The annual incidence of ICH is 24.6 cases per 100,000 inhabitants (95 % CI 19.7–30.7) [2], gradually increasing with age [2–4]. It accounts for 10–15 % of all brain strokes [5], with a variability relating to geographic areas and age. ICH is prevalent in the Asian population where it accounts for 30 % of all strokes [6] and in the age group under 45 years, in which ICH and SAH are 25–55 % of the total number of strokes [7].

ICH occurs more frequently in men than in women in all age groups [2, 8], especially in Japanese population [2]. As for ethnicity, there is a higher prevalence of ICH in Asian [2, 5] and black populations [9] which might be correlated to the higher prevalence of hypertension.

ICH can have primary and secondary causes.

Among the primary causes are hypertensive angiopathy and cerebral amyloid angiopathy (CAA).

Hypertensive Angiopathy

High blood pressure is the most important cause of ICH [10] and 50 % of all ICH is caused by hypertension. Hypertensive angiopathy is the predominant cause of ICH also in subjects aged between 40 and 50 years [11]. High blood pressure increases the risk of ICH particularly in patients who do not comply to hypertension therapy and in patients younger than 55 years old who are habitual smokers [12].

By contrast, appropriate control of arterial hypertension reduces the risk of ICH. The majority of ICH related to hypertensive angiopathy occurs because of the rupturing of small perforating arteries such as lenticulostriate, thalamus perforating arteries and the arteries that originate from the basilar artery.

ICH is the type of stroke which leads to higher mortality and disability. 30-day mortality is estimated at between 32 and 50 % [2, 4, 20] and the 1-year survival is 46 % [19]; only 28–35 % of the patients who survive are independent 3 months after the acute event. In ICH patients, the “do not resuscitate” order is quite frequent and reaches 35 % in a recent study [21]. Factors associated with the “do not resuscitate” order are advanced age, ICH severity and early clinical deterioration.

Neurological deterioration commonly occurs in the first hours after onset of symptoms: 20 % of patients present it in the pre-hospital phase; a further 15–23 % of patients show signs of clinical deterioration after arriving at the hospital [22]. Neurological deterioration often signals constant bleeding resulting in expansion of the hematoma.

ICH is a medical emergency and rapid diagnosis of the aetiology of the bleeding is also essential for appropriate management of the patient.

Treatment of the acute phase includes general interventions for critically ill patients, such as airway management, and specific treatments for ICH designed to prevent expansion of the hematoma, reduce intracranial hypertension and treat and prevent complications. Treatment of high blood pressure and intracranial hypertension as well as restoration of coagulation should be carried out beforehand in the emergency departments where patients with ICH are evaluated.

The risk of neurological deterioration and cardiac instability is higher in the first 24 h after onset of symptoms; clinical and haemodynamic monitoring should be carried out in intensive/semi-intensive care units. Admission to a stroke unit results in a significant improvement in the prognosis of patients. Furthermore, the mortality rate is significantly lower in patients treated in a stroke unit compared to those treated on a medical ward (risk ratio (RR) 0.73; 95% CI 0.54–0.97, p=0.02) [44].

The incidence of SAH is approximately 10 cases per 100,000 people per year with a range from 2 to 27 cases per 100,000 per year linked to the geographic areas under consideration. SAH is more common in Japan and Finland (22.7 and 19.7 cases per 100,000/year, respectively) and least frequent in China (2 cases per 100,000/year) and in Central and South America (4.2 cases per 100,000). In the United States and in Italy, the incidence of the disease is 9.7 and 10.8 cases per 100,000 per year, respectively [85, 86]. It is higher in black and Hispanic people than in white Americans [86].

In recent decades there has been a reduction of 0.6 % in the incidence of total SAHs, smaller than that observed for stroke in general [85].

The average age of SAH patients is lower than the age of patients suffering other types of stroke and, although the incidence increases with age, approximately 50 % of SAH patients are younger than 55 years [87]. In recent decades an increase has been observed in the average age of patients, from 52 to 62 years [88].

As far as gender is concerned, women have a slightly higher incidence than men (1.24 times), which is evident after 55 years and increases after that age [85, 88].

In 85 % of cases, SAH is secondary to spontaneous rupture of a cerebral aneurysm; in 10 % of cases it is an idiopathic SAH, frequently located in the perimesencephalic area; in 5 % of cases it is due to rarer causes such as arterial dissection, arteriovenous malformations, dural arteriovenous fistulas, mycotic aneurysms, fusiform aneurysms, cerebral amyloid angiopathy, reversible vasoconstriction syndrome and vascular lesions in the spinal cord.

The modifiable risk factors include active and passive cigarette smoking [89], high blood pressure and high alcohol consumption. Each of them doubles the risk of aneurysmal SAH (aSAH) [90]. Another risk factor is represented by the abuse of sympathomimetic substances (cocaine).

The non-modifiable aSAH risk factors are family history, understood as a first-degree relative having suffered an SAH, genetic diseases such as polycystic kidney disease with dominant autosomal inheritance, Ehlers-Danlos syndrome, deficiency of alpha-1-antitrypsin and female gender.

The risk of aSAH in women varies with age of menarche, pregnancies and menopause. Women who have had an early menarche, women who are going through menopause and first-time mothers have a higher risk of aSAH than other women. There is no evidence of increased risk in pregnancy, childbirth and the postpartum period [91, 92].

People with a family history of SAH tend to be younger than those with sporadic SAH and have larger and multiple aneurysms [93, 94].

The risk of SAH increases in the presence of large unruptured aneurysms of the posterior circulation, particularly if symptomatic, and in the presence of previous SAH with or without untreated residual aneurysm, as demonstrated by the long-term follow-up of ISAT study [95].

The International Study of Unruptured Intracranial Aneurysms (ISUIA) study has provided information on the risk of aneurysm rupture in relation to the location and size of the aneurysm. It is higher for aneurysms of the anterior communicating artery and of the basilar artery apex as compared to the middle cerebral artery.

For aneurysms with a diameter of less than 7 mm, the risk of rupture is low (approximately 0.1 % per year) and progressively increases as the size increases. In the study, 75 % of unruptured aneurysms had a diameter of less than 1 cm. Larger aneurysms (>8 mm) longitudinally followed by MRI tend to grow more over time and therefore they are at greater risk of rupture [96].

Also, the morphological characteristics of the aneurysm, such as a bottle-shaped neck and the relationship between the size of the aneurysm and the afferent vessel, are associated with the risk of rupture [97–99] but it is still unclear how to decline such information for individual patients to predict the risk of SAH.

A model has recently been developed derived from the analysis of six cohort studies, which has led to the creation of a map of the risk of rupture of an intracranial aneurysm [100]. The identified predictors of aneurysm rupture are age, high blood pressure, history of SAH, size of aneurysm, location of aneurysm and geographic area. They form the PHASES score which corresponds to a precise risk of rupture at 5 years [100]. According to this score, the risk of rupture of an aneurysm at 5 years varies from 0.25 % in patients below the age of 70 without vascular risk factors and with small aneurysms of the internal carotid artery (<7 mm) to over 15 % in individuals older than 70 with hypertension, a history of SAH and giant aneurysms (<20 mm) of the posterior circulation. In the Finnish and Japanese population, the risk increases by 6.3 and 8.2 times, respectively.

Some studies have demonstrated a connection between diet and SAH. In particular, high consumption of vegetables is associated with a reduced risk of aSAH [101].

The factors that precipitate the rupture of an aneurysm have not yet been clarified, but it is assumed that a sudden increase in transmural pressure can play a role. Some studies have found that activities such as physical exercise, sexual activity and stress precede the occurrence of aSAH in over 20 % of the cases; no triggers have been identified for the remaining 80 % of the cases [89, 102, 103].

The prognosis is severe and the death rate, although it has decreased by 17 % in recent decades thanks to improvements in the management of SAH in hospitalized patients [88, 104], still remains high, reaching 45 % at 30 days in some studies [105].

A meta-analysis of population studies showed a geographical variation in median mortality rate, significantly lower in Japan (27 %) as compared to Europe (44 %), the USA (32 %), Asia excluding Japan (38 %), Australia and New Zealand [88].

In population studies, 12–15 % of SAH patients die before reaching the hospital (at home or during transportation) and 25 % of the remaining patients die within the first two weeks after the event [88, 106].

More than a third of the survivors remain disabled and typically some cognitive deficits are observed, such as impaired memory, executive functioning and attention disorders that impact on daily life [83]. Cognitive deficits tend to improve during the first year; however, they persist in approximately 20 % of patients [107], leading to a low quality of life.

Nineteen percent of patients show residual disability in conditions of dependence 3–12 months after the event [88].

Deterioration of the clinical condition is due to the initial severity of the bleed and/or to the onset of complications such as re-bleeding and late ischemia.

Re-bleeding can also occur very early. In 15 % of cases early clinical deterioration is caused just by re-bleeding that occurred before the first CT scan or before the patient entered the hospital, and it accounts for pre-hospital mortality.

The risk of re-bleeding is higher in the first 24 h after the aSAH, with a peak in the first 6 h; [108] after the first day, the risk of untreated aSAH remains high in the following 4 weeks, reaching 30 % in a month and decreasing gradually from 1 to 2 % per day to around 3 % per year [109]. The prognosis after re-bleeding is severe, with 60 % mortality rate and residual disability in 30 % of cases [110].

Delayed cerebral ischemia occurs in about one third of cases, mainly in the first and second week after the aSAH (peak 4–12 days). About 25 % of patients die and 10 % are disabled.

Negative prognostic factors include advanced age, clinical severity at admission and the extent of bleeding visible on the CT scan.

Other negative prognostic factors are: ruptured aneurysms of the posterior circulation, large diameter aneurysm and more recently the Apo E lipoprotein allele ɛ4 [111, 112]. Clinical severity and in particular the level of consciousness are the most important prognostic factors.

Spontaneous SAH typically manifests with an unusually severe headache that is generally widespread and has a sudden onset. The so-called thunderclap headache is often described as “an explosion” because of the intensity and speed of onset of pain, with peak time from a fraction of a second to a few seconds in 75 % of cases [116]. However, what should raise the suspicion of SAH is not only the intensity of the pain but its mode of onset. The headache might be associated with an altered state of consciousness, focal neurological deficits and vomiting, but in a third of all cases, the headache is the only symptom. In up to 77 % of cases, the headache is accompanied by photophobia, nausea and vomiting [117]. These are not SAH-distinctive characteristics because they are present in about half of patients suffering from thunderclap headache which is not associated with SAH [116].

Impaired state of consciousness is frequent. It was observed in 53 % of 109 patients in a retrospective study [118] and in about two-thirds of 346 patients on arrival at the hospital in a prospective study; approximately half of them were in a coma [119]. The normal state of consciousness may be recovered or it may remain altered.

Rarely, the patient with SAH may present with an acute confusional state (1–2 % of cases) [120].

Focal neurologic deficits are detectable in 10 % of cases and are due either to the extension of the bleeding in the brain parenchyma or to focal cerebral ischemia which is secondary to acute vasoconstriction occurring immediately after rupture of the aneurysm.

The deficits vary in relation to the bleeding site. The complete or partial deficit of the third cranial nerve has a localizing value because it is the sign of a ruptured aneurysm typically located at the level of the internal carotid artery, at the origin of the posterior communicating artery. The deficit of the sixth cranial nerve, when present, has no localizing value.

A stiff neck is a common sign and is due to an inflammatory response to the blood present in the subarachnoid space. It does not appear immediately after the bleed but needs 3–12 h to occur and cannot be present in patients in deep coma and with negligible SAH [121]. Therefore, the non-accrual of neck stiffness does not exclude SAH.

Seizures on onset have been reported in 7 % of SAH patients [116, 122, 123] and are strongly indicative of SAH caused by aneurysm rupture, since they have not been described in patients with perimesencephalic SAH or in patients with thunderclap headache without bleeding [116].

It should be remembered that although the typical presentation has been described, SAH patients may show isolated headaches and a normal neurological examination. Presentation and severity of the headaches can vary, which may make diagnosis difficult [124].

Some studies have shown that 10–43 % of patients diagnosed with SAH presented a so-called “sentinel” headache from 2 to 8 weeks before the major bleed [125, 126]. Such headache is often milder than that which occurs with the major rupture and may persist for days; it may be associated with nausea and vomiting, but meningeal signs are rarely present. In the absence of a proper diagnosis, a major re-bleed is very frequent in the following days or weeks.

Therefore, it is necessary to maintain a high degree of clinical suspicion for the disease, which should be excluded by appropriate diagnostics, since a missed diagnosis is associated with mortality and disability rates at 1 year that are four times higher in patients with minimal or no neurological signs at the initial examination [124]. The necessary examinations must be carried out on all patients struck by an unusual headache with acute onset and/or unusual pain characteristics.

It should not be forgotten that SAH accounts for only about 1–3 % of all visits to the emergency room for acute headache [127].

The headache lasts 1–2 weeks, though sometimes it can last longer; the minimum duration of the headache is unknown [113].

SAH patients may present intraocular hemorrhages of all types: retinal, subhyaloid or vitreous. In a systematic review vitreous hemorrhage was observed in 13 % of aSAH patients and was more frequent in subjects with an impaired state of consciousness [128]. Intraocular hemorrhage is due to increased intracranial pressure that causes obstruction of the venous drainage of the central retinal vein when it crosses the optic nerve sheath. Linear or flame-shaped subhyaloid bleeds are visible in the vicinity of the optical disc and may extend into the vitreous to configure Terson’s syndrome. Patients may complain of a brown stain that obscures the view. The examination of the fundus oculi is essential in these patients.

SAH clinical severity is expressed by means of rating scales.

Pre-hospital phase management should include vital sign detection, assessment of the level of consciousness, including determination of the Glasgow Coma Scale (GCS) score, and a brief neurological assessment to detect focal neurological deficits by means of a formal scale in order to assess the severity of the symptoms. The patient must be stabilized, and, depending on the patient’s clinical condition, airway management must be performed.

Cardiac arrest occurs in 3 % of cases and it is important to perform cardiopulmonary resuscitation (CPR) because half of the survivors recover their independence [129].

A typical candidate for SAH diagnostic protocol is a patient presenting with severe headache of sudden onset. The headache is typically described by the patients as the worst of their lives, with a peak time of seconds and associated with nausea, vomiting and focal deficits. Most severe headaches have a benign cause; approximately 10–16 % are caused by a serious medical condition, including SAH. It is more difficult to decide when to be alarmed in subjects with a vague modification of the primary headache they suffer from and with a normal neurological examination.

Patients presenting with acute headache of sudden onset should undergo brain CT scan if any of these conditions are present: age >= 40, neck pain or stiffness, witnessed loss of consciousness, onset during coitus, thunderclap headache and neck stiffness at neurological examination. This decisional scheme has a 100 % sensitivity for diagnosis of SAH [130].

In case of suspected SAH, a brain CT scan without contrast is the first examination that must be performed, but CT scans have some limitations, related mainly to the time when the examination is carried out and the amount of bleeding. A CT scan guarantees a high level of sensitivity (95–99 %) in detecting blood in the subarachnoid space within the first few hours of onset of symptoms, i.e. in the first 24 h; such sensitivity progressively reduces as time goes by (over the following days) because of the degradation of the blood and its dilution secondary to the continuous circulation of cerebrospinal fluid [131–134].

Within the first 6 h from onset of headache in patients with suspected SAH, the sensitivity of a CT scan is 98.5 % (95 % CI 92.1–100 %) [135]. The predictive negative value of CT results is almost 100 % (95 % CI 99.5–100 %) if the exam is carried out with third-generation CT scans and interpreted by neuroradiologists or radiologists experienced in reading Brain CT scan images [134]. This result was confirmed by a recent retrospective study in non-academic centres (negative predictive value 99.9 %; 95 % CI 99.3–100.0 %) [136]. Over the 6 h following onset of headache, sensitivity is reduced to 90.0 % (95 % CI 76.3–97.2) [135].

MRI might have the same sensitivity as CT in detecting SAHs in the first two days after the event [38] but may be more sensitive in the following days when the hyperdensity on a CT scan is reduced. The use of sequences with fluid-attenuated inversion recovery (FLAIR) and SWI is particularly sensitive to paramagnetic substances such as haemoglobin degradation products [137].

If CT scan findings are negative and clinical suspicion is high, a lumbar puncture should be performed, taking into account that in only 3 % of patients with negative CT findings within 12 h of onset of symptoms, a lumbar puncture shows haemoglobin metabolites and an angiography confirms the presence of a ruptured aneurysm [132]. Account must also be taken in the negative predictive value of a third-generation CT carried out within 6 h of the onset of symptoms.

A lumbar puncture should be performed at least 6 h after the onset of symptoms and preferably 12 h afterwards. In fact, if the cerebrospinal fluid (CSF) is obtained early and contains blood, it is almost impossible to discriminate between blood actually present in the subarachnoid space and blood generated by a traumatic puncture.

If correct timing is observed, blood degradation products will be present only in the case of a hemorrhage; in particular there will be bilirubin resulting from fragmentation of erythrocytes in CSF [113, 138]. The three-tube test carried out in order to differentiate the presence of blood resulting from a traumatic puncture from the presence of blood in subarachnoid space is not reliable [139].

There are specific technical recommendations for CSF analysis when SAH is suspected, and the recommended method is spectrophotometry with quantitative analysis of bilirubin.

An increase of bilirubin in the CSF indicates the presence of blood in the subarachnoid space and excludes that it is due to a traumatic puncture. The increase in bilirubin is generally accompanied by the presence of oxyhaemoglobin; the isolated presence of oxyhaemoglobin is almost always an artefact but may occasionally occur in SAH [140].

The absence of oxyhaemoglobin and bilirubin does not support diagnosis of SAH. It is not advisable to proceed directly to a vascular imaging exam (CT angiography or MRA) before demonstrating the presence of SAH, as unruptured aneurysms with diameters of <5 mm could be detected, which are to be considered incidental findings [113].

A recent study has shown that CT scans and CT angiograms detect aneurysmal SAH in 99 % of cases [141] and lumbar puncture remains the correct approach.

Bleeding resulting in a ruptured aneurysm cannot be confined to the subarachnoid space and may extend to the brain parenchyma, the ventricular system and the subdural space.

An intraparenchymal hematoma has a localizing value for the ruptured aneurysm: frontal or fronto-basal hematomas in anterior communicating artery aneurysms, temporal area hematomas for posterior communicating aneurysms and hematomas in situ for middle cerebral artery aneurysm. The possibility that an intraparenchymal hematoma is secondary to the rupture of a cerebral aneurysm varies from 4 to 35 % of cases.

As previously mentioned, neurological severity on admission, age, amount of bleeding and endoventricular extension of bleeding are prognostic factors for outcome after SAH.

Among these factors, the neurological status of the patient on admission is the most important and can vary over time. In order to have standardized assessment, it is necessary to have a valid and reproducible assessment tool, having good correlation with prognosis and good interobserver agreement.

Three rating scales are basically used to assess SAH: the Hunt and Hess scale, the World Federation of Neurological Surgeons (WFNS) scale and the modified Fisher scale (see Appendix). These scales have been created for different reasons, none of which are strongly related to the prognosis.

The Hunt and Hess (HH) scale was created in 1968 to stratify patients in relation to the surgical risk [148] and is still widely used. It takes into account the level of consciousness, intensity of headache, neck stiffness and severity of focal deficits. The categories are not clearly defined, however, and the scale has low levels of reproducibility and validity [149, 150].

The World Federation of Neurological Surgeons scale was proposed in 1988 by a commission of the WFNS in an attempt to identify a valid and reproducible tool [151]. It is based on the Glasgow Coma Scale (GCS), which has the advantage of having good interobserver agreement [152–154], and a grade has been added to levels 13 and 14 to insert the presence or absence of focal neurological deficits. The scale is easy to use but the correlation between each grade and the prognosis is controversial, and the cut-offs within the scale are based on consensus and not on a formal assessment.

Another scale was later proposed, based solely on GCS: the Prognosis on Admission of Aneurysmal Subarachnoid Hemorrhage (PAASH) Scale [155]. This scale is divided into five categories and the cut-offs between them are different from the WFNS cut-offs. In fact, the PAASH cut-offs have been selected by calculating the point at which two consecutive categories correspond to a statistically different prognosis at 6 months. This scale has good internal and external validity [156].

In the HH, WFNS and PAASH scales, the higher the score, the worse the patient’s prognosis. A study which evaluated the prognostic accuracy of the three scales showed that the WFNS and PAASH scales have a good prognostic value. The PAASH scale seems to have a more linear and gradual correlation with the risk of poor prognosis in the higher categories than the WFNS scale, and every increase in grade relates to a poorer prognosis [156] (See Appendix).

The interobserver agreement is similar in the WFNS and PAASH scales (weighted kappa values: 0.60, 95 % CI 0.48–0.73, and 0.64, 95 % CI 0.49–0.79, respectively), while it is lower for the HH scale (weighed kappa values: 0.48, 95 % CI 0.36–0.59) [157].

The modified Fisher scale is based on CT findings and derives from the original Fisher scale. It was developed in order to better predict the risk of angiographic vasospasm. The correlation is linear: the higher the score, the higher the risk of angiographic vasospasm [158, 159] (See Appendix).

Another scale has recently been proposed to stratify the risk of cerebral ischemia: the VASOGRADE scale. It combines the WFNS scale and the Modified Fisher Scale (VASOGRADE green, grade 1 or 2 on the modified Fisher scale and WFNS 1 or 2; VASOGRADE yellow, grade 3 or 4 on Modified Fisher Scale and WFNS 1, 2, or 3; VASOGRADE red, WFNS 4 or 5) [160].

Emergency treatment includes airway management, close haemodynamic monitoring, support treatment, prevention and treatment of complications. This may require rapid transfer of the patient to a tertiary centre to secure the aneurysm.

The number of clinical interventions is limited in SAH patients, but it has been shown that by treating patients in specialized centres that have a large volume of patients and multidisciplinary teams, their prognosis improves [161, 162]. “Large volume” is intended as at least 35 cases per year, with best benefits found in centres that treat more than 60 cases/year; in general, the higher the number of cases treated, the better the prognosis [163, 164].

Patients who survive the early hours are struck by three major complications: re-bleeding, delayed ischemia and hydrocephalus. In addition, there may be systemic complications that have a negative impact on prognosis.