MIA 2012: Backes D, et al. Time-dependent test characteristics of head computed tomography in patients suspected of nontraumatic subarachnoid hemorrhage. Stroke. 2012 Aug;43(8):2115-9

Bottom Line:
  • 100% sensitive and specific if < 6 hours from headache onset
  •  90% sensitive if after 6 hours
A noncontrast head CT can effectively rule out atraumatic subarachnoid hemorrhage (aSAH) in patients who present with acute headache within six hours after ictus. Those who present outside this time window or present atypically for SAH (ie neck pain) require further workup, including a lumbar puncture.

Why It’s Important for Emergency Medicine
The majority of patients presenting to emergency departments with headache are suffering from primary headache. Emergency physicians are usually concerned with ruling out dangerous causes of secondary headache. In a majority of cases involving young, otherwise healthy people, this means evaluating for subarachnoid hemorrhage. A gold standard for making this diagnosis exists in the lumbar puncture, which is nearly 100% sensitive. However, LPs are an uncomfortable procedure prone to causing further headaches, not to mention time consuming and often technically difficult. Therefore, it is desirable to determine in which patients SAH can be safely excluded with a simple CT scan.

Major Points
Of 137 patients who underwent head CT < 6 hours after ictus: 
  • 68 had positive findings (mostly aSAH or perimesencephalic hemorrhage)
  • Of the 69 inconclusive head CTs who underwent CSF fluid analysis, only one dangerous secondary HA was identified, a bleeding cervical AVM. This patient presented not with headache but vomiting, neck pain and stiffness. 
Of the 113 patients who underwent head CT > 6 hours after ictus:
  • 37 had positive findings (primarily aSAH and PMH). 
  • Of the 76 who underwent CSF fluid analysis, four cases of aSAH were identified, as well as one thoracic AVM. 
  • Several other secondary headaches were also identified including a case of bacterial meningitis, viral encephalitis and viral meningitis. 
Design
Retrospective study from a tertiary referral center for patients with SAH in the Netherlands. Patients included were all those between 2005 and 2012 with a clinical suspicion of aSAH. Patients were retrieved from two prospective databases. One included consecutive patients with confirmed SAH and the other included all patients receiving LP with CSF spectrophotometry. GCS < 15, referral with confirmed SAH, unknown time of ictus and focal deficits on presentation.

1039 considered patients, 789 excluded. 137 underwent head CT < 6 hours, 113 > 6 hours.

Criticisms
The study is retrospective; there is a possibility of selection bias. There were many excluded patients, however, mostly for good reason (505 pts w GCS < 15). Almost universally, having a GCS < 15 will warrant further workup if the CT is negative. CT scans were read by an experienced neuroradiologist in this study, a luxury many institutions do not have.

Reviewed by A. Williams

Reference
Backes D, Rinkel GJ, Kemperman H, Linn FH, Vergouwen MD. Time-dependent test characteristics of head computed tomography in patients suspected of nontraumatic subarachnoid hemorrhage. Stroke. 2012 Aug;43(8):2115-9. Pubmed

MIA 2012 = Most Interesting Articles series of 2012