2023 AHA/ASA Aneurysmal Subarachnoid Hemorrhage Guideline: Summary for Neurologists
The 2023 AHA/ASA Guideline for the Management of Patients with Aneurysmal Subarachnoid Hemorrhage (Hoh et al.) replaces the 2012 guideline. This summary distills the key recommendations using the AHA COR/LOE classification system, organized for rapid clinical reference. aSAH remains a devastating condition with ~30% case fatality and significant long-term morbidity in survivors, particularly in cognitive, behavioral, and quality-of-life domains.
🔹 Bottom Line: 2023 aSAH Guideline — Key Take-Home Messages
- Antifibrinolytics — OUT: Routine tranexamic acid is NOT recommended. The ULTRA (SAH) trial showed no reduction in rebleeding and no improvement in functional outcomes. (COR 3: No Benefit)
- Nimodipine — Still the ONLY proven DCI therapy: Enteral nimodipine 60 mg q6h remains the only Class 1, LOE A recommendation for DCI prevention. Do not interrupt dosing even with hypotension if manageable.
- Hypervolemia is HARMFUL: Prophylactic "triple-H" therapy is out. Target euvolemia only. Prophylactic hemodynamic augmentation increases complications without improving outcomes. (COR 3: Harm)
- Statins — NOT recommended: No benefit in DCI or mortality despite vasospasm reduction. (COR 3: No Benefit)
- Magnesium — NOT recommended: MASH-2 and IMASH showed no outcome benefit. (COR 3: No Benefit)
- Phenytoin is HARMFUL: Associated with worse cognitive outcomes and excess morbidity. Use levetiracetam if seizure prophylaxis needed. (COR 3: Harm)
- Coiling vs Clipping: For good-grade anterior circulation aneurysms equally suitable for both → coiling preferred at 1 year (ISAT). Long-term outcomes equalize (BRAT). Age and location matter.
- Treat early: Secure the aneurysm within 24 hours of onset whenever feasible. (COR 1)
- Recovery focus expanded: Screen all survivors for depression, anxiety, cognitive dysfunction. MoCA preferred over MMSE. Fluoxetine does NOT improve functional recovery.
1. Clinical Presentation & Diagnosis
- Acute severe headache → prompt workup recommended to diagnose/exclude aSAH (COR 1, LOE B-NR)
- Presentation ≥6 hours from onset OR new neurological deficit: Noncontrast CT + LP (if CT negative) should be performed (COR 1, LOE B-NR)
- Presentation <6 hours, no new deficit: High-quality noncontrast CT alone is reasonable to exclude aSAH — sensitivity ~98.7% within 6 hours when read by board-certified neuroradiologist (COR 2a, LOE B-NR)
- Ottawa SAH Rule may be reasonable to identify low-risk patients (COR 2b, LOE B-NR). Criteria: age ≥40, neck pain/stiffness, witnessed LOC, exertional onset, thunderclap headache, limited neck flexion. Rule is 100% sensitive but only ~15% specific.
- Confirmed SAH with negative/inconclusive CTA: DSA is indicated (COR 1, LOE B-NR) — sensitivity of CTA for aneurysms <3 mm is only ~61%
- Confirmed aneurysm: DSA can be useful to determine optimal treatment strategy (COR 2a, LOE B-NR)
🔹 Clinical Relevance: Diagnostic Workflow
- A sentinel headache precedes aSAH in 10–43% of cases — misdiagnosis can be fatal
- CT sensitivity drops significantly after 6 hours → LP is essential for late presenters
- Spectrophotometric analysis for xanthochromia: sensitivity 100%, specificity 95.2%
- Perimesencephalic SAH pattern: CTA alone vs DSA remains debated
2. Clinical Grading & Prognosis
- Clinical grading scales (Hunt-Hess, WFNS) are recommended for severity classification and outcome prediction (COR 1, LOE B-NR)
- Composite scores: VASOGRADE, HAIR (HH grade, Age, IVH, Rebleed), SAHIT, and SAH score combine clinical + radiographic features for refined prognostication
- High-grade aSAH (HH 4–5): Aneurysm treatment is reasonable after careful prognosis discussion — ~40% can achieve favorable outcome at 12 months (COR 2a, LOE B-NR)
- Advanced age: Treatment is reasonable — 42% of patients >65 years achieved functional independence at 6-year follow-up in BRAT post hoc analysis (COR 2a, LOE B-NR)
- Irrecoverable injury: In patients with absent brainstem reflexes, no purposeful responses, global cerebral edema, or large completed infarct — aneurysm treatment is NOT beneficial, after correcting modifiable conditions (COR 3: No Benefit)
🔴 Before Declaring Futility — Correct Modifiable Conditions First
- Seizures / status epilepticus
- Hydrocephalus
- Electrolyte abnormalities (hyponatremia)
- Hypothermia
- High ICP without ventricular enlargement
- Time dimension matters: absent brainstem responses at presentation ≠ absent at 12–24 hours
3. Systems of Care
- Timely transfer to high-volume centers with neurocritical care, comprehensive stroke capabilities, and experienced cerebrovascular surgeons is recommended (COR 1, LOE B-NR)
- Dedicated neurocritical care unit with multidisciplinary team is recommended (COR 1, LOE B-NR)
- Stroke center designation associated with reduced in-hospital mortality
- Factors associated with treatment delay: older age, non-White race, Medicaid status, surgical clipping, admission to low-volume hospitals
4. Preventing Rebleeding (Medical Measures)
- BP management: Frequent monitoring with short-acting agents is recommended. Avoid severe hypotension, hypertension, AND BP variability (COR 1, LOE C-EO)
- No specific SBP target endorsed — previous guidelines suggested <160 or <180 mmHg
- Meta-analysis: rebleeding rates higher with SBP >160 mmHg
- Avoid sudden, profound BP reduction — risk of cerebral ischemia, especially with elevated ICP
- Anticoagulation reversal: Emergency reversal with appropriate agents should be performed (COR 1, LOE C-EO)
- Antifibrinolytics — NOT recommended: Routine use does not improve functional outcome (COR 3: No Benefit, LOE A)
- The ULTRA (SAH) trial: TXA did not significantly reduce rebleeding (10% vs 14%) and showed LOWER rates of excellent outcome (mRS 0–2) in the TXA group
- Good outcome (mRS 0–3): 60% TXA vs 64% control
5. Aneurysm Treatment: Clipping vs Coiling
Timing
- Treat as early as feasible, preferably within 24 hours of onset (COR 1, LOE B-NR)
- Meta-analyses support benefit of treatment <24h vs >24h from ictus
- No data to support emergency (≤6h) or 24/7 nighttime treatment — may create suboptimal conditions
- If presenting during DCI window (days 4–7): do NOT delay beyond 7–10 days
Treatment Goal
- Complete obliteration whenever feasible — incomplete occlusion carries substantially higher rebleeding and retreatment risk (COR 1, LOE B-NR)
- If complete obliteration not feasible: partial treatment to secure rupture site + delayed retreatment in 1–3 months is reasonable (COR 2a, LOE C-EO)
Modality of Treatment
| Clinical Scenario | Recommendation | COR / LOE | Key Evidence |
|---|---|---|---|
| Good-grade anterior circulation, equally suitable for both | Coiling preferred over clipping for 1-year outcome | COR 1 / A | ISAT: RR 0.77 for death/dependency at 1 yr |
| Good-grade anterior circulation, long-term outcome | Both coiling and clipping are reasonable | COR 2a / B-R | BRAT: No significant difference at 3- and 6-year follow-up |
| Posterior circulation, amenable to coiling | Coiling preferred | COR 1 / B-R | RR 0.41 (95% CI 0.19–0.92) for death/dependency |
| Large intraparenchymal hematoma with depressed LOC | Emergency clot evacuation + clipping | COR 1 / B-R | Mortality 27% vs 80% with conservative management |
| Age >70 years | No clear superiority of either modality | COR 2b / B-R | ISAT subgroup: RR 1.15 (CI 0.82–1.61) — no benefit of coiling |
| Age <40 years | Clipping might be preferred (better durability) | COR 2b / C-LD | ISAT: less benefit of coiling in <50 yrs |
| Ruptured wide-neck, not amenable to clipping or primary coiling | Stent-assisted coiling or flow diverters are reasonable | COR 2a / C-LD | — |
| Ruptured fusiform/blister aneurysms | Flow diverters are reasonable | COR 2a / C-LD | Comparable morbidity/mortality to surgical strategies |
| Saccular aneurysms amenable to clipping or primary coiling | Stents/flow diverters should NOT be used | COR 3: Harm / B-NR | Higher risk of hemorrhagic complications, esp. EVD-related hemorrhage |
Intraoperative Management
- Mannitol or hypertonic saline can be effective for ICP/cerebral edema (COR 2a, LOE B-R)
- Prevent intraoperative hyperglycemia and hypoglycemia (COR 2a, LOE B-NR)
- Intraoperative neuromonitoring (EEG, SSEPs, MEPs) may be reasonable (COR 2b, LOE B-NR)
- Adenosine for cardiac standstill may be considered for uncontrolled intraoperative rupture (COR 2b, LOE C-LD)
- Induced mild hypothermia — NOT beneficial in good-grade aSAH (COR 3: No Benefit, LOE B-R) — IHAST: 1000 patients, no improvement in 3-month outcomes
6. Medical Complications Management
Pulmonary
- Standardized ICU care bundle for mechanically ventilated patients >24h (COR 1, LOE B-NR) — includes lung-protective ventilation, early enteral nutrition, systematic extubation approach
- Severe ARDS with refractory hypoxemia: prone positioning and recruitment maneuvers may be reasonable WITH ICP monitoring (COR 2b, LOE B-NR)
Volume & Electrolytes
- Goal-directed euvolemia is reasonable (COR 2a, LOE B-R) — CVP alone is NOT adequate for volume assessment
- Fludrocortisone is reasonable for natriuresis/hyponatremia (COR 2a, LOE B-R) — reduces sodium excretion without significant morbidity beyond hypokalemia
- Hypervolemia is potentially HARMFUL — associated with excess morbidity without reducing DCI (COR 3: Harm, LOE B-R)
Other
- VTE prophylaxis: Pharmacological or mechanical prophylaxis is recommended after aneurysm is secured (COR 1, LOE C-LD)
- Glycemic control: Avoid both hyperglycemia and hypoglycemia (COR 2a, LOE B-NR) — tight control (80–120 mg/dL) reduced infections but did not improve outcomes
- Fever/TTM: Effectiveness of TTM for refractory fever is uncertain (COR 2b, LOE C-LD) — no modality has improved outcomes; TTM can cause shivering, prolonged sedation, longer ventilation
7. Vasospasm & DCI: Monitoring
DCI occurs in ~30% of patients, mostly between days 4–14. Clinical deterioration = focal deficit or GCS drop ≥2 points lasting ≥1 hour, not attributable to other causes.
| Modality | COR / LOE | Key Points |
|---|---|---|
| CTA / CT Perfusion | COR 2a / B-NR | CTA sensitivity 91% for central vasospasm; CTP positive predictive value 0.67 for DCI; can perform on day 3 to baseline risk |
| Transcranial Doppler (TCD) | COR 2a / B-NR | MCA mean velocity ≥120 cm/s + Lindegaard ratio ≥3; sensitivity 90%, NPV 92%; operator-dependent, limited by bone window |
| Continuous EEG (cEEG) | COR 2a / B-NR | Useful in high-grade aSAH; decreasing alpha/delta ratio, late epileptiform abnormalities predict DCI (96.2% sensitivity) |
| Invasive monitoring (PbtO₂, microdialysis) | COR 2b / B-NR | Lactate/pyruvate ratio and glutamate correlate with DCI; regional measurement — placement location matters |
8. Vasospasm & DCI: Treatment
Recommended / Reasonable
- Nimodipine 60 mg enteral q6h — beneficial for preventing DCI and improving functional outcome (COR 1, LOE A)
- Only FDA-approved therapy for neurological improvement after aSAH
- Meta-analysis of 16 trials (n=3361) confirms benefit
- Disruption of dosing correlates with higher DCI incidence — maintain full dosing even with hypotension if manageable
- IV and IA nimodipine: limited data, no recommendation
- Euvolemia can be beneficial (COR 2a, LOE B-NR) — volume depletion associated with 58% DCI rate; one study showed euvolemic protocol reduced DCI from 44% to 8%
- Induced hypertension for symptomatic DCI may be reasonable (COR 2b, LOE B-NR)
- HIMALAIA trial: terminated early for futility — underpowered, no conclusion
- Observational data: ~80% symptomatic improvement with induced hypertension
- Norepinephrine may be preferred over phenylephrine (94% vs 71% neurological improvement)
- Intra-arterial vasodilators may be reasonable for severe vasospasm (COR 2b, LOE B-NR) — multiple agents available; avoid papaverine (neurotoxicity); intermittent preferred over continuous infusion
- Cerebral angioplasty may be reasonable (COR 2b, LOE B-NR) — more durable than vasodilators; vessel rupture carries high mortality but contemporary safety profiles are favorable
NOT Recommended
| Intervention | COR / LOE | Evidence |
|---|---|---|
| Statins | COR 3: No Benefit / A | HDS-SAH + meta-analysis of 6 RCTs: vasospasm reduced but no benefit in DCI or mortality |
| IV Magnesium | COR 3: No Benefit / A | MASH-2 (n=1204): RR 1.03 for poor outcome. IMASH (n=327): OR 1.0. Two meta-analyses confirm no benefit. |
| Prophylactic hemodynamic augmentation | COR 3: Harm / B-R | No neurological benefit; higher complications including congestive heart failure |
🔹 Clinical Relevance: Investigational DCI Therapies in the Trial Database
- Clazosentan (endothelin receptor antagonist): CONSCIOUS-1 showed dose-dependent vasospasm reduction (65% RRR with 15 mg/h) but NO outcome benefit. CONSCIOUS-2 (clipping) and CONSCIOUS-3 (coiling) — neither improved functional outcomes. Vasospasm ≠ DCI.
- Tirofiban (GPIIb/IIIa inhibitor): ISPASM — small pilot (n=30) showed dramatic DCI reduction (6% vs 33%, NNT 3.7) but underpowered for outcomes.
- Nicardipine implants: NICARDIPINE Implant trial — vasospasm reduced (20% vs 58%), rescue therapy reduced (10% vs 58%) in clipped patients. NicaPlant Phase IIA showed safety and therapeutic CSF levels.
- Albumin: ALISAH — dose-escalation pilot; 1.25 g/kg/day appeared optimal but higher doses caused pulmonary edema.
- Other investigational: TRIVASOSTIM (trigeminal nerve stimulation — negative), SAS (sulforaphane — negative), SAHRANG (galantamine — QOL signal), Prostacyclin in SAH (trend only), Dapsone in SAH (promising small pilot), Dantrolene SAH Trial, Tiopronin Trial (safety established).
9. Hydrocephalus
- Acute symptomatic hydrocephalus: Urgent CSF diversion (EVD and/or lumbar drainage) should be performed (COR 1, LOE B-NR)
- Acute hydrocephalus occurs in 15–87% of patients
- EARLYDRAIN: lumbar drainage reduces DCI prevalence and improves early outcomes
- Centers without aneurysm treatment capability: stabilize, place EVD if needed, then transfer
- EVD bundled protocol is recommended — addresses insertion, management, education, monitoring. Reduces infection from 6–37% (pre-protocol) to 0–9% (post-protocol) (COR 1, LOE B-NR)
- Chronic symptomatic hydrocephalus: Permanent CSF diversion (VP shunt) is recommended (COR 1, LOE B-NR) — shunt dependency in 9–48% of aSAH patients
- Lamina terminalis fenestration — NOT beneficial for reducing shunt dependency: RR 0.88 (CI 0.62–1.24) (COR 3: No Benefit, LOE C-LD)
10. Seizure Management
Patients WITHOUT Seizures at Presentation
- cEEG monitoring is reasonable in patients with fluctuating exam, depressed mental state, MCA aneurysm, high-grade SAH, ICH, hydrocephalus, or cortical infarction (COR 2a, LOE B-NR)
- Seizure prophylaxis may be reasonable ONLY with high-risk features: MCA aneurysm, high-grade SAH (HH ≥3), ICH, hydrocephalus, cortical infarction (COR 2b, LOE B-NR)
- Without high-risk features: Prophylactic antiseizure medication is NOT beneficial (COR 3: No Benefit, LOE B-R)
- Phenytoin is HARMFUL — associated with worse cognitive outcomes, excess morbidity and mortality (COR 3: Harm, LOE B-NR). Use levetiracetam — randomized study showed same efficacy with fewer adverse effects.
Patients WITH Seizures at Presentation
- Antiseizure medication for ≤7 days is reasonable to reduce perioperative complications (COR 2a, LOE B-NR)
- Treatment beyond 7 days is NOT effective for reducing future seizure risk in patients without prior epilepsy (COR 3: No Benefit, LOE B-NR)
🔹 Clinical Relevance: Seizure Definitions in aSAH
- Onset seizures: At time of hemorrhage — predict poor outcome
- Early seizures: First week — may relate to acute injury
- Late seizures: After 1 week or postoperative — relate to treatment modality/infarction; coiling associated with lower late seizure rates than clipping
- American Clinical Neurophysiology Society definition: epileptiform discharges averaging >2.5 Hz for ≥10 seconds
11. Acute Recovery
- Validated screening for physical, cognitive, behavioral, and QOL deficits is recommended before discharge (COR 1, LOE B-NR)
- Depression and anxiety screening in the postacute period using validated tools (Hospital Anxiety Depression Scale, PHQ-9, GAD-7) (COR 1, LOE B-NR)
- Depression treatment: Psychotherapy and pharmacotherapy recommended — SSRIs reduce poststroke depression (RR 0.75) (COR 1, LOE B-NR)
- Cognitive dysfunction screening: Use validated tools; cognitive impairment in 40–70% of survivors, even with good functional outcomes (COR 1, LOE B-NR)
- Multidisciplinary team approach recommended — reduces LOS (mean hospital LOS from 21.6 to 14.1 days in one study) (COR 1, LOE B-NR)
- Early rehabilitation after aneurysm is secured is reasonable (COR 2a, LOE B-NR) — each mobilization step in first 4 days associated with 30% reduction in severe vasospasm risk
- Neurostimulants (amantadine, modafinil) may be reasonable for coma patients after reversible causes treated (COR 2b, LOE C-LD)
- Fluoxetine — NOT effective for enhancing poststroke functional recovery; increases fractures, osteoporosis, and seizures (COR 3: No Benefit, LOE A)
12. Long-Term Recovery
- Screen for depression, anxiety, and sexual dysfunction — recommended for long-term outcome improvement (COR 1, LOE B-NR)
- MoCA is preferred over MMSE for cognitive assessment — higher sensitivity for detecting impairment after aSAH (COR 2a, LOE B-NR)
- Counsel on dementia risk: HR 2.72 (95% CI 2.45–3.06) for dementia in aSAH survivors; median age at diagnosis 74 years (vs 79 for ICH, 81 for ischemic stroke) (COR 2b, LOE B-NR)
- Cognitive difficulties persist in ~50% at 1 year, even with good mRS
13. Follow-Up Imaging & Recurrence Prevention
- Perioperative cerebrovascular imaging is recommended to identify remnants/recurrence (COR 1, LOE B-NR)
- ISAT: 30-day rebleed risk — 1.9% coiled vs 0.6% clipped; incomplete occlusion is the primary risk factor
- Long-term follow-up imaging is recommended for recurrence/regrowth of treated aneurysm and de novo aneurysm detection (COR 1, LOE B-NR)
- Coiled aneurysms: higher rate of incomplete occlusion and recurrence
- Clipped with residual: regrowth 2.1%/year vs 0.26%/year without residual
- De novo aneurysm risk factors: younger age, family history, multiple aneurysms
14. Trial Comparison Table
| Trial | Year | N | Intervention | Key Finding | Outcome Impact |
|---|---|---|---|---|---|
| ISAT | 2002 | 2143 | Coiling vs Clipping | 7.4% ARR in death/dependency at 1 yr with coiling | ✅ Positive (1-yr); equalized long-term |
| BRAT | 2012 | 408 | Coiling vs Clipping | No difference at 3- and 6-year follow-up | ➖ Neutral long-term |
| IHAST | 2005 | 1000 | Intraoperative hypothermia vs normothermia | No improvement in 3-month outcomes | ❌ Negative |
| MASH-2 | 2012 | 1204 | IV MgSO₄ vs placebo | Poor outcome: 26.2% vs 25.3% (RR 1.03) | ❌ Negative |
| IMASH | 2010 | 327 | IV MgSO₄ + nimodipine vs placebo + nimodipine | GOSE 5–8: 64% vs 63% (OR 1.0) | ❌ Negative |
| HDS-SAH | — | — | High-dose simvastatin in aSAH | No benefit in DCI or mortality | ❌ Negative |
| CONSCIOUS-1 | 2008 | 413 | Clazosentan (1/5/15 mg/h) vs placebo | 65% RRR in vasospasm with 15 mg/h; no outcome benefit | ⚠️ Vasospasm reduced, no functional benefit |
| CONSCIOUS-2 | 2011 | 1147 | Clazosentan 5 mg/h vs placebo (clipped) | Primary endpoint: RRR 17% (p=0.10) | ❌ Negative |
| CONSCIOUS-3 | 2012 | 577 | Clazosentan 5/15 mg/h vs placebo (coiled) | 15 mg/h: primary endpoint reduced (15% vs 27%); no functional improvement | ⚠️ Halted early; vasospasm reduced, no outcome benefit |
| ISPASM | 2021 | 30 | Tirofiban vs placebo | DCI: 6% vs 33% (p=0.04); NNT 3.7 | ⚠️ Promising pilot — needs phase 3 |
| MASH | 2006 | 161 | Aspirin 100 mg vs placebo | No reduction in DIND (HR 1.83) | ❌ Negative |
| NICARDIPINE Implant | 2024 | 41 | Nicardipine implants vs standard care (clipped) | Vasospasm: 20% vs 58% (p=0.02); rescue: 10% vs 58% | ⚠️ Promising — needs larger trial |
| NicaPlant | 2023 | 14 | NicaPlant® implants (Phase IIA) | Safe; therapeutic CSF levels without systemic absorption | ⚠️ Safety established |
| ALISAH | 2012 | 47 | 25% albumin (dose-escalation) | 1.25 g/kg/day best balance; higher doses → pulmonary edema | ⚠️ Pilot — dose identified |
| GDHT After SAH | 2020 | 108 | Goal-directed therapy vs standard care | DCI: 13% vs 32% (p=0.021); GOS 5: 66% vs 44% | ✅ Positive (single-center) |
| EARLYDRAIN | — | — | Lumbar drainage for aSAH | Reduced DCI prevalence, improved early outcomes | ✅ Positive |
| CIAN | 2022 | 17 | Continuous IA nimodipine for refractory vasospasm | GOS 4–5 at 1 year: 76% | ⚠️ Feasibility/safety established |
| TRIVASOSTIM | 2023 | 60 | Trigeminal nerve stimulation vs sham | No difference in vasospasm-related infarction | ❌ Negative |
| SAS | 2024 | 105 | SFX-01 (sulforaphane) vs placebo | No reduction in vasospasm or functional improvement | ❌ Negative |
| SAHRANG | — | — | Galantamine in aSAH | Signal for QOL improvement (EQ5D VAS) at days 30–60 | ⚠️ Pilot — QOL signal |
| VANQUISH | — | — | Vagus nerve stimulation for SAH headache | Reduced post-stimulation headache intensity; no opioid reduction | ⚠️ Pilot |
| Prostacyclin in SAH | 2015 | 90 | IV prostacyclin vs placebo | DIND: 21% vs 38% (NS); no CBF difference | ❌ Negative |
| Dapsone in SAH | — | ~50 | Dapsone in aSAH | DCI: 26.9% vs 63.6% (p=0.011); infarction: 19.2% vs 63.6% | ⚠️ Promising small study |
| SAHARA | — | — | — | — | — |
| Tiopronin Trial | — | 9 | Tiopronin dose-escalation | Safe at 3 g/day; no vasospasm-related infarction | ⚠️ Safety pilot |
References
- Hoh BL, Ko NU, Amin-Hanjani S, et al. 2023 Guideline for the Management of Patients With Aneurysmal Subarachnoid Hemorrhage: A Guideline From the American Heart Association/American Stroke Association. Stroke. 2023;54:e314–e370. doi: 10.1161/STR.0000000000000436
- Molyneux AJ, Kerr RS, Birks J, et al. Risk of recurrent subarachnoid haemorrhage, death, or dependence and standardised mortality ratios after clipping or coiling of an intracranial aneurysm in the International Subarachnoid Aneurysm Trial (ISAT). Lancet Neurol. 2009;8:427–433.
- Spetzler RF, McDougall CG, Zabramski JM, et al. The Barrow Ruptured Aneurysm Trial: 6-year results. J Neurosurg. 2015;123:609–617.
- Todd MM, Hindman BJ, Clarke WR, Torner JC; IHAST Investigators. Mild intraoperative hypothermia during surgery for intracranial aneurysm. N Engl J Med. 2005;352:135–145.
- Dorhout Mees SM, Algra A, Vandertop WP, et al. Magnesium for aneurysmal subarachnoid haemorrhage (MASH-2): a randomised placebo-controlled trial. Lancet. 2012;380:44–49.
- Wong GK, Poon WS, Chan MT, et al; IMASH Investigators. Intravenous magnesium sulphate for aneurysmal subarachnoid hemorrhage (IMASH). Stroke. 2010;41:921–926.
- Macdonald RL, Higashida RT, Keller E, et al. Clazosentan, an endothelin receptor antagonist, in patients with aneurysmal subarachnoid haemorrhage undergoing surgical clipping: a randomised, double-blind, placebo-controlled phase 3 trial (CONSCIOUS-2). Lancet Neurol. 2011;10:618–625.
- Macdonald RL, Kassell NF, Mayer S, et al. Clazosentan to overcome neurological ischemia and infarction occurring after subarachnoid hemorrhage (CONSCIOUS-1). Stroke. 2008;39:3015–3021.
- Post R, Germans MMA, Tjerkstra MA, et al. Ultra-early tranexamic acid after subarachnoid haemorrhage (ULTRA): a randomised controlled trial. Lancet. 2021;397:112–118.