Difficult diagnostic cases were solved by Google in most cases:
Hangwi Tang and Jennifer Hwee Kwoon Ng, doctors at the Princess Alexandra Hospital, in Brisbane, simply entered words from the case records into Google. The words reflected the symptoms described, and for each case they picked between three and five.They then looked at the first three pages of the Google output “” thirty items “” and chose what seemed to be the most plausible of the diagnoses offered. In 58 per cent of the cases, Google came up with the right answer, or at least the same answer as given in the journal.
For example, when the case involved a 48-year-old man with multiple spinal tumours and skin tumours, the doctors searched Google by entering the words “multiple spinal tumours” and “skin tumours”. Google responded with items suggesting the man had neuro-fibromatosis type 1, the correct diagnosis.
I previously wrote about how I incorporated Google in my practice on a daily basis.
Related posts:
- "I hate doctors" and Google
- Death of the medical riddle and why Google is responsible
- PHRs: Google plays catchup
- Matthew Holt test drives Google Health
- "House" and Google hot trends
- Kidney Notes quoted in the Google Annual Report
- Goodbye Google Answers
KevinMD.com on Facebook
{ 1 comment }
Common things happen commonly, or, when you hear hoof beats, think horse not zebra.
Getting things right 58 percent of the time is about as much good as asking a medical student.
Although this might not be a completely apt comparison. When NEJM ran its weekly Case records of the Mass Gen hospital there were always diagnoses sought from the visiting expert, local attendings and local med students. The students there had an impressive track record, better than the visiting expert physician. It was however speculated that as the cases were generated from the local hospital, the students had the advantage of local institutional memory.
If one has no clue where to begin, then casting a net broadly to include as many ideas as possible is a good idea. Then one must wheedle the list down without spending as much money as possible. The rational practice of medicine is the parsimonious selection of the fewest and least invasive tests that will confirm the diagnosis sufficiently to lead to appropriate intervention.
Any fool can order every test possible and then let the consultants figure out what it all means. As a patient you have to choose whether you want every possible exam done to you or just enough to get the job done in all probability.
Comments on this entry are closed.