Subarachnoid haemorrhage caused by a ruptured aneurysm: diagnosis and management

Why the committee made the recommendations

People with a suspected subarachnoid haemorrhage can present with a wide range of symptoms and signs. There was little evidence to indicate which symptoms or signs specifically point to a diagnosis of subarachnoid haemorrhage. The committee agreed that the consequences of a missed diagnosis may be severe and include disability and death. Urgent investigation is required to confirm a diagnosis of subarachnoid haemorrhage and facilitate early treatment to prevent rebleeding from the ruptured aneurysm. Clinicians should therefore be alert to the possibility of subarachnoid haemorrhage in people presenting with unexplained acute severe headache, reduced consciousness, seizure, focal neurological deficit or other suggestive symptoms. Some people present with non-specific symptoms, such as nausea and vomiting, photophobia, or neck pain or stiffness. In these cases, a diagnosis of subarachnoid haemorrhage is more likely to be missed.

The committee's experience indicated that a 'thunderclap' headache is a presenting symptom in most people with a subarachnoid haemorrhage. However, they noted that this type of headache is a common presentation in emergency departments and around 10% of people who present with this symptom have a diagnosis of subarachnoid haemorrhage confirmed; the majority are diagnosed with other conditions, for example, migraine. The committee agreed that other symptoms and signs are also commonly seen in people with a subarachnoid haemorrhage. These can be used, together with clinical judgement, to support clinical assessment and guide decisions on further diagnostic investigations. The committee noted that these symptoms and signs of subarachnoid haemorrhage can also be caused by a number of other conditions. Therefore, a senior clinician should confirm the assessment of subarachnoid haemorrhage and arrange immediate referral for further investigation.

There is good evidence showing that non-contrast CT head scans carried out within 6 hours of symptom onset are highly accurate and can be used to rule out a diagnosis of subarachnoid haemorrhage, therefore avoiding the need for further investigation with a lumbar puncture. CT head scans done more than 6 hours after symptom onset are less accurate (see the section on diagnosing a subarachnoid haemorrhage for more information).

Evidence on decision tools, including the Ottawa Subarachnoid Haemorrhage Rule for Headache Evaluation, showed that these tools have a high level of accuracy in ruling out subarachnoid haemorrhage, but are less accurate at ruling it in, with a large number of false-positive identifications. Although these tools are beneficial in ensuring that no cases are missed, over-reliance on them risks harm from unnecessary imaging and invasive investigations. The committee noted that the tools use a broad range of symptoms and signs that are not specific to subarachnoid haemorrhage.

How the recommendations might affect practice

The recommendations may be useful for non-specialist clinicians, but are not expected to lead to significant changes in practice. Non-contrast CT head scans are the usual first-line investigation for suspected subarachnoid haemorrhage, and this is not expected to change.

Return to recommendations

Why the committee made the recommendations

Although there was limited evidence on the use of specific analgesia or sedation, the committee agreed that pain in adults with aneurysmal subarachnoid haemorrhage should be managed with analgesics in line with standard clinical practice. Headache is usually treated with simple analgesics such as paracetamol, escalating to opioids as needed. The committee agreed that the sedative and pupillary effect of opioid analgesics should be taken into account when doing a neurological assessment, but should not preclude their use.

How the recommendations might affect practice

The recommendations generally reflect current practice and are not likely to result in changes.

Return to recommendations

Why the committee made the recommendations

The committee looked at evidence on non-contrast CT head scans, lumbar puncture and MRI. In the studies, the CT scans were reviewed by either a general radiologist or a neuroradiologist. There was good evidence showing that CT head scans done within 6 hours of symptom onset have a high diagnostic accuracy. Taking into account the invasive risks of lumbar puncture, the difficulty of monitoring patients during MRI and the costs of both procedures, the committee concluded that these procedures should not be routinely offered if a CT head scan, done within 6 hours of symptom onset and reported and documented by an appropriately experienced radiologist, shows no evidence of a subarachnoid haemorrhage. In these circumstances alternative diagnoses should be considered and advice sought from a specialist, for example in neurosurgery, neuroradiology, neurology or stroke medicine.

If the CT head scan is done more than 6 hours after symptom onset, the evidence showed that diagnostic accuracy is reduced and false-negative results are more likely. The committee therefore agreed that further investigation with a lumbar puncture should be considered if a CT head scan done more than 6 hours after ictus does not confirm the diagnosis of subarachnoid haemorrhage.

When a lumbar puncture is indicated, the committee agreed that it should be done at least 12 hours after symptom onset, when bilirubin formation is sufficient to be detected reliably. The committee considered that the diagnostic accuracy of earlier lumbar puncture (to detect blood in the cerebrospinal fluid) is likely to be low because it can take several hours for blood to appear in the lumbar subarachnoid sac.

There was limited evidence on the relative accuracy of non-contrast CT head scanning at various time intervals greater than 6 hours after symptom onset, so the committee made a recommendation for research on timing of CT head scans.

How the recommendations might affect practice

Centres that routinely perform lumbar puncture after a negative CT head scan may see a reduction in the use of lumbar puncture in people with an early negative CT head scan. Reliance on severity scoring to determine transfer to specialist centres may reduce, leading to more people appropriately being transferred for treatment.

Return to recommendations

Why the committee made the recommendations

Evidence showed that CT angiography has a high level of accuracy in identifying aneurysms causing subarachnoid haemorrhage but evidence for the diagnostic accuracy of magnetic resonance angiography (MRA) was less compelling. CT angiography is the quickest to perform, is non-invasive, does not usually necessitate sedation or general anaesthesia and is available in most centres. MRA and DSA are more complex and time-consuming procedures that need specialist input and have a higher risk of complications. DSA is regarded as the 'gold standard' investigation and is currently commonly carried out when CT angiography is negative but there is a high suspicion of aneurysmal subarachnoid haemorrhage, whereas the complexities involved in obtaining high-quality MRA images make this less beneficial. The committee agreed that DSA should be reserved for instances when CT angiography has not detected a suspected aneurysm.

How the recommendations might affect practice

The recommendations largely reflect current practice and are not expected to lead to changes.

Return to recommendations

Why the committee made the recommendations

How the recommendations might affect practice

Why the committee made the recommendations

The committee noted that around half of people who survive an aneurysmal subarachnoid haemorrhage will have a second bleed from the culprit aneurysm within weeks, and the mortality from a second bleed can exceed 50%. Interventional treatment with endovascular coiling or neurosurgical clipping, if suitable, can secure the aneurysm and prevent rebleeding.

The committee acknowledged that interventional treatment is not suitable for some people who have a major neurological deficit after an aneurysmal subarachnoid haemorrhage, and that the costs of long-term nursing care or rehabilitation can be considerable. Very little evidence was found for this group, so the committee made a recommendation for research on interventions for aneurysmal subarachnoid haemorrhage in people with major neurological deficit.

Severity scoring systems might help clinicians identify people for whom intervention is likely to be justified, but none of the severity scoring systems currently in use reliably predicts morbidity and mortality in people with aneurysmal subarachnoid haemorrhage. In addition, the committee agreed that clinical state and severity score can vary over time, especially soon after symptom onset. Decisions on clinical management should therefore be based on a holistic patient assessment rather than solely on a severity score. Factors that should be considered by clinicians and discussed with the patient include: neurological status, performance status and comorbidities.

The evidence was not sufficient to determine the clinical effectiveness of endovascular coiling compared with neurosurgical clipping, although a small amount of evidence suggested that endovascular coiling might be more beneficial. Endovascular coiling is less invasive and potentially safer, so the committee agreed that it should be offered as the first option, taking factors such as aneurysm characteristics and the amount and location of subarachnoid blood into account. The committee agreed that neurosurgical clipping should be considered if endovascular coiling is not a suitable treatment option. They stressed that any procedure to secure the aneurysm should be performed without delay to minimise the risk of rebleeding.

Newer, more expensive interventional technologies are being used but there is little evidence on their effectiveness, so the committee made a recommendation for research on novel endovascular interventions.

How the recommendations might affect practice

Endovascular coiling accounts for around 75% of interventions done in current practice and the recommendation is not expected to change this. Most interventions are carried out within 48 hours of admission. However, this may vary according to the availability of interventional neuroradiologists and vascular neurosurgeons, and the capacity of neurosurgical centres. For people admitted during weekends, endovascular coiling should be available from the same interventional neuroradiology teams who will provide thrombectomy for stroke, in line with the recommendations on thrombectomy in the NICE guideline on stroke and transient ischaemic attack in over 16s. However, provision of neurosurgical clipping during weekends may necessitate changes to services such as the setting up of appropriate networks.

Return to recommendations

Why the committee made the recommendations

There was no evidence on the routine use of direct intracranial pressure monitoring for raised intracranial pressure. Very limited evidence from 1 study on transcranial doppler monitoring for vasospasm suggested an increase in mortality, morbidity and length of hospital stay compared with no transcranial doppler monitoring. The committee agreed that these outcomes were unlikely to be directly caused by the transcranial doppler monitoring. However, they were concerned that such monitoring might influence subsequent decisions, for example about investigations or interventions, and so indirectly affect outcomes. They therefore agreed that transcranial doppler monitoring should only be used in the context of clinical research. The committee noted that there is increasing enthusiasm for the use of this modality and made a recommendation for research on transcranial doppler monitoring.

There was no evidence on the investigation of unexplained neurological deterioration in people with aneurysmal subarachnoid haemorrhage. The committee agreed that in current practice unexplained neurological deterioration is initially investigated with a non-contrast CT scan, which can indicate the cause of deterioration including ventricular enlargement suggestive of hydrocephalus, cerebral ischaemia, intracranial haematoma or evidence of rebleeding.

How the recommendations might affect practice

In current practice the use of transcranial doppler monitoring varies widely. The recommendation is likely to stop this type of monitoring in centres that use it in routine practice. CT head scans are used to investigate unexplained neurological deterioration in current practice and the recommendation will not affect this.

Return to recommendations

Why the committee made the recommendations

How the recommendations might affect practice

The recommendations reflect current practice.

Return to recommendations

Why the committee did not make a recommendation

There was little evidence on diagnosing and treating intracranial hypertension. The committee agreed that the diagnostic accuracies of transcranial doppler and ultrasound measurement of optic nerve sheath diameter are too low for reliable detection of intracranial hypertension when compared with direct intracranial pressure measurement. The committee also acknowledged that there is no evidence that interventions to lower intracranial pressure improve clinical outcome.

The committee agreed that raised intracranial pressure is common in people with aneurysmal subarachnoid haemorrhage, but intracranial hypertension that impedes blood flow to the brain and contributes to brain injury is generally only seen in the most severely ill. These people are usually unconscious or need ventilation in an intensive care unit. They are a heterogeneous population and management varies widely, with some clinicians advocating routine monitoring of intracranial pressure to guide intervention (such as drainage of cerebrospinal fluid, hypertonic saline or vasopressor therapy) to lower intracranial pressure and maintain cerebral perfusion. Other clinicians favour management without intracranial pressure monitoring.

Intracranial pressure can be monitored by inserting an intracranial pressure bolt or using an external ventricular drain inserted to manage acute hydrocephalus. The committee acknowledged that insertion of a pressure bolt is associated with risk, and monitoring of intracranial pressure will only improve outcome if it leads to effective intervention.

The committee debated the variation in current practice and were not able to reach a consensus on the role of interventions to diagnose and treat intracranial hypertension, and so they made a recommendation for research.

Return to recommendation for research

Why the committee made the recommendation

There was very little evidence on delayed cerebral ischaemia so the committee based the recommendation on their clinical experience. They noted that current practice for managing delayed cerebral ischaemia is to maintain cerebral blood flow to prevent or limit cerebral infarction. Intravenous fluid is usually given to ensure euvolemia and if symptoms persist a vasopressor is administered to raise systemic blood pressure. The committee agreed with this approach, but added that the improvements seen after these measures may be temporary, and there was no evidence of impact on longer-term outcomes. Treatment for people whose condition is not improved by vasopressor therapy varies widely. Some clinicians recommend cerebral angiography and intra-arterial therapies, including intra-arterial vasodilators and angioplasty. However, the committee were unable to reach a consensus on the use of intra-arterial therapies. They made recommendations for research on intra-arterial therapies to manage delayed cerebral ischaemia and vasopressors to manage delayed cerebral ischaemia.

How the recommendation might affect practice

The recommendation reflects current practice.

Return to recommendation

Why the committee made the recommendations

Evidence from surveys and interviews with people who had an aneurysmal subarachnoid haemorrhage showed that those who did not receive clear information about their medical care after discharge felt anxious and 'abandoned'. Those who were given this information, together with details of who to contact for ongoing advice, said they felt more supported. The committee agreed that the follow-up care plan should be included in the person's medical record and discharge correspondence.

How the recommendations might affect practice

The amount and quality of information given to people about follow-up care after an aneurysmal subarachnoid haemorrhage varies. This recommendation can be expected to improve the provision of this information.

Return to recommendations

Why the committee made the recommendation

No clinical evidence was found on follow-up neuroimaging. Based on their knowledge and experience, the committee agreed that there is a risk of aneurysm recurrence, progression of non-culprit aneurysms and formation of new aneurysms. They agreed that follow-up imaging should be considered, based on the person's clinical situation and risk factors.

How the recommendation might affect practice

The recommendation reflects current practice.

Return to recommendation

Why the committee made the recommendations

There was not enough good evidence to enable the committee to recommend a preferred management option for non-culprit aneurysms, although they agreed that the risk of a non-culprit aneurysm rupturing is higher in people who have had an aneurysmal subarachnoid haemorrhage than those who have not. Based on their experience, the committee agreed that the overall probability of a non-culprit aneurysm rupturing is low, so interventional treatment of all non-culprit aneurysms is unlikely to be a cost-effective strategy. They agreed that conservative management usually includes monitoring of the aneurysm with MRA to detect changes in the aneurysm's size or shape. The committee were in agreement that the frequency of monitoring should be based on a balance between the risks of aneurysm rupture and the risks of interventional treatment, and take into account multidisciplinary team, neuroradiological and neurosurgical opinion, and the person's preferences.

The committee's experience was that people with unruptured aneurysms who are offered conservative management are often anxious about the possibility of a future rupture. They therefore recommended that all treatment options be discussed with the person.

How the recommendations might affect practice

Current management of unruptured aneurysms varies. The recommendations are not expected to lead to substantial changes in practice.

Return to recommendations

Why the committee made the recommendations

How the recommendations might affect practice

The recommendations on hypertension, smoking, headaches and seizures are not expected to change practice. The recommendations on managing conditions treated with antiplatelets or anticoagulants may reduce delays in prescribing these medicines for people who have had a successfully treated aneurysmal subarachnoid haemorrhage. Although specialist input may be needed, this is not expected to have a large impact on services.

Return to recommendations

Why the committee made the recommendation

No evidence was found on investigating relatives for intracranial aneurysms. The committee recognised that first-degree relatives of people who have had an aneurysmal subarachnoid haemorrhage are at higher risk of intracranial arterial aneurysm than the general population. The committee agreed that investigating for aneurysms in first-degree relatives may produce health benefits but could also lead to harm, including unnecessary anxiety and a consequent increase in visits to GPs and emergency departments. Moreover, the optimal timing and frequency of investigations to detect and monitor intracranial aneurysms in first-degree relatives is unknown.

The committee were aware that the NHS does not currently support routine screening for intracranial arterial aneurysms in relatives and any change in NHS policy could have a significant resource impact.

The committee acknowledged that in current practice investigation is typically only recommended for people thought to have a significant risk of having a brain aneurysm that could rupture at some point in the future, and this would usually only apply to people with 2 or more first-degree relatives who have had an aneurysmal subarachnoid haemorrhage. This reflects the advice given on the NHS website on screening for brain aneurysms.

The committee agreed that the lack of evidence for routine testing of relatives and current practice on testing should be explained to people who have had a subarachnoid haemorrhage, and their families as appropriate. Given the importance of this issue and the lack of evidence, the committee made a recommendation for research on investigations for relatives.

How the recommendation might affect practice

The recommendation is not expected to lead to a change in practice.

Return to recommendation

Why the committee made the recommendations

Evidence from studies using surveys and interviews with people who have had an aneurysmal subarachnoid haemorrhage showed that they value information that is clear, concise and tailored to their own needs. Participants in the studies highlighted how difficult it was to remember information given to them verbally at the time of hospital admission and the need to have this information in a written form that they can take home at discharge. They also stressed the importance of information about their medical care and what to expect after discharge, including common symptoms, possible complications and where they can get advice and support.

The committee used their experience to agree topics that should be included, as a minimum, in the information given. They highlighted the importance of ease of access to information for people throughout their care pathway.

In the committee's experience, some people want information about the risk of recurrence of subarachnoid haemorrhage and some do not. They agreed that discussions should include the effect of individual variables on the person's risk of recurrence. The committee discussed the lack of a validated risk tool to inform estimates of individual risk and made a recommendation for research on a risk stratification tool to estimate risk of recurrence.

How the recommendations might affect practice

The content, delivery and quality of information given to people with aneurysmal subarachnoid haemorrhage varies widely. The recommendations are expected to encourage discussion that is tailored to the person's preferences.

Return to recommendations