Isn't that CT Enough? - Water Cooler Breakdown of CT vs CT/LP for SAH

Heilman, J. Uploaded from Wikipedia Commons

Heilman, J. Uploaded from Wikipedia Commons

Perry et al. Sensitivity of computed tomography performed within six hours of onset of headache for diagnosis of subarachnoid hemorrhage: prospective cohort study. BMJ 2011; 343:d4277.                                                                                                                

Study Objective: To determine sensitivity of head CT within 6 hours in ED patients presenting for possible SAH

Design: Multicenter prospective cohort study at 11 tertiary care centers in Canada from 2000-2009

Main Results:

  • 3132 patients enrolled
  • 240 total cases of SAH (7.7% study prevalence)
  • Overall sensitivity of CT for SAH was 92.9%
  • Overall specificity was 100%
  • Negative Predictive Value was 99.4%
  • Positive Predictive Value was 100%
  • 953 patients (30.4%) were scanned within 6 hours

o   121 (12.7%) found to have SAH

o   All cases identified by head CT (100% sensitivity + 100% specificity)

Table 3: Sensitivity of Computed Tomography for SAH in patients with acute headache(recreated from original text)

Authors’ Conclusions

Modern multi-detector thin slice head CT is highly sensitive for detecting SAH if performed within 6 hours of headache onset and interpreted by qualified radiologist.

“These findings are robust enough to inform the clinical decision […] whether to perform lumbar puncture after negative results on CT.”

Why Should You Care?

  • Headache approximates 2% of presenting complaints to the ED, and SAH is identified in approximately 1% of those patients with headache in the ED.
  • Overall mortality of SAH is high, estimated at 25-50% of patients dying within 6 months
  • If not fatal, SAH leaves approximately 33% of survivors with some appreciable neurological deficit affecting their ADLs.

Summary of the Discussion:

Many thanks to Dr’s Knight, McDonough, Hooker, and Adeoye who actively participated in the discussion and provided valuable insight. Many excellent points were raised, and although a verbatim post might do the contributors justice, it will not benefit those of us wary of the scroll feature.

Although this study raised some interesting points, as of right now the ASA/AHA currently recommend the CT/LP approach to diagnosing aSAH (Class 1 Recommendation, Level B Evidence, Stroke 2012). Heck, if Ope and his cronies cannot diagnose what may or may not be an aneurysm, then what is left for the rest of us? As was stated, if SAH is suspected clinically, then those patients should probably continue to get LP’s to assist in diagnosis or exclusion.

An interesting discussion ensued about the acceptable miss rate for SAH – and other neurologic catastrophes – being essentially zero. Thus, although a “bad” test might still look good for a disease of such low prevalence, it remains essentially indefensible not to complete the currently recommended evaluation if the head CT was non-diagnostic.

Also, to satisfy you methodology wonks, thanks to Dr. Benoit for pointing out that this was indeed a prospective study with the time cut-off decided a priori!


  • Perry et al. Sensitivity of computed tomography performed within six hours of onset of headache for diagnosis of subarachnoid hemorrhage: prospective cohort study. BMJ 2011; 343:d4277.
  • Connolly, et al. AHA/ASA Guideline.  Guidelines for the Management of Aneurysmal Subarachnoid Hemorrhage. Stroke.  43. 1711-1737. doi: 10.1161/TR.0b013e3182587839