• Temple EM

Ottawa Subarachnoid Hemorrhage Rule

The article: Perry JJ, Stiell IG, Sivilotti ML, et al. Clinical decision rules to rule out subarachnoid hemorrhage for acute headache. JAMA. 2013;310:1248-1255.

The bottom line: Use of the Ottawa subarachnoid hemorrhage rule in the ED to help determine which patients need workup for possible SAH decreases the rate of missed SAH and allows clinicians to safely classify patients meeting none of the criteria as low risk. However, the rule does not decrease the investigation rate of SAH and still need to be validated externally to prove generalizability.


  1. Headaches make up 2% of ED visits; SAH is seen in 1-3%, with ½ presenting with GCS 15 with no new neuro deficit

  2. SAH workup is aggressive: CT -> LP (radiation risk followed by invasive/painful procedure)

Goal of Study: to create a clinical decision rule that would allow physicians to rule out SAH in a certain set of headache patients presenting to the ED without increasing the miss rate


  1. Prospective multicenter cohort study in 10 Canadian urban teaching hospital EDs

  2. Inclusion criteria: 16 or older, chief complaint headache, non-traumatic (no trauma in last 7 days), headache at maximal intensity by 1 hour, A&O, presented within 14 days of headache

  3. Exclusion criteria: 3 or more similar headaches of similar character/intensity over >6 months (chronic recurrent headaches), known SAH, repeat evaluation with CT/LP already performed at prior visit, presence of papilledema or new focal neuro deficits, or prior diagnosis of cerebrovascular aneurysm, SAH, brain neoplasm, hydrocephalus

  4. Assessments: performed by ED attending physicians or residents

  5. For each patient, 19 clinical findings thought to impact SAH risk were recorded prior to the physician ordering studies

  6. Physicians then treated as they would normally based on clinical judgement

  7. Outcome measures

  8. SAH defined as one of the following: Subarachnoid blood on CT, xanthochromia of CSF (visual inspection), or RBC in CSF with aneurysm or AVM on cerebral angiography

  9. Patients that didn’t get CT/LP were followed up at 1 month and 6 months to make sure they didn’t present with SAH within 6 months (would be considered a miss)

  10. Statistical Analysis

  11. 3 decision rules were assessed for sensitivity and specificity. For each rule, patients with any of the following characteristics were not considered low-risk for SAH:

  12. Rule 1: Age >40, neck pain or stiffness, witnessed LOC, or exertional onset

  13. Rule 2: Age >45, arrival via ambulance, presence of vomiting, diastolic BP >100

  14. Rule 3: age 45-55, neck pain or stiffness, arrival via ambulance, or systolic BP >160

  15. Authors also retrospectively analyzed a previous derivation cohort (1999 patients) to determine internal stability


  1. The study included 2131 patients out of 2736 potentially eligible patients

  2. 132 (6.2%) of the patients had SAH

  3. Factors found to be correlated with SAH: older patients, rapid peaking headache, exertional headache, LOC, presence of neck pain or stiffness, presence of vomiting, arrival by ambulance, worse headache of life, high BP, limited neck flexion

  4. Rule 1: 98.5% sensitive, 27.6% specific

  5. Rule 2: 95.5% sensitive, 30.6% specific

  6. Rule 3: 97.0% sensitive, 35.6% specific

  7. Good interobserver agreement for rule interpretation

  8. To increase sensitivity, Ottawa SAH Rule was developed. Patients with any of the following were not considered low risk for SAH:

  9. Age ≥40 years old

  10. Neck pain or stiffness

  11. Witnessed LOC

  12. Onset of headache during exertion

  13. “Thunderclap” headache (maximally intense pain at headache onset)

  14. Limited neck flexion on exam

  15. This rule was 100% sensitive, 15.3% specific – essentially “rules out” SAH

  16. Retrospectively internally validated with patients from a prior study used to develop the first 3 rules

  17. Doesn’t decrease the rate of testing (85.7% testing with this rule vs 84.3% baseline rate)


  1. Rule #1 has 99.2% sensitivity and 99.6% specificity for SAHs undergoing neurologic intervention (when applied to cohort + retrospectively to prior cohort) and would decrease investigation rate to 74%, however, this rule is not 100% specific

  2. Ottawa SAH rule has 100% specificity but is less specific and does not decrease testing rate

  3. Sensitivities of ALL 3 rules studied are higher than current practice which is estimated at 5% for first visit so may decrease miss rate of SAH

  4. It is important to remember these rules should not be applied to rule out any headache etiology other than SAH

  5. Rule still needs to be validated outside of the derivation population in an implementation study

  6. Patients with high (>50%) pretest probability should be investigated irrespective of this decision rule – e.g. strong family history, polycystic kidney disease, or any other reason to think patient is predisposed to aneurysm


0 views0 comments

Recent Posts

See All