Nonmodifiable Risk Factors

Modifiable Risk Factors

Nontraumatic Nonaneurysmal Subarachnoid Hemorrhage

One can broadly classify SAH into traumatic and nontraumatic etiologies. Trauma is the most common cause of SAH. SAH has been cited as occurring in up to 60% of traumatic brain injury patients, a population with an incidence of approximately 540 per 100,000 in the United States. The pattern of hemorrhage, associated injuries, and clinical history often make this diagnosis readily apparent.

Nontraumatic SAH may also occur in patients not harboring intracranial aneurysms. Perimesencephalic SAH is defined by a relatively distinct radiographic pattern, with hemorrhage centered anterior to the midbrain or pons, with or without extension of blood around the brainstem, into the suprasellar cistern, or into the proximal Sylvian fissures. A negative 4-vessel cerebral angiogram confirms the diagnosis. Nontraumatic SAH has an incidence rate of 0.5 persons per 100,000 in adults. These patients tend to less likely be female, hypertensive, or of older age than aSAH patients. Diffuse, angiographic negative SAH that does not fit the perimesencephalic distribution of blood is thought to be a distinct entity with an incidence rate approximately twice as high as perimesencephalic SAH, and a higher incidence of complications such as hydrocephalus and vasospasm as well as the need for cerebrospinal fluid shunting and frequency of poor outcomes.

Additional possibilities include arteriovenous malformations, vasculitis, tumor, cerebral artery dissection, rupture of a small superficial artery, coagulation disorder, sickle cell disease, rupture of an infundibulum, and pituitary apoplexy. However, no cause can be determined in 14% to 22% of nontraumatic SAH.

General Trends and Grading

Over the past 2 decades, mortality following SAH has decreased dramatically. Previous death rates consistently occurred in a range around 50%, whereas more recently they have been found in the 10% to 24% range. These reductions likely stem from advances in multiple aspects of SAH management, including improved diagnostic capabilities, aggressive neurocritical care management, and use of modern microsurgical and endovascular instruments and techniques.

The 3 SAH-specific scales most widely used for classifying clinical presentation of SAH patients are the Hunt and Hess (H&H) Scale, the WFNS Scale, and the Fisher Scale. The H&H Scale was designed to gauge surgical risk of admitted SAH patients by evaluating intensity of meningeal inflammatory reaction, severity of neurologic deficit, and level of arousal. Gradation of the scale is 1 to 5, with 1 representing a nearly asymptomatic state and 5 denoting a deep moribund coma. There is evidence that differences between each H&H grade may not correlate with a unique outcome, but dichotomizing the scale has demonstrated that a good H&H grade (1 to 3) predicts a better outcome than a high H&H grade (4 and 5). In a series of 230 patients, 19% of good grade H&H patients had an unfavorable outcome compared with 90% of poor grade patients. A later study showed that H&H grades of 4 and 5 had 4.87 times greater odds of unfavorable outcome compared with grades 1 to 3 (95% CI 2.57–9.21; P <.001).

The WFNS Scale, developed in 1988, is also a 5-grade system, but is based on the GCS and is designed to acknowledge the significance of a focal neurologic deficit. Higher grades on the WFNS Scale indicate worse clinical presentation of SAH. The predictive value of this scale has been repeatedly called into question because widely varying outcomes have been observed in patients presenting with the same grade. Some investigations have even failed to predict any difference in outcome among adjacent WFNS grades when assessing patients with Glasgow Outcome Scale (GOS) at 1 month after discharge. In contrast to this, a large study of approximately 3500 SAH cases evaluated patients with the GOS at 3 months following SAH and found that the likelihood ratio of a poor outcome varied linearly with increasing WFNS: WFNS grade 1 = 0.36, WFNS grade 2 = 0.61, WFNS grade 3 = 1.78, WFNS grade 4 = 2.47, and WFNS grade 5 = 5.22.

The Fisher Scale is a radiographically defined score primarily concerned with predicting cerebral vasospasm after SAH. A grade of 1 to 4 is assigned depending on the amount of blood visible on CT imaging and presence of intracerebral or intraventricular clot. Fisher grade 3 is associated with the highest incidence of clinical vasospasm. Although the scale has been used to predict outcome (for scores of 3 or 4, relative risk of poor outcome >4), it is not considered to be comprehensive enough to be used as a primary grading system for SAH.

Several other diagnostic factors on admission have been shown to correlate with outcome following SAH. Specific portions of the GCS can be strong predictors of outcome. In poor grade SAH patients (H&H Grades 4 and 5), an additional point on the GCS motor examination at admission predicted a 1.8-fold increase in the odds of achieving a favorable long-term outcome as defined by a mRS score of 3 or less (95% CI 1.4–2.3). At discharge, an additional point in the eye examination was associated with a 3.1-fold increase in favorable outcome (95% CI 1.8–5.4). In another study, pupillary reactivity at admission predicted a 6.44 increase in the odds of favorable outcome at 12 months ( P = .008) in a population of 204 poor grade patients.

The modified Rankin scale (mRS) is an instrument frequently used for outcome assessment following SAH, and the scale is often dichotomized such that scores of 0 to 3 represent a favorable outcome with functional independence, whereas scores of 4 to 6 report a poor outcome with loss of a patient’s functional independence. The GOS is a second important measure of outcome composed of 5 points that reflect the following states: death, persistent vegetative state, severe disability, moderate disability, and good recovery.

Rebleeding, Timing of Intervention, and Vasospasm

Rebleeding is the major cause of death in patients who survive the initial hemorrhage but do not undergo surgical intervention. In untreated SAH, the greatest risk of rebleeding occurs on the first day (4%), with a daily frequency of 1.5% until 13 days. By 2 weeks, the rebleed rate is 15% to 20%, and up to 50% by 6 months. The goal of surgical and endovascular treatment is to prevent this occurrence, and since the 1980s there has been a shift toward early intervention.

Although the timing of intervention is still a source of debate, there are substantial efforts being made to carry out early management protocols on account of broad-based support for their implementation garnered through favorable outcome data. In the pursuit of early intervention, there is a widely recognized trade-off between early surgical risk of operative mortality and the benefits it confers in terms of rebleeding prevention. No outcome difference has previously been found between intervention at 0 to 3 days after the original bleed versus 11 to 14 days, but outcomes were definitively worse in the 7- to 10-day interval. Subsequent studies and meta-analyses have argued that the benefits derived from reduction of rebleeding seem to outweigh the risks of early intervention.

In the current context, the rate of rebleeding is near 7% when pre-hospital events are excluded, although several studies refer to a 10% to 20% incidence of “ultra-early” rebleeds by taking into account events that occur before patients receive neurosurgical attention. Overall, rebleed events in the first day are associated with a drastically reduced chance of survival with functional independence at 3 months (mRS score, ≤4; OR 0.08; 95% CI 0.02–0.34).

Ischemic neurologic deterioration secondary to cerebral vasospasm represents another major cause of morbidity after SAH. On average, symptomatic vasospasm occurs in 20% to 30% of patients, but vasospasm can be identified by arteriogram in 30% to 70% of patients with SAH, resulting in observed infarction in 10% to 45% of patients. For compounds that attempt to mitigate the effects of vasospasm, class I evidence has been obtained only in demonstrating the beneficial effects of the calcium channel blocker nimodipine. Although this medication does not alter radiographic vasospasm, it does improve the odds of favorable outcome to 1.86:1 ( P <.005).

Age, Gender, and Race

Patient age is strongly associated with worse outcome following SAH. In a study of 409 patients undergoing craniotomy for SAH, the investigators found that increased age correlated with significantly worse outcome such that a patient older than 63 years was at a 30-times greater risk for a poor outcome (GOS scores 1–3) than a patient aged 43 to 52 years. Less drastically, a study of 98 patients treated for SAH demonstrated that those who were 65 years or older fared significantly worse than younger patients on mRS outcome measures (hazard ratio 6.6; 95% CI 1.8–24.1; P <.001). Despite a higher mortality rate for elderly SAH patients, Stachniak and colleagues determined that quality of life (QOL) scores appeared acceptable for elderly survivors of SAH, suggesting that surgery need not be ruled out as an option for this population.

Mortality rates for female SAH patients have been reported to be higher than those for males. In an analysis of national death certificate data of SAH (n = 27,334) from 1995 to 1998, women had a higher death rate compared with men following SAH (4.9% versus 3.1%; Rate Ratio = 1.58; 95% CI 1.54–1.62). Although there exists a significant difference in mortality among men and women, gender has not been found to predict severity of presentation, outcome, or survival following SAH. In a trial for high-dose intravenous nicardipine, after adjustment for age, no difference was observed between women and men in terms of favorable outcomes at 3 months as measured by the GOS (69.7% for women versus 73.4% for men, P = .243; n = 565 women and 320 men). Thus, some investigators conclude that higher death rates in women may simply be the result of higher SAH incidence among females.

Although there exist variations in incidence and presentation of SAH among racial groups, the same cannot be concluded about differences in outcome, as several studies have shown no relation between outcome and race. In a 1970 study of SAH patients (n = 319) selected from a 20% systematic sample of hospital veterans, there were no racial differences in terms of survival after SAH. A retrospective case series from a single-center study of patients (1971–1976) similarly reports that race was not associated with adverse events in surgery. Although these 2 studies were conducted before the establishment of current guidelines for early treatment of SAH, subsequent and more recent work corroborates their conclusions.

In an analysis of 107 patients prospectively identified from 2000 to 2006 in the Brain Attack Surveillance in Corpus Christi Project, no ethnic difference in outcome or discharge was found between whites and Mexican Americans. Likewise, a prospective/retrospective case series of patients undergoing craniotomy for clipping of ruptured aneurysms (n = 219, recruited from 1989 to 1994), demonstrated no racial difference in QOL score after clipping for SAH. Lastly, a retrospective study of prospectively collected data for a randomized-control trial of tirilazad in SAH patients (1991–1997) found no difference in 3-month outcome as measured by GOS. A retrospective case series of cranial surgery among Medicare beneficiaries, however, demonstrated that black SAH patients had a longer length of hospital stay than the average SAH patient (12.2 days versus average 9.6 days, P = .001). Interpretation of these results is inherently limited as only 5.7% of the study subjects were diagnosed with SAH (2526 SAH cases of 44,078 enrolled patients), while the stated results are for all cranial surgeries. The preponderance of studies that report a lack of association between race and outcome, coupled with the consistency of that observation before and after the shift to early treatment guidelines, would suggest that outcome following SAH is largely unaffected by race.

Other Prognostic Indicators

Hospitals with a high volume of SAH admissions generate better treatment outcomes for SAH than do low-volume hospitals. In a large-sample study that examined 16,399 hospitalizations for SAH from 18 states in the US, patients who were treated in hospitals that see a low volume of SAH had 1.4 times the odds of dying in the hospital (95% CI 1.2–1.6) as patients admitted to high-volume hospitals. Larger hospital size seems to be associated with lower mortality rates especially in patients who undergo aneurysm clipping ( P <.01). In New York state hospitals, those treatment centers that performed more than 30 craniotomies per year reported a 43% (95% CI 29%–57%) lower mortality rate for SAH patients compared with hospitals performing less surgery. Furthermore, country or continent has no bearing on outcome after SAH, and outcome is unaffected by weekend versus weekday admission.