Letters Subarachnoid haemorrhage

Reinventing the wheel, square

BMJ 2010; 341 doi: https://doi.org/10.1136/bmj.c7414 (Published 30 December 2010) Cite this as: BMJ 2010;341:c7414
  1. Alistair J Jenkins, neurosurgeon1
  1. 1Regional Neurosciences Centre, Newcastle on Tyne NE4 6BE, UK
  1. alistair.jenkins{at}nuth.nhs.uk

Perry and colleagues analysed the cases of 1999 patients to try to formulate “a clinical decision rule to rule out subarachnoid haemorrhage” in patients presenting to emergency departments.1

Neurosurgeons—the clinicians ultimately dealing with the condition—have an extremely low threshold for investigating patients for subarachnoid haemorrhage. We will investigate, or agree on investigations for, pretty much anyone with a sudden severe headache continuing for more than a few minutes. That’s it. The rest doesn’t really matter—and the idea that someone would not refer a patient because they didn’t arrive in an ambulance, or that the headache didn’t come on with exertion, really doesn’t bear thinking about. Most of the small amount of medicolegal work I do deals with missed subarachnoid haemorrhage, and this paper is guaranteed to increase that.

One or two things are simply wrong: computed tomography should not be preceded by lumbar puncture, for very obvious reasons.

While it may be interesting in retrospect to assess positive diagnostic features, experience shows that these are often absent or confusing. In such a potentially lethal condition the “clinical decision rule” is simple: if you think it might be a subarachnoid haemorrhage, it is worth a scan, and if results are negative, a lumbar puncture. In that order.


Cite this as: BMJ 2010;341:c7414


  • Competing interests: None declared.

  • Editor’s note: A correction regarding the order of computed tomography and lumbar puncture has been published to the print version of the paper by Perry and colleagues.


View Abstract

Sign in

Log in through your institution