medpundit |
||
|
Saturday, November 11, 2006So a team at the Princess Alexandra Hospital in Brisbane identified 26 difficult diagnostic cases published in the New England Journal of Medicine last year, including obscure conditions such as Cushing's syndrome and Creutzfeldt-Jakob disease. They selected three to five search terms from each case and did a Google search while blind to the correct diagnoses. Google gives users quick access to more than three billion medical articles. The researchers then selected and recorded the three diagnoses that were ranked most prominently and appeared to fit the symptoms and signs, and compared the results with the correct diagnoses as published in the journal. Google searches found the correct diagnosis in 15 (58 per cent) of cases. Respiratory and sleep physician Dr Hangwi Tang, who led the study, said: "Doctors adept at using the internet use Google to help them diagnose difficult cases. I confess, I have done this on occasion. I once made the mistake of doing it in front of the patient only to be met with an expression of quizical disappointment. I think she was thinking "I could have done that myself!" posted by Sydney on 11/11/2006 02:43:00 PM 3 comments 3 Comments:
I remember doing a preceptorship with a rural doc (a pulmonologist by specialty, a generalist of necessity) who was working with some obscure outfit that was trying to develop "diagnostic software." There was a race to develop many similar projects back then, in the early 90's, so my preceptor kept the details to himself (but he did let me play with the prototype for a few minutes). The idea was that a practicing physician could sit down at the computer after evaluating a patient with an unclear diagnosis, enter a series of signs, symptoms, and test results, and the software would regurgitate a list of possible diagnoses. The physician then did his or her part: applying intuition to glue the correct diagnosis to the patient. By 2:13 PM , atThe problem in diagnosis is usually not finding disease, it's excluding the healthy worried well from running up a fortune in tests. Remember in medical school how easy it was to self diagnose things like lymphoma in one's self? Computer based diagnostic programs are no smarter than the people writing them. The advantage of computers lies in their ability to work without getting tired or irritable and handling mounds of data. Maybe we should have computers to sign off on the myriad of home care and nursing home orders that plague me, and let the human MDs diagnose illness. Oh ,and one more thing: if you are stressed and sad, would you rather get a hug from a human or a programmed diagnosis from a PC? By 9:50 AM , atComputers maybe used for checking from a series of data gathered from a patient, maybe used to help patients remind them of medication time. But it is impossible for computers, because healing really comes within us, which are the nutritions/cells and emotions/way of thinking By 3:08 AM , at |
|