Acute headaches account for 1-2% of all ED visits. Of these patients, 1-3% will actually have a subarachnoid hemorrhage (SAH) (Goldstein 2006). This makes it a rare, challenging to make diagnosis but, one with a significant associated mortality. The key for survival in patients with an aneurysmal SAH is recognition followed by prompt endovascular repair or surgery. These patients have a 1-month mortality of 40% and 1.5% re-bleeding risk per day that the condition goes undiagnosed. Unfortunately, population studies have shown that more than 1 in 20 patients with SAH are misdiagnosed on their initial ED visit (Vermeulen 2007).

The current guidelines from the American Heart/Stroke Association in 2012 state that the diagnostic workup for suspected SAH should be a noncontrast head CT (NCHCT), which, if nondiagnostic, should be followed by a lumbar puncture (LP) (Class I; Level B evidence) (Connolly 2012). The Ottawa SAH decision instrument was previously derived from prospective cohort studies of patients who visited EDs with acute headache. The goal of this clinical decision instrument is to determine which patients with headache require investigation to rule out SAH, and which do not. This clinical decision instrument (CDI) is now being validated by a separate prospective cohort study performed by the same authors.

The previously derived clinical decision instrument is as follows, any patient who met all inclusion criteria and none of the exclusion criteria is eligible. If the patient is positive for one or more criteria, they are Ottawa SAH positive. Only those patients with all negative criteria are considered Ottawa SAH negative.

  • Age >40
  • Neck Pain or Stiffness
  • Witnessed loss of consciousness
  • Onset during exertion
  • Thunderclap headache (instantly peaking headache)
  • Limited neck flexion and extension

Clinical Question

Can the Ottawa SAH decision instrument be validated as a highly sensitive CDI in a separate prospective cohort study?


Prospective multi-center cohort study in the EDs of 6 university-affiliated Canadian tertiary-care hospitals from Jan 2010 to Jan 2014. Consecutive patients (age >16) with nontraumatic headache that reached maximal intensity within one hour of onset were considered for enrollment.


Use of Ottawa SAH decision instrument at physicians discretion.


Patients diagnosed with SAH defined by any one of the following: subarachnoid blood visible on NCHCT (from the final report of the local radiologist); xanthochromia in the cerebrospinal fluid (by visual inspection); or the presence of erythrocytes (> 1 × 106/L) in the final tube of cerebrospinal fluid, with an aneurysm or arteriovenous malformation visible upon cerebral angiography.
Follow up performed by structured telephone interview at 1 and 6 months.


Prospective cohort observational study


Glasgow Coma Scale <15, Direct head trauma in previous 7 days, Presentation more than 14 days after onset of the headache, Any history of 3 or more recurrent headaches of similar character and intensity over a period of 6+ months, If the patient came with a confirmed diagnosis of SAH from outside hospital, Papilledema on fundoscopic exam, It patient returned for reassessment of the same headache that was already investigated with NCHCT and LP, Any new focal neurologic deficits, Previous diagnosis of cerebral aneurysm, SAH, brain neoplasm, ventricular shunt or hydrocephalus.

Primary Results

        1153 patients enrolled in the study

  • 590 patients identified as ‘missed eligible’ meaning they met inclusion criteria for the study however were not enrolled by the ED physician
  • 1153 patients enrolled in the study
  • No significant difference in baseline demographic characteristics between the two groups
  • 67/1153 (5.8%) of enrolled patients had SAH and 33/590 (5.6%) missed eligible had SAH
  • Ottawa SAH CDI had a sensitivity of 100% (95% CI 94.6%–100%) with a specificity of 13.6% (95% CI 13.1%–15.8%)

Critical Findings

  • Sensitivity: 100%
  • Specificity of 13.6%
  • (+) LR: 1.16
  • (-) LR: 0


  • The research group performed a prospective validation of their retrospectively derived CDI which is the next step in creation of a robust CDI
  • Sensitivity of 100% makes this an ideal screening test. Every single patient with a subarachnoid hemorrhage was correctly identified when the test was used
  • The same inclusion and exclusion criteria are uniformly applied from the prior derivation study. The exclusion criteria are reasonable as they exclude patients that most physicians would not include as a low risk population
  • Follow up was nearly complete


  • Validation study was performed at the same site as the derivation study, which may overestimate the ability of the physicians to use the CDI correctly and limits external validity
  • Strict inclusion and exclusion criteria may cause incorrect over-application of the CDI by the physician and potentially increase investigation for SAH (indication creep)
  • A large number of eligible patients were not included in the study and it is unclear why these patients were missed
  • Lack of gold standard definition for positive SAH
  • While follow up was nearly complete, it was performed by phone interview which introduces recall bias
  • While the sensitivity was high in the study, the lower limit of the confidence interval was just 94.6% which may be too low for some providers comfort.
  • Both aneurysmal and nonaneurysmal SAH were included in the definition of SAH, given that nonaneurysmal SAH are treated without intervention and are of questionable significance, that means the actual true prevalence of aneurysmal SAH with clinical impact is even lower than reported
  • The low specificity of the CDI means that very few patients will be “ruled out” by application of the CDI and, misinterpretation of how to apply a CDI (I.e. this is a one-way decision instrument) may paradoxically lead to an increase in workup

Author's Conclusions

“We found that the Ottawa SAH Rule was sensitive for identifying subarachnoid hemorrhage in otherwise alert and neurologically intact patients. We believe that the Ottawa SAH Rule can be used to rule out this serious diagnosis, thereby decreasing the number of cases missed while constraining rates of neuroimaging.”

Our Conclusions

The Ottawa SAH CDI when correctly applied in patients that meet all inclusion criteria, has 100% sensitivity for identifying subarachnoid hemorrhage in patients with headache. These patients will most likely not need further work up for SAH unless they are otherwise at high risk. This clinical decision instrument will not decrease the number of patients that receive a work-up for SAH. Conversely, all patients that are Ottawa SAH positive do not necessarily deserve a SAH work-up. Given the specificity of only 13.6% and extremely low positive predictive value of this clinical decision instrument, physicians should rely on their clinical judgement instead.

Potential Impact To Current Practice

The high sensitivity of the Ottawa SAH CDI makes it an appealing addition to clinicians’ approach to headache evaluation but, it is unlikely that this clinical instrument will decrease testing and, it should be noted that the lower level of the CI for sensitivity is just 95%.

Bottom Line

The Ottawa SAH clinical decision instrument was created with 100% sensitivity in mind, which makes it an excellent screening CDI when negative. However, a positive Ottawa SAH screen should not be misinterpreted as evidence to necessarily push for further investigation of a subarachnoid hemorrhage as it is a one-way clinical decision instrument.

Read More

Read More

EM Lit of Note: It’s SAH Silly Season Again!

St. Emlyn’s: JC: Subarachnoid Haemorrhage, Decision Rules & Overtesting Headaches


Goldstein JN et al. Headache in United States emergency departments: demographics, work-up and frequency of pathological diagnoses. Cephalalgia 2006; 26(6): 684-90. PMID: 16686907

Vermeulen MJ, Schull MJ. Missed diagnosis of subarachnoid hemorrhage in the emergency department. Stroke 2007; 38(4): 1216-21. PMID: 17322078

Connolly ES et al. Guidelines for the management of aneurysmal subarachnoid hemorrhage: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2012; 43(6): 1711-37. PMID: 22556195