Key facts:
Authors: Katharine Elliott and Lewis Thorne
Top Tip: All investigations in SAH have problems; you need to know a lot about their pros + cons
Key Differential Diagnoses
- Meningitis
- Migraine
- Other forms cerebral haemorrhage
Key Investigations
- FBC, ESR, CRP
- Coagulation screen
- U+E, LFT, Bone, Glucose
- ECG, CXR
- CT ('N' does not exclude)
- ± LP (not needed if CT confirms; don't do, if drowsy)
Key Treatments
- PO NIMODIPINE 60 mg 4 hrly
- Endovascular coil, or open surgery (clip)
Key Management Decision
- CT angiogram
This page is dedicated to the cousin of one of the editors (AS), Susannah Simpson, who experienced a SAH in 2008; but made a full recovery after endovascular coiling
Background
Diagnosis often missed; either due to poor history taking, or lack of understanding of investigations. These are more complicated than you think
Introduction
- The history is vital, this takes experience
- CT and LP can be normal
- Spontaneous bleeding into subarachnoid space, can present as a range of syndromes, from mild meningism, to deep coma. Patients may have earlier experienced a small warning ('sentinel') headache due to warning leak from an aneury sm
- The main point of SAH 'treatment' is to prevent a second SAH
- It seems an 'easy syndrome' to diagnose. It is not at all. It is prone to both over and under-diagnosis
- The diagnosis is often missed, especially in the Emergency Department[Ref]
[Ref]
Definition
- Bleeding into subarachnoid space
Epidemiology
- Incidence (Aneursymal) SAH = 9/100,000 pa
[Ref] - 2% normal population have a berry aneurysm; most don't rupture
- When rupture and cause SAH, 25% die on first day; further 25% die in first month
- 15% Berry aneurysms multiple
- 4% Occur after intercourse
- Close relatives have a 2-50x increased risk of SAH (see below)
Note: associated with PCKD, coarctation of aorta, Ehlers-Danlos
Causes
Berry aneurysm (80%)
- May be associated with various congenital diseases, eg PCKD
- 90% present as SAH, 7% as signs of SOL, 3% incidental finding at postmortem
- Peak age 40-50y
- Only aneursymal SAH has a high rate of (usually) fatal rebleed (rebleed 70% mortality); this is why it's so important not to miss diagnosis; [Ref]
Other causes (20%)
- Other types aneurysm (eg mycotic)
- 15%: no cause is found
- 5%: Arteriovenous malformation
- Peak age 20-40y
- Other (warfarin, cocaine use, tumours, vasculitis etc)
Risk factor
- Alcohol (moderate reduces likelihood of ischaemic stroke though)
Symptoms
- Sudden onset of meningism (classic triad of headache, neck stiffness and photophobia)
- 'Thunderclap headache'
- 'Like being chopped in the back of the neck'
- Vomiting, collapse
- Of focal neurological deficit (eg unilateral weakness)
Key questions
- "When did the headache start, and did it start suddenly?"
- "What is your normal pattern of headache, and do you have migraine?"
- "Do you have family history of SAH, brain haemorrhage or sudden death at a young age?"
Signs
- Nil, or
- Neck stiffness (Kernig's Sign?), mild fever
- Of focal neurological deficit (III nerve palsy suggests post communicating artery aneursym)
- Reduced conscious level
- Seizure, or confusion/decreased conscious level, post seizure (postictal)
Note: papilloedema suggests raised ICP (don't do LP, as can cause 'coning'). Rising BP, and bradycardia = Cushing's Reflex, and is something that happens near death
9 reasons why SAH harder to diagnose, exclude and manage than you would think
- History is vital, and this takes experience
- Normal CT does not exclude
- CT done too late (or early) can miss diagnosis
- Normal LP does not exclude
- LP done too early (or late) can miss diagnosis
- Blood in LP does not diagnose
- CSF bilirubin has to be measured in right way (to diagnose SAH retrospectively)
- 2% normal people have a berry aneursym. So, this may not have bled (and is not going to) and they have a non-SAH related headache. Unnecessary coiling or clipping can do them more harm than doing nothing
- Neither coiling or clipping treats the current bleed
Investigation
It is all in the history: anyone with a good story needs further investigation, even if CT -ve
Blood
- FBC, ESR, CRP
- Coagulation screen
- U+E, LFTs, Bone, Glucose
- ABG, if unwell
Other
- CT ('normal' does not exclude; but if done early (<12h), will detect 98%)
- Then, if required, LP
Note: ie, if CT confirms, no need for LP- Classically, at least three tubes of CSF are collected. It is said that, if an elevated number of red blood cells is present equally in all bottles, this indicates SAH. If the number of cells decreases per bottle, 'traumatic tap' more likely
- If subarachnoid hemorrhage may have occurred some days ago, CSF is also examined for xanthochromia-the yellow appearance of centrifugated fluid. More sensitive is spectrophotometry for bilirubin. These tests may be useful for retrospective diagnosis of SAH. Problems with all techniques are debated below
- Bilirubin will become +ve after 8h, so if you do LP too early (see risk stratification below), you may miss it - ie sometimes 'better' to delay it a little
- Don't do LP if clinical, or CT evidence of raised ICP, see above; also remember to record 'opening pressure' (10-20 cm H20) - if low, and CSF otherwise normal, consider diagnosis of low pressure headache)
[Ref]
[Ref]
- CXR, ECG
Problems with LP
- The role of LP in SAH is controversial. There are problems with all the commonly used techniques: the three bottle technique has not been validated, and the ability of the operator to determine whether the tap was traumatic, using this technique, is 50% - ie you can toss a coin
- The LP should be done after a minimum of 8hrs, as the only valid test is spectrophotometry for billirubin. Done any earlier the cells will not have lysed to produce billirubin. Holding up the specimen to the light again has a reliability of 50%
- The LP is only of any use if it is done in daylight, the laboratory is expecting it, a decent amount is sent (around 10ml, if you send three drops as some people have been told the lab just lets it down with saline, making it very inaccurate), the specimen is spun down and the supernatant stored in the dark (otherwise the cells from a traumatic tap will continue to lyse and produce a false +ve)
- Few hospitals have spectrophotometry on site. So, if the results come back two days or more down the line, too late for a clinical decision
Other diagnostic difficulties
- The difficulties of diagnosis are demonstrated in this paper: [Ref]
Key investigations
- CT/LP
Unlike this CT ('Star of David' sign), the CT can be normal
Specialist investigation
- CT angiogram
Differential diagnoses
- Meningitis
- Migraine
- Other forms of cerebral haemorrhage (intracerebral, extradural)
- Cluster headache
- Cerebral venous thrombosis (CVT)
Treatment
Sudden onset headache with maximal severity <15 minutes is a subarachnoid haemorrhage until proven otherwise
Treatment (first line)
Drugs
- PO NIMODIPINE 60 mg 4 hrly, started within 4 days of SAH, and continued for 21 days; reduces Ca mediated cell death associated with ischaemia; though use is debated:
[Ref] ##title##Centre for reviews and dissemination: Efficacy of prophylactic nimodipine for delayed ischemic deficit after subarachnoid hemorrhage: a meta-analysis. Barker FG et al, 1997 - Analgesia (eg PO CODEINE PHOSPHATE 30-60 mg 4-6 hrly) + antiemetics
Procedures
- IV (+ fluids), keep well hydrated
- Advocated by neurosurgeons; little evidence for:
[Ref] - Protect airway, if drowsy
- OXYGEN, if hypoxic
- Avoid hypertension
Key management decision
- CT angiogram
Stop
- Aspirin/warfarin
Treatment (second line)
- Patients are normally put on bed rest (not clear why), and laxatives, to prevent straining secondary to opiate use
- If unwell, urinary catheter, CVP line, arterial line
- CT angiogram
Note: CT angiography may be used to improve the diagnostic power of CT and reduce or remove the need for lumbar puncture. But the problem of performing CTAs is that incidental aneurysms that exist in around 2% of the population maybe picked up
[Ref] - Then, if CT angiogram positive, endovascular coil (if aneurysm)
[Ref] - Or, surgery ('clip'; if aneurysm): coiling is probably better than surgery:
[Ref] Timing of surgery (early vs late) is debated:
[Ref]
Prescribing issues
- Make sure nurses are giving NIMODIPINE 4 hrly; write up at 2400, 0400 and 0800; so patient gets a little sleep
Treat Early/Late Argument
- The primary issue of tertiary SAH care is preventing harm. With the exception of treating hydrocephalus, or using hypertension to reverse vasospastic ischaemic deficit, you do not do anything to make people 'better'
- Treating the aneurysm is all about preventing a second haemorrhage, which has a mortality of 70%
- However, treating aneurysms in the badly injured brain, or during the vasospastic period (days 4-10, on the whole), is associated with a worse outcome than the natural history of the disease, due to the high incidence of ischaemic stroke, whether surgery or coiling. This is the basis of the early/late argument
Admit
- Always
Bed plan
- Neurosurgery
- ± ITU
Referrals
Medical
- ± Neurosurgery
- ± ITU
Other
- Neuro rehabilitation
The Rest
If drowsy, focal neurology (including papilloedema) or fitting, do CT first; then decide on ?LP
Complications
Early
- Fitting
- Arrthymias
- ECG abnormalities (including ST elevation)
- Vasospasm (can cause cerebral ischaemia; nimodipine may reduce)
Later
- Rebleeding (30%; more headache, or further reduction LoC)
- Hydrocephalus, 20%; may require shunt
Late
- Further SAH from other aneursyms
Risk stratification (who is 'safe' to send home)
- As a VERY general rule, if a <12 hr CT and all components of the LP are negative, it is safe to discharge the patient. BUT as the LP will be negative for bilirubin in the first 8 hrs, in some patients (eg with recent headache) it is better to delay the LP a little, to look for bilirubin in the CSF
- This is a complex area, with no one perfect test - you will need senior advice. Routine CT angiography is not advisable (it has risk, and may pick up an incidental aneursym that has not bled)
[Ref]
Prognosis
- Depends on conscious level (see table below)
[Ref] - 25% die on first day; further 25% in first month; if rebleed 70% mortality
[Ref]
2° Prevention + Health Promotion
- CT angiogram, endovascular coil or surgery (clip) in relatives?
Note: screening in normal person may be worthwhile; close relatives have 2x inc risk SAH, if 1 relative has had a SAH (but 50x, if has 2)
[Ref]
[Ref]
Don't forget
- If drowsy, focal neurology (including papilloedema) or fitting, do CT first; then decide on ?LP
- Ie, DON'T DO LP IF SUSPECT RAISED ICP (like meningitis)
- No one test absolutes excludes SAH
- The main point of SAH 'treatment' is to prevent a second SAH
Red flags
- Focal neurology
- Reduced conscious level
- Rebleeding
