The first investigation should be a non-contrast CT scan of the head [11,18].
If there is abnormal neurology or a reduced conscious level this should take place as soon as possible after the index episode. Ideally, this would hold true for all patients as the sensitivity of CT scanning is higher the closer it is performed to the index episode.
The largest studies on ED patients reveal an overall sensitivity of 92-93% though this may be substantially higher, perhaps even 100%, if performed in the first 6 hours [43-45] The sensitivity drops off to about 85% at 3 days and 50% at 1 week  as blood diffuses away from the site of haemorrhage.
Interpretation of the results of the CT scan may also be subject to spectrum bias (patients who are neurologically well will likely have smaller bleeds harder to see on CT) and the fact that neuroradiology reporting is superior to that of a general radiologist .
CT appearances of SAH
The distribution of blood on the initial CT Head scan can be helpful in distinguishing aneurismal SAH (Fig 4) from perimesencephalic haemorrhage (Fig 5) [18,19,37].
| || |
|Fig 4: aneurismal SAH (click to enlarge) ||Fig 5: Perimesencephalic haemorrhage (click to enlarge) |
Ruptured aneurysms are most often found in the anterior communicating artery (blood in interhemispheric fissure) (Fig 4) followed by the internal carotid artery, middle cerebral artery (blood in the Sylvian fissure) and vertebrobasilar circulation.
About 20% of patients with SAH will have multiple aneurysms so the CT pattern of blood is important in identifying the probable culprit .
Patients with perimesencephalic patterns of SAH (blood localised to the midbrain cisterns) do very well  with no specific treatment. However, non-contrast CT brain appearances are not unique  and CT angiography (CTA) is required in these patients to exclude a ruptured vertebrobasilar aneurysm [47,48].
All patients with CT-proven SAH should undergo CT or formal angiography to identify the aneurysm responsible or confirm the absence of such in cases of perimesencephalic haemorrhage [47-51].
Since CT does not have 100% sensitivity, the concern is that a SAH may be missed despite a normal scan. Traditional teaching and expert opinion still mandate a lumbar puncture (LP) and cerebrospinal fluid (CSF) analysis for xanthochromia in every patient with a negative or non-diagnostic CT head scan [18,32,37,42,45] as evidenced by national guidelines in the United Kingdom (UK) [49,50] and the United States (US) [11,51].
A negative CT alone is not yet enough evidence to exclude SAH [43,45,47-51].