ARTICLE IN BRIEF
Neurointensivists respond to a study in the Sept. 25 issue of theJournal of the American Medical Association that identified a new clinical decision rule with a 100 percent sensitivity rate for ruling out subarachnoid hemorrhage in patients who presented with acute headache in the emergency department.
In the emergency department (ED), when is acute headache an indication of a life-threatening condition, such as a subarachnoid hemorrhage (SAH)? And how can neurointensivists ensure an accurate diagnosis when diagnostic delays can be fatal? In a new study published online ahead of print in the Sept. 25 issue of the Journal of the American Medical Association (JAMA), Jeffrey J. Perry, MD, from the Ottawa Hospital Research Institute, and colleagues found that six clinical factors resulted in a 100-percent sensitivity rate for recognizing SAH (95% CI, 97.2%-100.0%) and a 15.3 percent specificity for only finding SAH (and not another condition) in patients who presented with acute headache in the ED.
Dr. Perry and colleagues found that having one or more of these clinical factors best predicted 130 of 132 cases of SAH: age ≥ 40 years; neck pain or stiffness; witnessed loss of consciousness, onset during exertion — plus “thunderclap headache” (defined as instantly peaking pain) and limited neck flexion on examination (defined as inability to touch chin to chest or raise the head 8 centimeters off the bed if supine). But the authors wrote the tradeoff for “near-perfect sensitivity” may be “loss of specificity, increased testing, and increased associated costs.”
Dr. Perry, associate professor in the department of emergency medicine, senior scientist at the Ottawa Hospital Research Institute, and research chair in emergency neurological research at the University of Ottawa, and colleagues had previously proposed a set of three clinical decision rules for ruling out SAH in a 2010 paper in the British Medical Journal. (See “The Ottawa SAH Rule.”)
In the current multicenter cohort study, they set out to validate and improve the reliability of these three proposed rules, and identified the risk factors that were most sensitive for ruling out SAH.
“Among patients presenting to the emergency department with acute nontraumatic headache that reached maximal intensity within one hour and who had normal neurologic examination findings, the Ottawa SAH Rule was highly sensitive for identifying subarachnoid hemorrhage,” they wrote. “These criteria will allow physicians to use evidence-based medicine to determine which patients are at high risk for SAH and who should be investigated,” added Dr. Perry, cautioning that additional implementation studies are required before this rule may be standardized for use in the ED.
Dr. Perry and colleagues reviewed the medical records of 2,131 adults who presented with headache which peaked within one hour but who had no neurologic deficit upon examination. They had been seen at 10 university-affiliated Canadian tertiary care emergency departments from April 2006 to 2010.
The main outcome measure was subarachnoid hemorrhage, determined by computed tomography (CT) and lumbar puncture (LP) tests. Most participants (82.9 percent) had CT done; 39 percent had a lumbar puncture performed. The investigators identified SAH in 132 patients (6.2 percent).
“Prior to this study, there has not been an evidence-based method of determining which headache patient is high risk for SAH. As such, physicians have been forced to use their gestalt to determine which patients to investigate,” Dr. Perry told Neurology Today. “We strongly encourage physicians to utilize the [Ottawa SAH Rule's] high-risk features to identify patients who require investigation for SAH. For patients who are deemed low risk by the rule, physicians will still have to use their clinical judgment to determine if their patient requires investigation or not for SAH.”
The next step is a study “to determine how the Ottawa SAH Rule performs in clinical practice, assess the actual effects on patient care and patient outcomes, and conduct a formal health economic analysis. Additional study could assess the relative benefits in rural vs urban settings.” Dr. Perry said they are currently conducting a validation study of the Ottawa SAH Rule.
Neurointensivists not involved with the study noted that identifying SAH in patients coming to the ED with a headache has been a challenge for the past two decades. Headaches are among the most common presenting symptoms in the ED, they said.
Dileep Yavagal, MD, associate professor in the department of neurology & neurosurgery at the University of Miami Miller School of Medicine, told Neurology Today that this study adds to the emergency physicians' and neurologists' diagnostic skills, because sometimes patients without thunderclap headache are discharged from the ED without a CT or an LP, “and I think that's an important finding from the study, that even when the headache is somewhat slower in its peaking, if the patient has other high-risk features, CT is indicated, and if CT is negative then LP is indicated.”
The other clinically useful finding, he said, is that it is important to consider neck pain or limited neck flexion as high-risk feature for suspicion of SAH. In his experience, Dr. Yavagal said this risk factor is often overlooked in SAH, and may be ignored or viewed “as part of benign headache syndrome.” The study by Dr. Perry and colleagues “is one of the most well-done and definitive clinical rule studies,” he said.
Rajat Dhar, MD, an attending physician in the neurology/neurosurgery intensive care unit at Barnes-Jewish Hospital in St. Louis, MO, agreed, adding that although SAH was the cause in only 6 percent of headache cases in the JAMA study, “missing this uncommon but serious condition can and often does lead to worsening [events], including aneurysm rebleeding which can even be fatal.”
Dr. Dhar said, with the Ottawa SAH rule, “if you don't have any [high-risk] features — which few don't — then you can be discharged without further investigation or imaging.” This is the first rule, he said, “to demonstrate such high negative predictive value in a large cohort of emergency department patients. As such, it may be useful in the emergency department to guide triage and work up.”
Dr. Yavagal noted that 94–96 percent of the emergency department physicians queried in the study said they were comfortable with applying this rule, indicating a high potential for routine clinical application.
But the low specificity “fundamentally limits this rule from majorly affecting care,” and may actually lead to more testing, Dr. Dhar told Neurology Today.
Using the clinical rule in the study, one would probably do more CT scans than necessary, said Dr. Yavagal, and in some patients you would do a potentially painful procedure (a lumbar puncture), but you wouldn't miss any patient who had SAH. Dr. Yavagal said that high sensitivity and low specificity is acceptable with a dangerous condition like SAH where “the outcome really depends on early treatment and the morbidity of a missed diagnosis is extremely high.”
The other limitation, which is important to keep in mind, he said, “is that [the rule] does not give us any guidance about clinical decision making in acute headache developing over days.” As the authors stated, “this is only for those patients in whom headache peaked within one hour and SAH was suspected. So, it's important to not try to apply this rule in all headache scenarios,” Dr. Yavagal warned.
Lastly, Dr. Dhar said that there may be differences in the findings in different countries. He noted that “almost all headache patients get CT scans in the US, which is not the case in Canada,” where the study was conducted. The rule may not reduce the use of CT in the US, he said, “but it could be used to determine who needs and who can avoid a lumbar puncture.”
Future studies, wrote David E. Newman-Toker, MD, PhD, and Jonathan A. Edlow, MD, in an accompanying editorial in the same issue of JAMA, “should seek to validate the Ottawa SAH Rule using larger samples. Realistically, though, this may require use of administrative data and imputation of missing results. The rule should also be studied for the effect on patient outcomes as part of a clinical care pathway for headache diagnosis, ideally with direct comparison to an alternate care pathway based on the CT-LP rule.” Before full implementation, Dr. Dhar added, a validation study of how using the rule impacts outcomes and costs (versus standard care) would be optimal.
Dr. Yavagal said: “My expectation would be that [implementing the Ottawa SAH rule] would not significantly increase the current costs. Actually, if it did lead to any decrease, that would be important to know.” He said that the costs of implementation in different health systems would be interesting to study. For example, “in the US, the litigation rates are much higher and therefore one might end up finding that this rule actually saves some cost in avoiding unnecessary work-ups in the US. It might not be the same in the Canadian system,” he said.
THE OTTAWA SAH RULE
For alert patients older than 15 years with new severe nontraumatic headache reaching maximum intensity within 1 hour; not for patients with new neurologic deficits, previous aneurysms, SAH, brain tumors, or history of recurrent headaches (≥ 3 episodes over the course of ≥ 6 months).
Investigate if ≥ 1 high-risk variables present:
- Age ≥ 40 years
- Neck pain or stiffness
- Witnessed loss of consciousness
- Onset during exertion
- Thunderclap headache (instantly peaking pain)
- Limited neck flexion on examination
Source: JAMA 2013; 310(12): 1248–1255.
TUNE IN: In a new study published in the Sept. 25 online issue of the Journal of the American Medical Association, Jeffrey J. Perry, MD, from the Ottawa Hospital Research Institute, and colleagues found six high-risk clinical factors which resulted in a 100-percent sensitivity rate for ruling out SAH in patients who presented with acute headache in the emergency department. Here, Dileep Yavagal, MD, associate professor in the department of neurology & neurosurgery at the University of Miami Miller School of Medicine, discusses the significance of these findings and their potential clinical implications: http://bit.ly/rCBryX.