Table 1

Summary of the vignettes

Summary of vignette	Key elements of diagnostic uncertainty
40-year-old man with 3 years of intermittent abdominal pain, bloating and diarrhoea. No ‘red flag’ symptoms for cancer. A medical history of migraines and depression. No recent travel and no relevant family history. Normal examination, negative FIT (looked for blood in stool) and negative stool cultures. All tests, including FBC, LFT, thyroid function tests, coeliac serology, faecal elastase and faecal calprotectin, are normal. You believe IBS is the most likely cause of his symptoms.	The exact diagnosis here is not 100% certain—although IBS is the most likely diagnosis, there is no definitive test to confirm this. Without colonoscopy±biopsies, there is still a (very) small chance that this could be IBD or even a colorectal malignancy. Most doctors would agree that the chance of these alternate diagnoses are so low that the risks of doing further more invasive tests (such as a colonoscopy) outweigh the benefits.
45-year-old man with no medical history, who presents with central chest pain which came on with mild exertion and lasted 30 min. Normally he cycles 10 miles per day and has never had chest pain before. His maternal uncle died of a myocardial infarction aged 70, but he has no other cardiac risk factors. His examination is normal. CXR, ECG, D-dimer and serial troponins are all normal. You plan to discharge him with no further follow-up.	The investigations are all very reassuring and have essentially excluded serious pathology such as a pneumothorax, pulmonary embolus or myocardial infarction. The cause for the chest pain is not clear—it may be something benign such as acid reflux or a muscular strain, but this is uncertain. There is a small chance that this is a first presentation of angina, although this is less likely given the patient’s lack of risk factors and the fact that he cycles regularly and has never had such pain before.
30-year-old man with no medical history who presented to A&E with a severe headache, which came on at rest over a period of approximately 10 min. No associated loss of consciousness, neck pain, rash, photophobia or vomiting. His examination and observations are normal, as are his routine blood tests. He has a CT of the head within 3 hours of headache onset, which is reported by a neuroradiologist as normal. His headache has improved with paracetamol and is now a dull 3/10 severity. You are going to discharge him without a lumbar puncture (LP).	The normal examination/observations, blood tests and CT scan have essentially ruled out meningitis or a lesion inside the brain (such as a brain tumour). An important diagnosis to consider is a subarachnoid haemorrhage (SAH). Traditionally, a CT was not considered sensitive enough to rule out such a bleed, so if there was a sufficient degree of suspicion patients would go on to have an LP (which is more sensitive at detecting a small bleed). NICE guidance recommends that if the CT scan is done within 6 hours of headache onset, it can be used to exclude an SAH. For this patient, then, we cannot rule out an SAH with 100% certainty, but the risks of doing an LP most likely outweigh the benefits. We do not have a clear cause for the headache—it may be a migraine, but this is uncertain.

A&E, Accident and Emergency department; FBC, full blood count; FIT, faecal immunochemical test; IBS, irritable bowel syndrome; LFT, liver function tests; NICE, National Institute for Health and Care Excellence.