Subarachnoid Hemorrhage- a clinical pathway
Subarachnoid Hemorrhage – a clinical pathway
PGY3, Emergency Medicine
Acute headaches are commonly seen in the ED, with 1-2% of all ED visits primarily for this cause. While many of the headaches we see have benign etiologies, a concerning potential cause of an acute headache is subarachnoid hemorrhage (SAH). Aneurysmal SAHs occurs at a rate of 6-8 events per 100,000 person-years, and prognosis is poor for patients presenting with this condition: one month mortality is 40-45%, and almost one-third of patients who survive will have severe neurologic deficits. Patients may present with a “sentinel headache” before a larger, more devastating bleed occurs. In this situation, the risk of re-bleeding is 1.5% per day or 15-20% in the first 2 weeks. If the aneurysm is discovered at the time of the sentinel bleed, however, neurosurgical intervention can be curative. Because of this, there is a low-threshold for investigating this potential cause of headache despite the relative infrequency of aneurysmal SAH.
Traditionally, the workup for SAH consists of obtaining a non-contrast head CT, followed by a lumbar puncture (LP) if the CT is negative. If either of these tests are positive, a cerebral angiogram is performed to evaluate for aneurysm; if both are negative, SAH has been ruled out. Current ACEP clinical guidelines reflect this practice with B level recommendations. These guidelines, however, do not reflect the results of several new studies that are starting to change the way many clinicians now evaluate SAH. These studies help us address questions and concerns not addressed by ACEPs guidelines. In this post we will look at several studies which contributed to the development of Temple’s subarachnoid hemorrhage clinical guideline. Specifically, we will evaluate a clinical rule that has been developed to help determine a subset of patient who do not need workup for SAH, explore the sensitivity and specificity of detecting SAH with newer CT imaging[6-8], and address the large number of uninterpretable or inconclusive LP findings.
The first study, published in 2013, is “Clinical decision rules to rule out subarachnoid hemorrhage for acute headache” by Perry et al. In this study, 2131 adult patients were evaluated for SAH. Each of these patients presented with acute, severe, non-traumatic headaches that reached maximal intensity within 1 hour. Exclusion criteria included new neurological deficits, personal history of aneurysm, SAH, or brain tumor, or if the patient reported having at least three similar headaches over a time span of at least 6 months (suggesting a benign etiology). The presence or absence of potential risk factors were collected for each patient, and workup proceeded as per standard of care. The study identified the following six risk factors for SAH.
Age ≥40 years old
Neck pain or stiffness
Onset of headache during exertion
“Thunderclap” headache (maximally intense pain at headache onset)
Limited neck flexion on exam
These characteristics became known as the Ottawa SAH Rule, which when applied to the derivation population had 100% sensitivity for detecting SAH, suggesting that an appropriate patient (meeting inclusion criteria and not meeting exclusion criteria) without any of the six risk factors is unlikely to have a SAH and does not need further investigation for this headache etiology. While this decision rule still needs to be evaluated in a separate implementation study to confirm its validity outside of the derivation population, it offers a standardized, evidence-based algorithm for determining which patients can be reasonably classified as low-risk.
As CT technology has improved over the past several years to produce better imaging, it has been proposed that LP may no longer be necessary to evaluate for SAH in some cohorts. Three studies evaluated the sensitivity of CT for SAH when obtained early in the course of a headache to determine if LP can be eliminated from the algorithm for these patients. The first was a prospective cohort study also by Perry et al published in 2011 that looked at 3132 patients evaluated for SAH, 240 (7.7%) of which were diagnosed with SAH. For the purposes of the study, SAH was defined as either subarachnoid blood seen on head CT, CSF with more than 5×106 RBCs in the final CSF collection tube or visual xanthochromia, or cerebral aneurysm identified on angiography. This study found that head CT performed <6hrs after headache onset had a 100% (97-100% CI) sensitivity for SAH in this study population, while sensitivity was only 85.7% (78.3-90.9% CI) for CTs performed >6 hours after headache onset. While not every patient with a negative head CT subsequently underwent LP, these patients were followed clinically for 3 months and during this time period there were no deaths or negative outcomes for patients who had negative head CTs within the 6 hour window.
The next study was a retrospective single center study published by Backes et al in 2012 in which 250 patients were evaluated for SAH. In this study, all 250 patients had CTs read by neuroradiologists and each patient with a negative head CT underwent LP. Backes obtained results similar to the previous study: of 69 patients with a negative CT performed within 6 hours of headache onset, only one was found to have a SAH due to a vascular abnormality, giving CT scan within 6 hours of headache onset a 98.5% (92.1-100% CI) sensitivity. In this study, CT outside the 6-hour window had a sensitivity of 90% (76.3-97.2% CI).
One criticism of the previous two studies is that all of the CTs were read by neuroradiologists, which led to questions about the generalizability of these results as many hospitals do not have neuroradiology coverage. To address this issue, a third study by Blok et al published in 2015 evaluated this rule in a different patient population. They performed a multicenter retrospective study that took place in non-academic settings in which 760 patients were evaluated for SAH by both CT within 6 hours of headache onset read by “staff radiologists” as well as LP. Fifty-two of the patients, or 6.8%, had positive LPs after a normal head CT as defined by the presence of bilirubin in the CSF, which reflects only a 93.2% sensitivity for SAH. When CT results were overread by neuroradiologists, only one of these 52 (2%) was determined to be positive for a perimesencephalic nonaneurysmal SAH, suggesting that there is not a large miss-rate of subarachnoid blood among staff radiologists compared to neuroradiologists. Additionally, 29 of these patients underwent cerebral angiogram, and only 8 were found to have an aneurysm, 3 of which were previously known and had already been coiled, and the remaining 5 of which were determined to be unruptured (incidental). This showed that the negative CT/positive LP combination was most often due to either previously known aneurysms (patients with which were excluded in the original study by Perry) or nonaneurysmal SAH, which does not require intervention and has a much better prognosis than aneurysmal SAH. Accordingly, there were no deaths or negative outcomes in patients with negative CTs performed within 6 hours. This study helped demonstrate the generalizability of Perry’s results. Overall, this group of studies shows that while generally benign causes of SAH may be missed by skipping the LP after CT in patients who receive imaging within 6 hours of symptom onset, there is only a very small chance of missing a cause that portends bad outcomes or requires neurosurgical intervention.
Finally, one frustration with LP results is they can often be equivocal: 10-30% of LPs are traumatic and it can be difficult to distinguish between a positive test and a traumatic tap. To address this issue, Perry et al published a paper in 2015 to help classify “inconclusive” LP results as either positive or simply the result of a traumatic tap. The study was a prospective observational study of 1739 patients who underwent LPs to evaluate for SAH. 36.9% of the LPs were positive for RBCs in the final tube or had visual xanthochromia, however, only 15 (0.9%) of these patients were found to have SAH due to aneurysm. When comparing LP results, all patients with aneurysmal SAH had either visual xanthochromia or ≥2000×10^6 RBCs in their final CSF tube, suggesting that patients with LP results without either of these qualities are at low risk for aneurysmal SAH.
These studies, when taken as a whole, help to guide the emergency physician in the workup of the patient with an acute headache concerning for subarachnoid hemorrhage. The Ottawa SAH rule provides guidance as to which patients can clinically be ruled out without further investigation and the studies by Perry, Backes, and Blok provide evidence that CT alone is very sensitive for SAH if performed within 6 hours. In the cases we judge LP is necessary, the final study by Perry helps us interpret the LP results.
Sayer, D et al. An Observational Study of 2,248 Patients Presenting With Headache, Suggestive of Subarachnoid Hemorrhage, Who Received Lumbar Punctures Following Normal Computed Tomography of the Head. Journal of Academic Emergency Medicine. 2015 Nov; 22(11):1267-73.
Beck, J, Raabe, A, Szelenyi, A et al. Sentinel headache and the risk of rebleeding after aneurysmal subarachnoid hemorrhage. Stroke. 2006;37:2733-2737.
Perry JJ, Alyahya B, Sivilotti ML, et al. Differentiation between traumatic tap and aneurysmal subarachnoid hemorrhage: prospective cohort study. BMJ. 2015;350:h568.
Edlow, JA, Panagos, PD, Godwin, SA et al. Clinical Policy: Critical Issues in the Evaluation and Management of Adult Patients Presenting to the Emergency Department With Acute Headache. Annals of Emergency Medicine. 2008 Oct; 52(4): 407-418.
Perry JJ, Stiell IG, Sivilotti ML, et al. Clinical decision rules to rule out subarachnoid hemorrhage for acute headache. JAMA. 2013;310:1248-1255.
Perry JJ, Stiell IG, Sivilotti ML, et al. Sensitivity of computed tomography performed within six hours of onset of headache for diagnosis of subarachnoid haemorrhage: prospective cohort study. BMJ. 2011;343:d4277.
Backes D et al. Time-Dependent Test Characteristics of Head Computed Tomography in Patients Suspected of Nontraumatic Subarachnoid Hemorrhage. Stroke 2012; 43(8): 2115 – 9.
Blok KM et al. CT Within 6 Hours of Headache Onset to Rule Out Subarachnoid Hemorrhage in NonAcademic Hospitals. Neurology 2015; 84(19): 1927 – 32.