Diagnosis is an odd thing: occasionally it is easy and satisfying (at least to the doctor). More often it is elusive, taking time, tests, and possibly time again to become clear. I’m not even sure if I do make a definite, cast-iron diagnosis that often. Another way of looking at symptoms, and which seldom appear in a textbook list of classical complaints, is whether or not these symptoms are likely to be part of something sinister. By this I mean something likely to maim or kill. I am less happy about ‘watchful waiting’ when there is potentially serious underlying disease. One does not wish to over-investigate, a problem which has adverse consequences for the NHS and patient. Equally, one wishes to act promptly when it will help.
The journal Pediatrics has asked a lot of paediaticians if they made mistakes. Over half said they did, at least once or twice a month. But their ‘mistakes’ weren’t all what I’d class as a ‘mistake’. They listed, as an example, giving antibiotics when the diagnosis was viral, and antibiotics wouldn’t work. They claimed this as a ‘mistake’ because the child could suffer side effects of the medicines with no potential gain.
Is treating this as a mistake useful? Supposing a child came in with suspected meningitis – the correct, safest, treatment includes immediate antibiotics. So when the test results eventually come back negative and the child is better: does that mean it was a ‘mistake’ to use the drugs? I don’t think it’s fair or useful to categorise that as a ‘mistake’ . I can’t say that is either poor practice either: because the risk of getting the diagnosis of meningitis wrong is so dramatic, it seems to me to be appropriate care to run the risk of ‘over-use’ in this situation.
There are plenty of mistakes when medicine is practiced, but getting diagnoses wrong is not usually because of a clear error. A ‘mistake’ to me is when you give penicillin when someone is known to be allergic. Diagnosis in practice is a bit messier. You can certainly get it wrong, but usually it’s a case of calling the odds first.