> By: Richard G. Lanzara, Ph.D. http://www.kevinmd.com/blog/2009/05/death-of-the-medical-riddle-and-why-google-is-responsible.html/comment-page-1#comment-91478 Richard G. Lanzara, Ph.D. Mon, 25 May 2009 22:53:39 +0000 http://www.kevinmd.com/blog/?p=29907#comment-91478 Odine's curse. Odine’s curse.

]]> By: VendorMD http://www.kevinmd.com/blog/2009/05/death-of-the-medical-riddle-and-why-google-is-responsible.html/comment-page-1#comment-91477 VendorMD Mon, 25 May 2009 22:37:43 +0000 http://www.kevinmd.com/blog/?p=29907#comment-91477 Cool Riddle. I saw the answer at Verghese's blog. :) Cool Riddle. I saw the answer at Verghese’s blog. :)

]]>