Subarachnoid Haemorrhage (SAH): Introduction

Patients with headache account for approximately 1% of all emergency department (ED) visits, and of these 1–4% have been demonstrated to have subarachnoid haemorrhage. Early accurate diagnosis of aneurismal subarachnoid haemorrhage is imperative, as early occlusion of the aneurysm has been shown to reduce early complications of re-bleeding and vasospasm and improve outcome.

Pathology and Epidemiology

Subarachnoid Haemorrhage (SAH) is the presence of extravasated blood within the subarachnoid space. The incidence is 5–7 per 100 000 patient-years, but is significantly higher (around 20 per 100 000) in Japan and Finland, for reasons that are unclear. Although incidence increases with age, about half of those affected are under 55, the condition being most common in the 40–60 age group. Excluding head trauma, which remains the most common cause, non-traumatic or spontaneous SAH results from rupture of a cerebral aneurysm in approximately 85% of cases, non-aneurysmal perimesencephalic haemorrhage in 10%, and the remaining 5% from other rare causes including rupture of mycotic aneurysms, intracranial arterial dissection, aterio-venous malformations, vasculities, central venous thrombosis, bleeding diatheses, tumors and drugs such as cocaine, amphetamines and anticoagulants.

  1. Aneurysms

Intracranial aneurysms are not congenital. Rather, they develop during the course of life. An estimate of the frequency for an adult without risk factors is 2.3%, with the proportion increasing with age. Most aneurysms will never rupture, but the risk increases with size. Paradoxically, as the vast majority of aneurysms are small, most aneurysms that rupture are small. An aneurysm of the posterior circulation is more likely to rupture than one of comparable size in the anterior circulation.

Risk factors can be considered as those that are modifiable and those that are not. Modifiable risk factors include cigarette smoking, hypertension, cocaine use and excessive alcohol intake. Non-modifiable factors include a family history of first-degree relatives with SAH, heritable connective tissue disorders (particularly polycystic kidney disease and neruofibromatosis), sickle cell disease and α1–antitrypsin deficiency.

  1. Non-Aneurysmal Perimesencephalic Haemorrhage

This type of SAH is defined by the characteristic distribution of blood in the cisterns around the midbrain in combination with normal angiographic studies. It usually carries a relatively benign prognosis. A small proportion of patients with this distribution of blood may have a ruptured aneurysm of a vertebral or basilar artery.

Clinical Features

  1. History

The history is critical to the diagnosis of SAH:

  • Headache is the principal presenting symptom, being present in up to 95% of patients with SAH and being the solitary symptom in up to 40% of patients. It is typically of sudden onset (75% within a few seconds) and severe, often being the worst headache ever experienced by the patient. It may be the only symptom in up to one-third of patients. Approximately one in four patients presenting with sudden severe headache will have SAH. Other causes include benign thunderclap headache (40%), migraine, cluster headache, headache associated with sexual exertion, vascular headaches of stroke, intracranial haemorrhage, venous thrombosis, and arterial dissection, meningitis, encephalitis, acute hydrocephalus, intracranial tumor, and intracranial hypotension.
  • Up to 50% of patients experience a warning leak (sentinel haemorrage) in the hours to days before the major bleed. This headache may be mild, generalized or localized, resolve spontaneously within minutes to hours, or respond to analgesic therapy. It does, however, tend to develop abruptly and differ in quality from other headaches that the patient may have previously experienced. Hence a patient’s worst or first headache is suggestive of SAH.
  • Upper neck pain is common.
  • One-third of patients will develop SAH during strenuous exercise, e.g. bending or lifting, whereas in the remaining two-thirds it will occur during sleep or routine daily activities.
  • Nausea and vomiting are present in 75% of patients.
  • Brief or permanent loss of consciousness occurs in the majority of patients. Severe headache is usually experienced when the patient regains consciousness, although a brief episode of excruciating headache may occur prior to losing consciousness.
  • Seizures occur in 15% of patients and when associated with a typical headache are a strong indicator of SAH, even if the patient is neurologically normal when assessed.
  • Prodomal symptoms particularly third cranial nerve with papillary dilatation and sixth cranial nerve palsies are uncommon, but may suggest the presence and location of a progressively enlarging unruptured aneurysm.
  • No clinical feature can reliably identify SAH.
  1. Examination

There is a wide spectrum of clinical presentations, the level of consciousness and clinical signs being dependent on the site and extent of the haemorrhage:

  • On ED presentation, two-thirds of patients have impaired level of consciousness — 50% of these have coma. Consciousness may improve or deteriorate. An acute confusional state can occur which may be mistaken for a psychological problem.
  • Signs of meningism, including fever, photophobia and neck stiffness, are present in 75% of patients, but may take several hours to develop and may be absent in the deeply unconscious. Absence of neck stiffness does not exclude SAH.
  • Focal neurological signs may be present in up to 25% of patients and are secondary to associated intracranial haemorrhage, cerebral vasospasm, local compression of a cranial nerve by the aneurysm (e.g. oculomotor nerve palsy by posterior communicating aneurysm) or raised intracranial pressure (sixth-nerve palsy) or bilateral lower limb weakness (anterior communicating aneurysm).
  • Ophthalmological examination may reveal unilateral or bilateral subhyaloid haemorrhages or papilloedema.
  • Systemic features associated with SAH include severe hypertension, hypoxia and acute ECG changes that may mimic acute myocardial infarction.
  • A small proportion of patients present in cardiac arrest. Resuscitation attempts are vital, as half of survivors regain independent function.


  1. Imaging

A brain CT scan without contrast is the initial investigation of choice. In the first 24 hours after haemorrhage it can demonstrate the presence of subarachnoid blood in more than 95% of cases. The sensitivity, however, decreases with time owing to the rapid clearance of blood, with only 80% of scans positive at 3 days and 50% positive at 1 week. CT will also demonstrate the site and extent of the haemorrhage, indicate the possible location of the aneurysm, and demonstrate the presence of hydrocephalus and other pathological changes.

Magnetic resonance imaging (MRI) with FLAIR (fluid attenuated inversion recovery) is reliable in demonstrating early SAH and is superior to CT in detecting extravasated blood in the days (up to 40 days) following haemorrhage. Availability and logistical considerations make MRI impractical for use in the initial diagnostic work-up of SAH, but it may be considered in patients who present late.

CT angiography (CTA) is the preferred angiographic technique once SAH has been identified. Compared to catheter angiography it has a sensitivity of 95%, is readily available, and has a lower complication rate than catheter angiography. It should be performed as soon as the diagnosis is made. Where diagnosis has been made by CT, CTA should preferably be performed while the patient is still in the scanner. CTA is usually of sufficient quality to allow planning of endovascular or neurosurgical interventions.

Four-vessel cerebral angiography is the gold standard for confirming the presence of an aneurysm, its location and the presence of vasospasm, and was previously the preferred agiographic test. It is not, however, without risk. Neurological complications occur in ~1.8% of cases, with re-rupture of an aneurysm reported in 2–3%. It is also less available than CTA. These factors have seen it become less favored and used in selected cases only.

MR angiography is currently useful as a screening tool for the diagnosis of intracranial aneurysms in patients at increased risk.

In patients where SAH is present and no cause is found, then the distribution of extravasated blood on the CT scan should be reviewed. If this conforms to the perimensencephalic distribution of non-aneurysmal haemorrhage, then no repeat investigations are warranted. If, however, an aneurismal pattern of haemorrhage is present, then a second CTA is recommended as occasionally an aneurysm may have gone undetected on the original test.

  1. Lumbar Puncture

Lumbar puncture is necessary when there is clinical suspicion of SAH, the CT scan is negative, equivocal or technically inadequate, and no mass lesion or signs of raided intracranial pressure are found. In about 3% of patients with SAH the CT scan will be normal.

The diagnosis of SAH, then, is dependent on the finding of red blood cells not due to traumatic tap, or red blood cell breakdown products within the CSF.

Lumbar puncture should be delayed for at least 6 and preferably 12 hours after symptom onset to allow bilirubin to be formed from cell breakdown in SAH. Detection of bilirubin and xanthochromia is the only reliable method of distinguishing SAH from a traumatic tap. Proceeding to angiographic studies in every patient with bloodstained CSF would be expected to identify an incidental finding of a small unruptured aneurysm in about 2%.

It is important to measure the opening pressure when performing a lumbar puncture, as CSF pressure may be elevated in SAH or in other conditions such as intracranial venous thrombosis or pseudotumor cerebri, or low in spontaneous intracranial hypotension.

Xanthochromia, the yellow discoloration of CSF caused by the haemoglobin breakdown products oxyhaemoglobin and bilirubin due to lysis of red blood cells, is generally agreed to be the primary criterion for diagnosis of SAH and differentiates between SAH and traumatic tap. It is usually present within 6 hours of SAH and has been demonstrated in all patients with SAH between 12 hours and 2 weeks following the haemorrhage. Xanthochromia is not reliably detected by visual examination of centrifuged CSF. Spectrophotometric analysis of CSF for bilirubin is considered the most sensitive means of detecting xanthochromia. Owing to the time taken for haemoglobin to degrade into bilirubin and oxyhaemoglobin, xanthochromia may take up to 12 hours to develop. Hence controversy exists as to the optimal timing of lumbar puncture. Early lumbar puncture within 12 hours may have negative or equivocal CSF findings, whereas delayed lumbar may result in an increased risk of early re-bleeding as well as having practical implications for the ED.

In general, at least 6–12 hours should have elapsed between the onset of headache and lumbar puncture. Although detection of xanthochromia is indicative of SAH, it does not entirely rule out traumatic lumbar puncture and can occur in extremely bloody taps (>12 000 RBC/mL) or where the lumbar puncture has been repeated after an initial traumatic tap.

Other studies of the CSF, such as three tube cell counts, D-dimer assay and detection of erythrophages, have been found to be inconsistent in differentiating SAH from traumatic tap.

  1. General Investigations

General investigations to be performed include full blood examination, erythrocyte sedimentation rate, urea, electrolytes including magnesium, blood glucose, coagulation screen, chest X-ray and 12-lead ECG. ECG changes are frequently present and include ST and T-wave changes which may mimic ischaemia, QRS and QT prolongation and arrhythmias.


  1. Early Complications
  • Rebleeding: Up to 15% within hours of the initial haemorrhage, and overall 40% of patients re-bleed within the first 4 weeks without intervention. Re-bleeding is associated with a 60% mortality, and half of the survivors remain disabled.
  • Subdural haematoma or large inracerebral haematoma can be life-threatening and require immediate drainage. Similarly, a large intracerebral haematoma may be contributing to the poor clinical condition and warrant drainage simultaneously with treatment of the aneurysm.
  • Global cerebral ischaemia: Irreversible brain damage resulting from haemorrhage at the time of aneurysm rupture. This is probably secondary to a marked rise in incranial pressure resulting in inadequate cerebral perfusion.
  • Cerebral vasospasm: Clinically significant vasospasm occurs in approximately 20% of patients with SAH and is a major cause of death and morbidity. It tends to occur between days 3 and 15 after SAH, with a peak incidence at days 6–8. Vasospasm causes ischaemia or infarction and should be suspected in any patient who suffers a deterioration in their neurological status or develops neurological deficits. The best predictor of vasospasm is the amount of blood seen on the initial CT scan.
  • Hydrocephalus occurs in approximately 15% of patients with SAH. It can occur within 24 hours of haemorrhage and should be suspected in any patient who suffers a deterioration in mentation or conscious state, particularly if associated with slowed papillary responses.
  • Seizures.
  • Fluid and electrolyte disturbances: Patients with SAH may develop hypovolaemia secondary to excessive natriuresis (cerebral salt wasting), or alternatively may develop a syndrome of inappropriate ADH (SIADH).
  • Hyperglycaemia and hyperthermia, both of which are associated with a poor outcome.
  • Medical complications include pulmonary oedema, cardiogenic or neurogenic (23%), cardiac arrhythmias (35%), sepsis, venous thromboembolism and respiratory failure.
  1. Late Complications
  • Late re-bleeding, from a new aneurysm or regrowth of the treated aneurysm, is estimated at ~1.3% in 4 years for coiling and ~ 2–3% in 10 years for surgical clipping.
  • Anosmia: up to 30%.
  • Epilepsy: 5–7%.
  • Cognitive deficits and psychosocial dysfunction are common even in those who make a good recovery; 60% of patients report personality change.


The management of SAH requires general supportive measures, particularly airway protection and blood pressure control, as well as specific management of the ruptured aneurysm and the complications of aneurismal haemorrhage.

  1. General Measures
  • Stabilization of the unconscious patient, with particular attention to the airway. Endotracheal intubation with oxygenation and ventilation will be required in patients with higher-grade (4–5) SAH.
  • Close observation or GCS and vital signs.
  • In all patients, maintain oxygenation and circulation ensuring adequate (euvolaemic) blood volume.
  • Analgesia, using reversible narcotic analgesic agents, sedation and antiemetics as required. Ensure bed rest with minimal stimulation. Avoid aspirin and non-steroidal analgesic agents (NSAIDs).
  • Blood pressure control: Blood pressure levels are often of the order of 150/90 immediately following SAH, and in most patients can be adequately controlled by sedation and analgesia. Normotensive levels extending to mild to moderately hypertensive levels, especially in patients with pre-existing hypertension, are acceptable. Antihypertensive therapy should be reserved for patients with severe (mean arterial pressure > 130 mmHg) hypertension, or where there is evidence of progressive endorgan dysfunction, and short-acting antihypertensive agents (e.g., esmolol or nitroprusside) and intensive haemodynamic monitoring should be employed.
  • Seizures should be treated as they occur. The use of prophylactic phenytoin is controversial and has been linked with unfavorable functional and cognitive outcomes.
  • Correct electrolyte imbalances, hyponatraemia of excessive natriuresis must be differentiated from that of SIADH. Hypovolaemia is to be avoided.
  • Venous thromboembolism prophylaxis, initially with compressive devices and later with subcutaneous heparin following treatment of the aneurysm.
  • Treatment of hydrocephalus by ventricular drainage may be required.

Specific Treatment

  1. Prevention of Re-bleeding

Early occlusion within 72 hours secures the aneurysm, prevents re-bleeding, removes the clot and reduces the incidence of early complications and improves outcomes.

Endovascular occlusion by placing detachable coils in aneurysms under radiological guidance (coiling) has largely replaced surgical occlusion as the method of choice for prevention of re-bleeding in suitable cases.

Current evidence suggests that the relative risk reduction for poor outcome (death or severe disability) at 1 year with coiling versus surgical occlusion (clipping) is of the order of 24%, with an absolute risk reduction of 7%. Choice depends somewhat on anatomical considerations, as aneurysms are not equally amenable to this option.

Surgical clipping is now a second-line option for most patients. Modern techniques provide an estimated absolute risk reduction for poor outcome of 10%, with a relative risk reduction of 19%. Clipping is usually done early — when 3 days, and preferably within 24 hours.

Antifibrinolytic agents, including ɛ-aminocaproic acid, which inhibit clot lysis, reduce the incidence of re-bleeding after initial aneurismal rupture. Their use has, however, been associated with an increase in neurological deficits and failed to improve outcome. They are not advocated for routine use in SAH.

  1. Prevention of Delayed Cerebral Ischaemia

Cerebral ischaemia is often gradual in onset and involves the territory of more than one cerebral artery. Peak frequency is at 5–14 days after SAH. Calcium channel antagonists improve outcome in SAH, with a relative risk reduction of 18% and an absolute risk reduction of 5.1%. The current standard regimen is nimodipine 60 mg orally every 4 hours for 3 weeks. It should be commenced within 48 hours of haemorrhage.

Magnesium sulphate may also be useful as hypomagnesaemia is common and associated with the occurrence of delayed cerebral ischaemia and poor outcome. There is currently insufficient evidence to assess its effectiveness.

There are conflicting data about whether antiplatelet agents reduce the rate of cerebral ischaemia, and there is no evidence that they reduce the proportion of patients with poor outcome.

There are no proven treatments for delayed cerebral ischaemia, although induced hypertension, hypervolaemia and haemodiluation are plausible. In vasospasm unresponsive to medical management, emergency cerebral angiography with intra-arterial vasodilator infusion or transluminal balloon angioplastry may be considered where focal vessel narrowing is demonstrated.


SAH has a 40–60% mortality rate from the initial haemorrhage, with up to one-third of survivors having a significant neurological deficits generally being associated with a poor prognosis. Survival rates have been reported at 70% for grade I, 60% for grade II, 50% for grade III, 40% for grade IV and 10% for grade V SAH.

It is worth noting, however, that survival without brain damage is possible even after respiratory arrest. Even patients who make a good recovery may suffer cognitive and psychosocial dysfunction.

Aneurysm screening in patients who have survived aneurismal SAH is not advocated, as although these patients are at increased risk of new or current aneurismal bleeds, screening cannot be demonstrated to be cost effective or increase quality of life.

Incidental Unruptured Aneurysms

If an unruptured aneurysm is found incidentally, it raises the dilemma of the risk-benefit rationale between intervention and conservative management. Factors taken into account include age, aneurysm size and location, gender, country, comorbidity and family history. Such patients should be referred to a neurosurgical service for advice and counseling.


Clinical suspicion of the diagnosis of SAH gained from a history of sudden, severe or atypical headache demands a full investigation, including brain CT scan and, if necessary, lumbar puncture. Once SAH has been diagnosed, urgent neurosurgical referral and management are required.

related literatures:
    Adapted from: Textbook of Adult Emergency Medicine E-Book. Authored By Peter Cameron, George Jelinek, Anne-Maree Kelly, Lindsay Murray, Anthony F. T. Brown. References as cited include:
  1. Al-Shahi R, White PM, Davenport RJ, et al. Subarachnoid haemorrhage. British Medical Journal 2006; 333: 235–240.
  2. de Gans K, Nieuwkamp DJ, Rinkel GJ, et al. Timing of aneurismal surgery in subarachnoid hemorrhage: a systematic review of the literature. Neurosurgery 2002; 50: 336–340.
  3. Dorhoust Mees SM, RInkel GJE, Vermeulen M, van Gijn J. Calcium antagonists for aneurismal subarachnoid haemorrhage. Cochrane Database of Systematic Reviews, (4): CD000277. DOI: 10.1002/14651858.CD000277.pub3, 1999.
  4. Dorhout Mees SM, van den Bergh WM, et al. Aniplatelet therapy for aneurismal subarachnoid haemorrhage.
  5. Cochrane Database of Systematic Reviews (4): CD006184. DOI: 10.1002/14651858.CD006184.pub2, 2007.
  6. Edlow JA, Caplan LR. Primary care: Avoiding pitfalls in the diagnosis of subarachnoid hemorrhage. New England Journal of Medicine 2000; 342: 29–36.
  7. Naval NS, Stevens RD, Mirski MA. Controversies in the management of subarachnoid haemorrhage. Critical Care Medicine 2006; 34: 511–524.
  8. Roos YBWEM, Rinkel GJE, Vermeulen M, et al. Antifibrinolytic therapy for aneurismal subarachnoid haemorrhage.
  9. Cochrane Database of Systematic Reviews (4): CD001245. DOI: 10.1002/14651858.CD001245, 1998.
  10. Sawin PD, Loftus CM. Diagnosis of spontaneous subarachnoid hemorrhage. American Family Physician 1997; 55: 145–156.
  11. Suarez JI, Tarr RW, Selman WR. Current concepts: Aneurysmal subarachnoid haemorrhage. New England Medical Journal 2006; 354: 387–398.
  12. van Gijn J, Kerr RS, Rinkel GJE. Subarachnoid haemorrhage. Lancet 2007; 369: 306–318.