Having declared myself not just a fan but a follower of evidence-based medicine, I feel a bit embarrassed to admit that I practise gut-reaction medicine too. When I speak to colleagues, though, I feel reassured: others also talk about the time they ordered a chest X-ray, a series of blood tests or even admitted someone to hospital for woolly symptoms to which no evidence-based protocol would respond. Trainee GPs are wisely warned to beware the person who describes seemingly innocuous symptoms – but whose attendance at the surgery is low and whose previous medical notes are sparse.

There are other times, too, when we should let our guts lead: when chest infections have gone on too long; when faced with a clutch of disparate yet worrying symptoms; and when a neat diagnosis is elusive.

Recently, I’ve been wondering if gut reactions are actually evidence-based in their own way. Medicine is all about managing uncertainty: there is rarely a point where you can be entirely sure of a diagnosis or the impact of a treatment. Surely the gut reaction is just a synthesis of honed experience, information about the patient and knowledge of the patterns of illness in that patient’s community?

Bayes’ theorem – that’s Thomas Bayes, an 18th-century mathematician and church minister – states that the odds of a hypothesis being true change as it is put to the test. In practice, this means the hypothesis will change, too. Could gut reactions actually be quick Bayesian responses emerging as pieces of evidence are collected? And could they be as good as other, more formal, diagnostic tests?

We’re on the road to finding out. A study in the journal BMC Family Practice tried to define “gut feelings”. The researchers found that there were two types – one of alarm and one of reassurance. For example, the instinctive reaction at seeing a patient who has lost lots of weight might well be concern, whereas doctors probably won’t panic when they hear a young, healthy-looking, loquacious patient describe chest pain – but also say that it’s not brought on by exercise; this is unlikely to be coronary artery disease.

In general practice medicine, sometimes it’s just as important to exclude serious problems as it is to come to an exact diagnosis. There’s time, and specialists, for the winnowing away of uncertainty – but the patient will be better off going through that process having been reassured by his GP that nothing serious is wrong. To go a step further, should we sometimes follow gut reactions to such an extent that we set about excluding any serious illness – then skip the specialists and see if those “unlikely to be serious” symptoms go away on their own?

Margaret McCartney is a GP in Glasgow
margaret.mccartney@ft.com

For lively discussion of the latest medical issues go to the FT’s Health Blog