Tool Helps Rule Out Bleeds in Acute Headache

Pauline Anderson

September 26, 2013

A newly developed decision tool may help clinicians determine which patients presenting in the emergency department (ED) with headache require investigation for subarachnoid hemorrhage (SAH).

A study to validate prospective decision rules found that a tool that includes 6 possible clinical features — age 40 years or older, having neck pain or stiffness, loss of consciousness, onset during exertion, thunderclap headache (instantly peaking pain), and limited neck flexion — was highly sensitive, correctly identifying SAH cases presenting to the ED with acute nontraumatic headache with normal neurologic findings.

"We are ready at this point to say that anyone with one of the high-risk features should be investigated for subarachnoid hemorrhage," said the study's lead author, Jeffrey J. Perry, MD, an emergency physician and senior scientist, Ottawa Hospital Research Institute, and associate professor, University of Ottawa, Ontario, Canada.

"I would argue that it's ready to use in that regard, but where we are not quite ready is with patients who have no high-risk features; in that case, we cannot without some further validation say that they do not require investigation."

It's important to try to perfect a decision tool to detect SAH in the ED because some 5% to 6% of cases are missed on their first visit. This, said Dr. Perry, "is probably an unacceptably high number" given that prompt treatment leads to a better prognosis.

"While most cases are identified, there is an important number of patients who are missed, and we hope that our rule will be able to basically get that to as close to zero as possible."

The validation study was published in the September 25 issue of JAMA.

Actual Costs

Dr. Jeffrey J. Perry

Patients suspected of having SAH are typically evaluated with computed tomography (CT), followed by lumbar puncture if results of the CT are negative. Both treatments involve some costs, said Dr. Perry.

"There are the actual costs to the healthcare system, and there are time delays to doing these tests in already overcrowded emergency departments. For patients, there is exposure to radiation with the CT scan and with lumbar puncture, it's uncomfortable, it's frightening, and it can be associated with complications including rare but serious ones, such as meningitis or epidural hematoma, that can result in paralysis," he pointed out. "And there is a real possibility of getting a post–lumbar puncture headache that can be as bad or worse and more prolonged than the original headache."

In previous research, investigators developed 3 potential clinical decision rules for SAH. Each rule calls for investigation of whether 1 or more of the high-risk findings are present. These rules are as follows:

Rule 1:

  • Age 40 years or older

  • Neck pain or stiffness

  • Witnessed loss of consciousness

  • Onset during exertion

Rule 2:

  • Age 45 years or older

  • Arrival by ambulance

  • 1 or more episode of vomiting

  • Diastolic blood pressure of 100 mm Hg or over

Rule 3:

  • Age 45 to 55 years

  • Neck pain or stiffness

  • Arrival by ambulance

  • Systolic blood pressure of 160 mm Hg or over

The objective of the current prospective multicenter study was to assess the accuracy, reliability, clinical acceptability, potential for rule refinement, and potential effect of these candidate rules in a new cohort of neurologically intact patients with acute headache.

The study was conducted in the EDs of 10 university urban Canadian tertiary care teaching hospitals, some of which participated in the previous derivation study. The study enrolled 2131 patients aged 16 years and older whose chief reason for visiting the ED was a headache that reached maximal intensity within 1 hour, had a Glasgow coma score of 15 (and so were alert and oriented), had not sustained a full or direct head trauma in the previous 7 days, and had presented within 14 days of headache onset.

Patient Assessments

All patient assessments were made by attending physicians certified in emergency medicine or emergency medicine residents supervised by staff physicians. After assessing patients, but before ordering imaging or cerebrospinal fluid analysis, physicians recorded 19 clinical findings on data forms (these were variables found to be significant in the previous derivation study or to be possibly clinically important for potential rule refinement).

Physicians answered the following questions related to interpretation and use of the 3 proposed SAH rules:

  • Are investigations indicated for this patients according to the decision rule (yes or no)?

  • How comfortable would you be in actually using the rule for this patients (5-point scale from every comfortable to every uncomfortable)?

The primary outcome — subarachnoid hemorrhage — was defined by any of the following: subarachnoid blood on unenhanced CT of the head; xanthochromia in the cerebrospinal fluid; or red blood cells in the final tube of cerebrospinal fluid, with an aneurysm or arteriovenous malformation on cerebral angiography.

Patients underwent evaluation with nonenhanced CT, lumbar puncture, or both, with cerebrospinal fluid analysis according to the judgment of treating physicians, who were instructed not to alter their practice according to the proposed rules.

Patients discharged without both CT and normal lumbar puncture findings were evaluated by using the study's Proxy Outcome Assessment Tool, which included a structured telephone interview at 1 month and 6 months after ED assessment as well as medical records review to identify patients who developed SAH.

Of the 2131 patients, 132 (6.2%) had SAH. Decision rule 1 had the highest accuracy, with 98.5% sensitivity (95% confidence interval [CI], 94.6% - 99.6%) and 27.5% specificity (95% CI, 25.6% - 29.5%). Rule 1 missed 2 patients, rule 2 missed 6 patients, and rule 3 missed 4 patients.

The best model to predict all cases of SAH was rule 1 plus thunderclap headache (instantly peaking pain) and limited neck flexion (defined as inability to touch chin to chest or raise the head 8 cm off the bed if supine). This new rule, which had 100% sensitivity (95% CI, 97.2% - 100.0%) and 15.3% specificity (95% CI, 13.8% - 16.9%) was labeled the Ottawa SAH Rule.

"We aimed to improve the rule which is what we did with the Ottawa SAH Rule, so we refined the rule, added 2 additional variables which then identified all cases of SAH in this cohort and indeed in previous cohort of 1999 patients," said Dr. Perry. "So it's quite robust as far as our data."

However, although the new rule has near-perfect sensitivity and may lead to fewer missed cases, it comes with a loss of specificity and increased testing and associated costs, note the authors.

"The Ottawa SAH Rule does not lead to a reduction of testing (ie, CT, lumbar puncture or both) vs current practice; however, it may help to standardize which patients with acute headache require investigations, and its widespread use could help decrease missed subarachnoid hemorrhages," they write.

Headache accounts for about 2% of all ED visits, and SAH is one of the most serious diagnoses, accounting for 1% to 3% of these headaches.

Important Caveats

In an accompanying editorial, David E. Newman-Toker, MD, PhD, Department of Neurology, The Johns Hopkins University School of Medicine Baltimore, Maryland, and Jonathan A. Edlow, MD, Harvard Medical School, Boston, Massachusetts, question whether the Ottawa SAH Rule is clinically useful. They also point out several important caveats for application of this rule.

"Effective use of any decision rule requires careful attention to clinical details affecting its generalizability. Does the patient meet all original inclusion criteria, such as having a headache that peaked in less than an hour? Has an examination been performed carefully enough to verify that neurological status is truly normal, including no papilledema? Is subarachnoid hemorrhage the only target diagnosis being considered, or are unstudied, rare, yet important issues of sudden-onset headache (eg, cerebral venous sinus thrombosis, pituitary apoplexy, arterial dissection) still part of the differential diagnosis? Are other unstudied variables (eg, family history of brain aneurysms) present that might complicate interpretation of the rule?"

The rule may miss diagnostic causes of headache other than SAH if the headache develops over hours or days instead of over seconds to minutes, say the editorialists. "Medical emergencies such as obstructive hydrocephalus, giant cell arteritis, bacterial brain abscess, and fugal meningitis can present with more gradual-onset headaches without focal neurologic or other red-flag features," they write.

Any reduction in missed cases of SAH assumes that the rule is correctly and consistently applied, they added. "Because some aspects of the rule depend on subjective physician interpretation (eg, headache peaking 'instantly'), subtle physician biases (eg, linked to physician risk tolerance) might lead to underuse or overuse of imaging unrelated to true disease risk."

According to Dr. Newman-Toker and Dr. Edlow, CT and lumbar punctures present relatively low risk and are only modestly inconvenient, and the Ottawa SAH Rule may have greater utility in situations, although uncommon, in which the CT–lumbar puncture combination is not readily available or cannot easily be applied.

"While awaiting further scientific advances, clinicians may find the refined Ottawa SAH Rule helpful to guide diagnostic decisions, but they should limit its use to patients with acute headache who are similar to those among whom the rule has been evaluated."

Dr. Perry has disclosed no relevant financial relationships. Dr. Newman-Toker reported serving as a board member for the Society to Improve Diagnosis in Medicine and receiving grants or grants pending from the Agency for Healthcare Research and Quality, National Institutes of Health, and Centers for Medicare & Medicaid Services. Dr. Edlow reported providing expert testimony for both plaintiff and defense in cases related to neurologic emergencies.

JAMA. 2013;310:1248-1255, 1237-1239. Abstract  Editorial


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.