The diagnosis of subarachnoid hemorrhage (SAH) requires suspicion on the part of medical personnel in initial contact with the patient. Although the abrupt onset of a severe headache is a common presentation, less severe headache with or without an associated brief loss of consciousness, nausea, or vomiting may also signal SAH, and an initial, mild hemorrhage is sometimes missed. A milder but unusual headache for an individual patient that occurs days or up to several weeks before a subsequent presentation with more severe symptoms that then leads to the SAH diagnosis is called a sentinel headache or warning leak. This may occur in 10% to 25% of those with later, clinically apparent SAH. Although there are some reports of SAH occurring with strenuous activities and exertion, most occur during lighter activities related to daily life. In any patient with new onset of an unusually severe or atypical headache, particularly if associated with a brief loss of consciousness, nausea, vomiting, stiff neck, or any focal neurologic findings, computed tomography (CT) of the head without contrast should be performed to determine whether intracranial hemorrhage is present (see algorithm for the management of SAH in Appendix E-4).

If obtained within 24 hours, the CT scan is abnormal in approximately 98% of cases of SAH and reveals the increased attenuation caused by hemorrhage in the subarachnoid space (Fig. 14-1). If the CT scan is negative, including no evidence of subtle SAH in the posterior horns of the lateral ventricles or over the sulci, then a lumbar puncture (LP) should be performed if there are no contraindications for the procedure (see Chapter 7 for details). If the CT scan shows evidence for SAH or intraparenchymal hemorrhage, an LP will not contribute significant additional diagnostic information and sometimes can be dangerous, especially if parenchymal blood is present. Magnetic resonance imaging (MRI), with various techniques including gradient echo imaging, diffusion-weighted imaging, and fluid-attenuated inversion recovery sequences, is as reliable as CT in demonstrating subarachnoid blood in the first few days after SAH but is less often available on an emergency basis. After a few days, MRI becomes increasingly superior to CT in demonstrating subarachnoid blood up to approximately 40 days after a hemorrhage.

The typical patient with SAH has grossly bloody spinal fluid. However, traumatic LP must be differentiated from true SAH. Three or four successive tubes of cerebrospinal fluid (CSF) are collected, and, if the specimens show progressively less blood, a traumatic LP is suggested. Clotting of the specimen virtually
never occurs with true SAH. Xanthochromia, or the yellow-tinging of CSF caused by hemoglobin breakdown products, is present in the supernatant as early as 6 hours after SAH and remains in the spinal fluid for an average of 2 to 3 weeks. Erythrocytes often disappear within several days after SAH. The CSF may not show xanthochromia if small numbers of erythrocytes are present from SAH (approximately ≤400), and xanthochromia has been reported in rare instances with traumatic LP if the erythrocyte count is more than 200,000 per µL. Spectrophotometry of CSF, when available, for the detection of oxyhemoglobin or bilirubin may also be of value in distinguishing between traumatic LP and true SAH if the visual inspection for xanthochromia is not definitive.

In addition to the classic severe, striking, generalized headache with meningismus to a much milder headache that resolves within 24 to 48 hours, other aspects of the neurologic history and examination are important in clarifying the potential presence of SAH. Because aneurysms may cause a mass effect on adjacent structures, including the cranial nerves and brainstem, symptoms that precede SAH may include diplopia, facial weakness, extremity weakness, and unsteadiness. In addition, aneurysms can occasionally present with transient symptoms caused by ischemia or seizures. Ischemic spells typically are thought to be caused by thrombus formation within or adjacent to the aneurysm and
lead to distal embolization and ischemia. In addition to focal symptoms, more generalized symptoms, including transient loss of consciousness caused by a sudden increase in intracranial pressure (ICP), may occur at the time of the hemorrhage. Approximately one third of patients develop focal signs at or around the time of headache onset, and almost half develop unresponsiveness that lasts an hour or more.

A comprehensive neurologic examination should be performed in the patient with SAH, but detailed muscle strength testing, particularly of proximal muscles, should be avoided because of the potential increased risk of rebleeding. On neurologic examination, the presence of any focal findings is important and may aid in localizing the aneurysm. Particular findings that may be useful include monocular visual deficits, which may indicate an ophthalmic artery aneurysm with optic nerve compression. Extraocular muscle abnormalities may indicate internal carotid artery, basilar artery, or ophthalmic aneurysms. A unilateral third cranial nerve palsy early in the course of the SAH strongly suggests an aneurysm of the posterior communicating artery. Sixth cranial nerve palsy may result from increased ICP or from a basilar artery aneurysm. Hemorrhage or mass effect from a middle cerebral artery aneurysm may cause aphasia and hemiparesis. Leg weakness and changes in behavior may be associated with rupture or mass effect from an anterior communicating artery or anterior cerebral artery aneurysm that affects the frontal lobe(s). A vitreous hemorrhage in the eye, also called Terson’s syndrome, can occur in SAH or in other disorders associated with rapidly increased ICP. The characteristic hemorrhage in a preretinal location will appear on funduscopic examination as a well-delineated, focal area of hemorrhage.

The initial CT scan may reveal the location of the hemorrhage as well as the possible source for the event, although standard CT imaging without contrast is unlikely to detect an uncalcified aneurysm, especially one that is less than 7 mm in size. An aneurysmal cause must be suspected in all cases of SAH or intracerebral hemorrhage (ICH) that cannot be clearly explained by an alternative cause. Computed tomography angiography (CTA) may reveal a causative aneurysm, with high sensitivity for aneurysms greater than 3 mm in maximal diameter, and also guide the selection of treatment modality. MRI and magnetic resonance angiography (MRA) may similarly reveal an underlying cause, but the availability of urgent MRA and the longer duration of the MRA assessment make it less useful than CT/CTA in the emergency setting. After SAH is diagnosed and the patient is initially stabilized, urgent cerebral arteriography, now often available as three-dimensional rotational arteriography, is indicated if the etiology of the SAH is uncertain from the CTA and for decision making regarding the potential for endovascular treatment. Cerebral arteriography is the most sensitive study available; it reveals small saccular aneurysms and arteriovenous malformations (AVMs) as well as the associated morphology and anatomic characteristics.

If the initial arteriogram performed for SAH fails to reveal an aneurysm, a repeat study is usually performed 1 to 3 weeks later in an attempt to detect an aneurysm that was not visible on the earlier study. An exception to this recommendation for repeat arteriogram is an SAH that is localized solely to a small clot anterior to the brainstem, the so-called perimesencephalic or pretruncal SAH. Posterior circulation aneurysms can cause such a CT appearance in some cases, so there remains some controversy regarding the following issues in perimesencephalic SAH: (a) whether a negative CTA alone is sufficient to rule out an aneurysmal cause and (b) whether there is a need for repeat arteriography. In general,
a cerebral arteriogram is often recommended in these patients, but a repeat arteriogram is not usually needed if the classic pretruncal SAH pattern is noted. Other studies (Table 14-1) may also be undertaken to identify other potential underlying factors that are causing or exacerbating SAH, such as hypocoagulable states.