Subarachnoid haemorrhage (SAH) can be a difficult diagnosis in patients who present to the emergency department (ED) with acute headache but are alert with no neurological deficit.1

The classical teaching is that SAH cannot be excluded by non-contrast CT and so patients should proceed to lumbar puncture (LP) if the initial CT is normal. However, there has been a recent shift in practice—driven by emerging evidence1 2—away from routinely offering LP to exclude SAH.

In this issue of the EMJ, the Trainee Emergency Research Network (TERN) report on the subarachnoid haemorrhage in the Emergency Department (SHED) study.3 SHED was a multicentre prospective cohort study that asked whether a normal CT brain within 6 hours is sufficient to exclude SAH in patients attending an ED with acute headache. The SHED investigators recruited adults with non-traumatic acute headache reaching maximal intensity within 1 hour and that were fully alert. They found that the sensitivity of CT for SAH was 97% (95% CI 87% to 99%) if performed within 6 hours, which is consistent with findings from earlier prospective studies and a recent diagnostic test accuracy systematic review. They also reported that the sensitivity of CT persisted after 6 hours: 95% (87% to 99%) at 6–12 hours, 94% (75% to 100%) at 12–18 hours and 75% (47% to 93%) at 18–24 hours.

SHED represents a monumental effort by emergency medicine trainees and research teams working in half of UK …