According to the Daily Mail, tests have shown that researchers can make successful diagnoses of difficult conditions, using a Google serach based on symptoms. Once junior doctor commented that a rare illness popped right out of a search.
The accuracy rate (measured by comparing the results with correct diagnosis in 26 difficult cases) was 58%. So its not quite a replacement for the local GP.
According to the Mail, “Studies of autopsies have shown doctors seriously misdiagnose fatal illnesses about 20 per cent of the time.So millions of patients are being treated for the wrong disease. And the more astonishing fact may be that the rate has not really changed since the 1930s. ”
I suspect that the rare and difficult diseases may in some ways be easier for Google to tackle than the common ones. Presumably the main mis-diagnosis issues are diseases that look like something less serious. (Meningitis in children, for instance. Or a man I knew who died of malaria he caught in Africa, which his local GP in Scotland took to be a less serious form of the disease.)
But this is yet another interesting demonstration that broad patterns can be helpful, and that database searches can help us to find them. The interesting thing would be to link diagnoses to massive datasets about a persons life, and see if other, causative, patterns emerged.