Excluding Subarachnoid Haemorrhage: Do we need the needle?

The current practice for the diagnosis of subarachnoid haemorrhage (SAH) in the ED follows a dependable pathway of non-contrast CT as soon as possible, followed by a lumbar puncture (LP) after 12 hours from headache onset, if the CT is negative.

As new data are generated, this approach continues to be brought to question. Not only is there little evidence behind the timing of the LP, there is now data to suggest that an LP may not be required at all.

Lay down your needles

One of the strong proponents for not doing an LP following a negative CT is David Newman of The NNT and Smart EM fame. He calculated out that the "NNT" for LP to diagnose one extra SAH amenable to intervention after a negative CT scan to be 700. His argument is that the risks of performing this many LPs outweighs the benefits of this one positive diagnosis, a viewpoint not agreed on by all.

The key study in this push comes from Perry et al in Ottawa—home of many great studies into the efficient use of resources. Published in the BMJ in 2011, this study showed that the sensitivity of modern CT in the subgroup of patients scanned within 6 hours of headache onset to be 100%.


This study sample had a SAH incidence of 7.7%, which seems a little high. This may be due to excellent patient selection by the clinicians, or the SAH definition used: positive blood on computed tomography, visual xanthochromia (spectrophotometry being largely unavailable), or red blood cells in the final tube of CSF (which may result from a traumatic tap). The use of LP to confirm/exclude diagnosis was not universal, but at clinician discretion; those who did not have an LP were followed up to 6 months.

Importantly, this study benefited from highly experienced neuroradiologists interpreting the scan results, which may not generalise to the immediate service in the average ED. A number of scans were initially misinterpreted in this study prior to neuroradiology review.

Despite these limitations, these data formed the foundation of suggesting that any modern CT within 6 hours could be taken as excluding SAH.

Another needle in the coffin

This article, published in Neurology in May, adds further weight to this movement. This study tested the assumption that the accuracy of the CT within 6 hours was dependent on having a specialist neuroradiologist, or whether the findings could be generalised to non-academic hospitals with general staff radiologists.


In this multicentre, retrospective case series, 760 patients who presented with acute headache, normal level of consciousness and no focal deficits had a CT within 6 hours reported as negative and subsequent CSF sampling. Of these, 52 had positive CSF spectophotometry, but only 1 of these was found to have any SAH on review of imaging, and this was thought to be non-aneurysmal. Follow up imaging of those with a positive CSF yielded 8 patients with aneurysms, 3 already clipped, and the rest not thought to have ruptured—all thought to be incidental to the presentation.


The initial radiology interpretation was performed by staff radiologists, not neuroradiologists. The imaging review was performed by 2 neuroradiologists and a neurologist. This study concluded that the negative predictive value of a staff radiologist report was 99.9%, missing only 1 case of the 760, with this case being deemed as non-aneurysmal with no readmission for SAH for over 2 years.

It is important to note that CTs interpreted by residents or registrars are not included, and that the setting for this study, in the Netherlands, dictates the experience of a staff radiologist to be at least 5 years.

From this, the authors conclude that:

  • The missed diagnosis rate following a negative CT would be 1 in 15,200
  • The false positive rate in subsequent LPs (6.7%) would lead to significant ongoing investigation and potential management of incidental aneurysms or other pathology
  • These findings support:

a change of practice wherein a lumbar puncture can be withheld in patients with a head CT scan performed within 6 hours after headache onset and reported negative for the presence of SAH by a staff radiologist


What are your thoughts? Will you stop doing LPs in this setting, or start to include these data in your shared-decision-making discussions with your patients?



Blok KM, Rinkel GJE, Majoie CBLM, et al. CT within 6 hours of headache onset to rule out subarachnoid hemorrhage in nonacademic hospitals. Neurology 2015;84:1927–32. doi:10.1212/WNL.0000000000001562

Perry JJ, Stiell IG, Sivilotti MLA, et al. Sensitivity of computed tomography performed within six hours of onset of headache for diagnosis of subarachnoid haemorrhage: prospective cohort study. BMJ 2011;343:d4277. doi:10.1136/bmj.d4277

Share your thoughts