Text Size:    +   -

Early Management of Acute Subarachnoid Hemorrhage

5th Edition
June 2015

The Canadian Stroke Best Practice Recommendations for Hyperacute Stroke Care, 5th Edition (2015) is published in the International Journal of Stroke (IJS) and available freely online. To access the specific recommendations for Early Management of Acute Subarachnoid Hemorrhage and all other sections of the Hyperacute Stroke Care recommendations, please click on this URL which will take you to the recommendations online in the IJS: http://onlinelibrary.wiley.com/doi/10.1111/ijs.12551/full

For the French version of these recommendations, open the appendix at this link :  http://onlinelibrary.wiley.com/store/10.1111/ijs.12551/asset/supinfo/ijs12551-sup-0001-si.zip?v=1&s=cdf3d494242426450aaa522f104ace17857f037a

All other supporting information, including performance measures, implementation resources, evidence summaries and references, remain available through www.strokebestpractices.ca, and not through the IJS.  Please click on the appropriate sections on our website below for this additional content.


Subarachnoid hemorrhage is a catastrophic neurosurgical emergency that is prevalent in approximately seven percent of adults with stroke, and also in children, and accounts for prolonged hospital lengths of stay. Recent mortality rates in Canada for patients with subarachnoid hemorrhage are just over 40 percent within 30 days of the event. Over the past decade, several advances have been made in early treatment of subarachnoid hemorrhage, including endovascular techniques. Prompt recognition and access to expert medical professionals may reduce mortality and morbidity and improve long-term outcomes.

System Implications

  • The stroke system design should include processes to rapidly access neurosurgical services which would treat a high volume of cases and have onsite expertise to treat acute hemorrhage (10 clips, 20 coils, and 35 SAH’s per year, based on the Joint Commission criteria).
  • Awareness and education for physicians and nursing staff to recognize subarachnoid hemorrhage as a medical emergency and usually has different distinct presentation, even in those patients with other headache disorders.
  • Protocols for acute SAH management, including rapid access to neurosurgical specialists for hemorrhagic patient management, including rapid referral process if neurosurgical services not available within the initial treating hospital.
  • Protocols for discussing organ donation for patients with SAH who are potential organ donors should be in place and staff should be trained with appropriate communication techniques for addressing this sensitive topic. Access to organ donor coordinators where available.

Performance Measures

  1. Risk-adjusted mortality rates for subarachnoid hemorrhage in-hospital, 30-day and one year (core).
  2. Percentage of subarachnoid hemorrhage patients who receive a consult to a neurosurgeon within 6 and within 12 hours of hospital arrival.
  3. Percentage of subarachnoid hemorrhage patients who receive a CT scan or MRI within 6 and 12 hours of hospital arrival.
  4. Rebleeding rate for subarachnoid hemorrhage patients (stratified by whether patient under-went surgical or endovascular intervention) within 7 days and 30 days of hospital presentation.

Measurement Notes

Risk adjustment should include age, gender, and initial stroke severity scores, as well as co-morbidities.

Implementation Resources and Knowledge Transfer Tools
Summary of the Evidence, Evidence Tables and References

Evidence Table 6 Acute Subarachnoid Hemorrhage

Subarachnoid hemorrhage (SAH) is a medical emergency with the potentially devastating effects of early mortality or significant morbidity. Early identification is critical when patients present with suspected SAH. With improvements in technology, modern CT scanners have become a more accurate diagnostic tool for SAH. Using a lumbar puncture as the diagnostic standard. Conrtnum et al. (2010) reported that CT alone was able to correctly identify 295/296 patients with confirmed SAH within the first week of symptoms onset. A lumbar puncture was required to confirm the diagnosis in a single patient on day 6. Overall, CT scanning had a sensitivity of 99.7 % (95% CI: 98.1 to 99.99%) and specificity of 100% (95% CI 98.2% to 100%). In another study examining the accuracy of CT performed early after suspected SAH, data from 11 tertiary-care emergency departments across Canada on 3,132 patients who presented with non-traumatic acute headache, were included (Perry et al. 2011). In this study, 240 patients (7.7%) were diagnosed with SAH. The overall sensitivity and specificity of CT imaging were 92.9% (95% CI 89.0 to 95.5%) and 100% (95% CI 99.9% to 100%), respectively. The sensitivity increased to 100% (95% CI: 97 to 100%) for the 953 patients who were scanned within 6 hours of headache onset. The ability of clinical decision-making rules to detect SAH has also been investigated. Perry et al. (2013) reported on the development of the Ottawa SAH Rule. Among 2131 patients, presenting with non-traumatic acute headache peaking within 1 hour, 6.2% were eventually diagnoses with SAH. The best model to identify SAH included age ≥40 years, the presence of neck pain or stiffness, witnessed LOC, onset during exertion, limited neck flexion and thunderclap headache yielding a sensitivity of 100% (95% CI 97.2%-100%), specificity of 15.3% (95% CI 13.8%-16.9%).

A number of pharmacological treatments have been evaluated in the management of SAH. A Cochrane Review (Dorhout Mees et al. 2007) including the results from 16 RCTs trials (3361 patients), evaluated the use of calcium antagonists for blood pressure control following aneurysmal SAH. Three of the studies examined magnesium sulphate in addition to nimodipine. Overall, calcium antagonists reduced the risk of poor outcome (RR=0.81, 95 %CI 0.72 to 0.92; NNT=19). Oral nimodipine alone was also associated with a significant reduction in the risk of a poor outcome (RR= 0.67, 95% CI 0.55 to 0.81), while the use of other calcium antagonists or intravenous administration of nimodipine were not. The addition of magnesium to standard treatment with nimodipine, was associated with the borderline reduction in the risk of poor outcome (RR= 0.75, 95%CI 0.57 to 1.00) and significantly reduced the clinical signs of secondary ischemia (RR=0.66, 95% CI 0.45 to 0.96). The use of the neuroprotective agent, tirilazad in addition to standard treatment was not found to significantly reduce the risk of death or poor outcome in patients with aneurysmal SAH (Zhang et al. 2010).

Endothelin receptor antagonist (ETRA) to prevent endothelin-mediated cerebral vasospasm after SAH was investigated in a series of the three related trials. In CONSCIOUS-1 (Macdonald et al. 2008), a dose finding study, patients were randomized to receive 1, 5 or 15 mg/hour of clazosentan or placebo for 14 days after aneurysmal SAH. Treatment with any dose was associated with a significant reduction in the risk of moderate or severe vasospasm, but did not reduce all-cause mortality or vasospasm-related morbidity. In CONSCIOUS-2 (Macdonald et al. 2011), patients were randomized to receive 5 mg/h of clazosentan or placebo for 14 days. Treatment with clazosentan was not associated with a significant reduction in all-cause mortality or vasospasm-related morbidity at 6 weeks (the primary outcome) or the risk of a poor outcome at 12 weeks. Finally, CONSCIOUS-3 randomized patients to receive 5 or 15 mg/h clazosentan or placebo within 56 hours of surgery, for up to 14 days (Macdonald et al. 2012). The trial was stopped prematurely due to non-significant findings from CONSCIOUS- 2 trial. The risk of the primary outcome, which was the same as the CONSCIOUS-2 trial, was reduced significantly for patients in the 15 mg group (OR=0.47, 95% CI 0.28-0.82, p=0.007), but not for patients in the 5 mg group (OR=0.79, 95% CI 0.48-1.29, p=0.34); however, the risk of a poor outcome was not reduced significantly for patients in either of the clazosentan groups compared with placebo. A meta-analysis that included the results from the 3 CONSCIOUS trials plus two other placebo-controlled RCTs reported that treatment with ETRAs was not associated with the reduced risk of a poor outcome at 3 months (RR=1.06, 95% CI 0.93-1.22) (Vergouwen et al. 2012). The occurrences of angiographic vasospasm, new cerebral infarction, vasospasm related infarction and case fatality was similar between groups. The use of statin medications as a potential agent to prevent vasospasm and improve clinical outcome has been investigated recently. The potential mechanisms of action were thought to be related to enhanced blood flow, anti-inflammation and up regulation of the endothelial nitric oxide synthase pathway. In the STASH trial, Kirkpatrick et al. (2014) randomized 803 patients to receive 40 mg simvastatin or placebo for 3 weeks. At hospital discharge there was no significant difference in the proportion of patients who experienced a good outcome (mRS 0-2) 59% (simvastatin) vs. 62% (placebo), adj OR=1.08, 95%CI 0.81-1.43, p=0.61. At 6 months, there was no significant difference in the proportion of patients who experienced a good outcome (72% vs. 72%; adj OR=0.97, 95% CI 0.75-1.25, p=0.81), or death (10% vs. 9%, p-0.59). The use of prophylactic transluminal balloon angioplasty (TBA) to prevent delayed ischemic neurological deficit (DIND) was investigated in 170 patients with Fisher grade III aneurysmal SAH (Zwienenberg-Lee et al. 2008). Patients were randomized to prophylactic TBA or no prophylactic treatment within 96 hours of aneurysm rupture. The trial was halted 3 times due to treatment-related complications. There was no difference in the number of patients who developed a DIND between groups (23.5% vs. 31.8%) or who experienced a favourable outcome at 3 months (49.4% vs. 43.5%). A high rate of vessel perforation was observed during the trial, resulting in three treatment-related deaths. The treatment protocol was subsequently revised to exclude angioplasty of the bilateral A1 and P1 segments where the majority of vessel perforations occurred.

The prophylactic use of anticonvulsant agents has not been well studied. In fact, a Cochrane review that sought RCTs that examined the use of anti-epileptic drugs following SAH, failed to identify any relevant studies. In one of the few studies published on the subject, a historically controlled trial, including 3,552 patients, the results suggested that outcome was worse in those who received prophylactic anticonvulsants (Rosengart et al. 2007). Patients treated with the anticonvulsants tirilazad had an increased risk for poor outcome (adjusted OR=1.56, 95% CI: 1.16 to 2.10; p = 0.003) based on the Glasgow Outcome Scale; and an increased risk for cerebral vasospasm, neurological deterioration, cerebral infarction and for elevated temperature during hospitalization.

Endovascular treatment is used with increasing frequency as alternative to neurosurgical clipping for SAH. The safety and efficacy of the procedure have been examined in several multi-site RCTs. The International Subarachnoid Aneurysm Trial (ISAT) included 2,143 patients from 42 institutions with definitive SAH occurring within the previous 28 days (Molyneux et al. 2002). Patients were randomly assigned to receive neurosurgical clipping or endovascular treatment. Clinical outcomes were assessed at two months and at one year with interim ascertainment of rebleeds and death. Although recruitment was stopped following planned interim analysis, follow-up of patients enrolled continues. At one year, 23.5% (250/1063) patients allocated to endovascular treatment were dead or dependent, compared with 30.9% (326/1055) patients allocated to neurosurgery, representing an absolute risk reduction of 7.4% (95% CI 3.6 to 11.2, p=0.0001), the equivalent of 74 patients avoiding death or dependency at one year for every 1,000 patients treated. The early survival advantage was maintained. At 5 years, fewer patients in the endovascular therapy group were dead (11% vs. 14%, RR=0.75, 95% CI 0.58-0.97, p=0.03) (Molyneux et al. 2009). The 10-year findings reported among 1256 patients recruited from UK centres, re-enforced the finding that endovascular treatment is associated with a reduced risk of mortality; however, there was no benefit in achieving independence and there was a slightly increased risk of a rebleed from the target aneurysm > 1 year following the procedure (Molyneux et al. 2014). The ISAT has been criticized for its selection bias and threats to external validity, since only 22% of patients screened were recruited. The Barrow Ruptured Aneurysm Trial (BRAT) (McDougall et al. 2012) was designed to be more inclusive and included 65% of patients screened. In this trial there were many protocol violations. Sixty-five patients who had been randomized to receive treatment with coil, in fact received clipping. More patients in the surgical group had a poor outcome (33.7% vs. 23.2%, RR=1.68, 95 CI% 1.08-2.61, p=0.02). Of the 245 patients who actually underwent surgery, more patients had a poor outcome compared with 113 patients who actually underwent coil embolization (33.9% vs. 20.4%, OR= 2.01, 95% CI 1.20–3.46, p = 0.01). Three years following SAH (Spetzler et al. 2013), there was no significant difference between groups in the risk of a poor outcome (35.8% clipping vs. 30% coil, adj OR=1.31, 95% CI 0.82-2.10, p=0.25).

A recent systematic review and meta-analysis examining clipping versus coiling for SAH included the results from 27 studies, of which four were RCTs and the remainder, observational (Li et al. 2013). Overall, treatment with coil was associated with an increased change of better outcome (OR=1.25, 95% CI 1.12-1.40, p<0.0001). When only the results from the RCTs were included, coiling was associated with better odds of avoiding an unfavorable outcome (mRS 3-6) (OR= 1.48; 95% CI; 1.24 to 1.76), while the benefit was lost when only the results from observational studies were included (OR= 1.11; 95% CI: 0.96 to 1.28). In subgroup analysis, the odds of a good outcome were higher in patients with good perioperative prognosis who underwent coiling (OR= 1.51; 95% CI: 1.24 to 1.84) compared with patients with poor initial prognosis (OR= 0.88; 95% CI: 0.56 to 1.38). Overall, there was no difference in 1-year mortality between groups (OR=1.07, 95% CI 0.88-1.30, p=0.51). Rebleeding rate was lower in patients who had received clipping. (OR=0.43, 95% CI 0.28-0.67, p=0.001). Vasospasm was more common after clipping, whereas ischemic infarct, shunt-dependent hydrocephalus and procedural complication rates did not differ significantly between techniques.