Introduction

A review of 51 studies describing 58 study populations in 21 countries, reported a range of incidence of subarachnoid hemorrhage (SAH) from 22.7 (95 % confidence interval [CI] 21.9–23.5) in Japan, 19.7 (95 % CI 18.1–21.3) in Finland, 4.2 (95 % CI 3.1–5.7) in South and Central America, and 9.1 (95 % CI 8.8–9.5) in the other regions per 100,000 person-years [1]. The review did not differentiate between aneurysmal and non-aneurysmal SAH. Most studies have focused on aneurysmal SAH [2–4]. Estimation of incidence of non-aneurysmal SAH has been difficult because of inability to differentiate between SAH related to undiagnosed aneurysms and true non-aneurysmal SAH. In a nationwide study, 40 % of 77,352 patients admitted with non-traumatic SAH underwent either endovascular or surgical treatment for intracranial aneurysms [5]. The rates of any treatment (neurosurgical or endovascular) in population-based SAH-patient studies using case ascertainment by chart review have ranged from 38 to 70 % [6–8]. The relatively low rate of treatment is presumably attributable to relatively high rates of non-aneurysmal SAH [9] and early mortality (within 24 h) among non-traumatic SAH [6]. We performed this study to determine the actual incidence of non-aneurysmal SAH in a population-based cohort using a standard definition and associated case fatality.

Methods

Stearns and Benton Counties are situated in central Minnesota with a total area of 1,751 square miles or 1,120,915 acres. Both counties contain 38 cities and 46 townships. As of the US 2010 census, there were 189,093 people, 71,653 households, and 46,396 families residing in both counties. The population density was 108 people per square mile (67/km2). The racial/ethnic makeup of both counties was 92.34 % White, 2.85 % African American, 0.33 % Native American, 1.8 % Asian, 0.03 % Pacific Islander, 1.06 % from other races, and 1.56 % from two or more races; 2.55 % of the population were Hispanic of any race. The median age was 33.7 years and 23.48 % were under the age of 18 years, 14.59 % from 18 to 24, 25.46 % from 25 to 44, 24.40 % from 45 to 64, and 12.05 % who were 65 years of age or older. St. Cloud Hospital is the only acute care hospital and certified Primary Stroke Center with 489 licensed beds for both Stearns and Benton Counties. This hospital is the only hospital that provides endovascular and vascular neurosurgical services within Central Minnesota, and all patients with SAH within the catchment area are admitted to St. Cloud Hospital.

Case Ascertainment

Patients admitted with SAH from June 1st, 2012 through June 30th, 2014 were identified through several overlapping sources. Procedure notes for all diagnostic and interventional endovascular procedures were reviewed within a prospective database nested within EPIC electronic medical record system. A list of patients admitted with primary or secondary diagnoses of SAH (ICD-9-CM) code 430 was retrieved from data entered in the Get With The Guidelines®-Stroke registry. Cerebral angiography is performed for all patients with SAH documented by CT scan or identified in CSF analysis. Patients in whom an intracranial aneurysm was not documented underwent a repeat cerebral angiogram within 7–10 days. Patients who had primary residence within the 11 zip codes included in Stearns and Benton Counties were included in further analysis.

Data Collection

Patients admitted to St. Cloud Hospital with the diagnosis of SAH were identified using the above-mentioned methodology. Within EPIC electronic medical record system, patient charts were reviewed for information regarding demographic and clinical characteristics, preexisting risk factors, Hunt and Hess Scale at admission and Fisher grade based on initial CT scan findings. Details regarding aneurysm location, type of aneurism obliteration procedure, and details of medical treatment given were also extracted. Length of hospital stay and functional status defined by modified Rankin scale (mRS) at discharge was also ascertained. All vital status was ascertained by review of in-hospital and clinic records or by direct contact with patient or family by the treating physician to determine mortality at 1 and 3 months. In the event, the patient’s vital status could not be traced, we used the Social Security Administration Death Master File for ascertainment of vital status.

Sensitivity Analysis

A list of patients admitted with primary diagnoses of SAH (ICD-9-CM) code 430 and who had primary residence within the 11 zip codes included in Stearns and Benton Counties was retrieved from Minnesota Hospital Association data from June 1st, 2012 to December 31st, 2013. The methodology of data collection by Minnesota Hospital Association has been previously described [10, 11]. A total of 13 patients with the aforementioned ICD-9-CM codes were admitted from study population of which 12 were admitted to St. Cloud Hospital during the same period of time providing a yield of 92.3 %.

Data Analysis

We calculated age-adjusted incidence rates (age-adjusted to the 2010 US Standard Population) for all SAH, and aneurysmal and non-aneurysmal SAHs. The number of SAHs within each of the 11 age and gender groups were ascertained. We calculated crude rates using population data with the counts for each of the 11 age/gender groups expressed as the number of SAH events per 100,000 population at risk [12]. The age-adjusted rate was calculated by multiplying each crude rate by the appropriate weight and summing the products. We compared the incidence of aneurysmal and non-aneurysmal SAHs among men and women using standardized incidence ratio (SIR) with 95 % CI.

We also report the case fatality rate and mortality rate within 1 and 3 months of SAH onset for all SAH, and aneurysmal and non-aneurysmal SAHs. We calculated age-specific SAH-related death rate (ASDR) per 100,000 population for each age group as follows: ASDR = deaths in age group/estimated population of that age group × 100,000. Each ASDR was then multiplied by the proportion of the standard population for the same age group. The age-specific results were summed to get the age-adjusted death rate for SAH-related deaths as follows: AADR = Summation of (ASDR X standard proportion) [13].

Results

Of the 18 identified SAH, 5 were true incident cases of non-aneurysmal SAH, and one SAH was related to intracranial dissection, in this population-based study. Mean age (±SD) at diagnosis was 59.8 years (±18.8 years) compared with aneurysmal SAH 56.3 years (±11.3 years p = 0.6).The crude incidence of non-aneurysmal SAH was 2.7 [95 % CI 0.3–5.1] per 100,000 person-years which was lower than aneurysmal SAH incidence of 7.1 [95 % CI 3.3–10.9] per 100,000 person-years. The age-adjusted incidence of non-aneurysmal SAH was 2.8 [95 % CI 2.7–2·9] per 100,000 person-years which was lower than aneurysmal SAH incidence of 7.2 [95 % CI 7.1–7.4] per 100,000 person-years (SIR 0.38, 95 % CI 0.1–0.8, p = 0.1). There was a significantly higher age-adjusted incidence of aneurysmal SAH (compared with non-aneurysmal SAH) among women (12.1 [95 % CI 12.0–12.3] per 100,000 person-years vs. 2.3 [95 % CI 2.2–3.0] per 100,000 person-years; SIR 5.26, 95 % CI 3.3–8.5, p < 0.001). The age-adjusted incidence of aneurysmal SAH was similar (compared with non-aneurysmal SAH) among men (2.2 [95 % CI 2.1–2.3] per 100,000 person-years vs. 3.2 [95 % CI 3.1–3.3] per 100,000 person-years; SIR 0.66, 95 % CI 0.3–1.6, p = 0.5).

At admission, 6 of 13 patients with aneurysmal SAH and 0 of 5 patients with non-aneurysmal SAH were categorized as poor grade (Hunt and Hess grade of 4 and 5), respectively. Eleven of 13 patients with aneurysmal SAH and 2 of 5 patients with non-aneurysmal SAH were categorized as Fisher grade of 3 or 4, respectively, based on initial CT scan. The distribution of blood was categorized as perimesencephalic in all of 5 patients with non-aneurysmal SAH versus diffuse in 13 patients with aneurysmal SAH. Angiographic cerebral vasospasm was documented in 4 patients and required endovascular treatment in 3 patients with aneurysmal SAH. No cerebral vasospasm in patients with non-aneurysmal SAH. Among survivors was identified with non-aneurysmal SAH, mild (mRS 1–2) and moderate (mRS 3–5) disability was seen in 3 and 2 patients, respectively, Among survivors with aneurysmal SAH, mild (mRS 1–2) and moderate (mRS 3–5) disability was seen in 5 and 4 patients, respectively. The 1-month case fatality rate of non-aneurysmal SAH was significantly lower compared with aneurysmal SAH (0.00 vs. 23.0 %, p = 0.2). The 3-month case fatality rate of non-aneurysmal SAH was significantly lower compared with aneurysmal SAH (0.0 vs. 30.8 %, p = 0.2). The 3-month mortality rate was higher among patients with aneurysmal SAH (4.3 per 100,000 persons) compared with non-aneurysmal SAH (0 per 100,000 persons). The age-adjusted case fatality rate at 3 months 4.4 and 0.0 per 100,000 persons for aneurysmal and non-aneurysmal SAH, respectively.

Discussion

The incidence of non-aneurysmal SAH was relatively high with the incidence of aneurysmal and non-aneurysmal SAH being similar among men. Although, the case fatality was quite low, the rate of disability of any severity was considerable. The frequency of non-aneurysmal SAH in recent studies is low and ranges from 11 to 19 % in data derived from tertiary care specialized centers [14–16]. The incidence estimates from such reports are often artifactually lower because of transferred patients and selective triages based on presence of aneurysms on noninvasive imaging results. Incidence studies pertaining to SAH are particularly prone to such biases because of high rates of transfer for aneurysm treatment [17, 18]. At tertiary care centers, approximately 50–70 % of the SAH patients are transferred from other hospitals [19, 20]. Therefore, the impact of skewed estimates can be considerable. We used a population-based design to avoid the biases introduced by selective triages and admission patterns. We also used serial angiographic criteria for confirmation of absence of intracranial aneurysm or other underlying etiology. Such methodology avoids lack of detection of small intracranial aneurysms seen with CT angiography [21] and delayed appearance of aneurysms [14]. There is a possibility that the relatively higher than expected incidence ratio of non-aneurysmal SAH maybe related to decreasing rates of SAH related to aneurysmal rupture because of increasing treatments provided at an unruptured state [22]. The incidence of aneurysmal and non-aneurysmal SAH was similar among men but the incidence of aneurysmal SAH was prominently higher than non-aneurysmal SAH among women. Although the incidence of non-aneurysmal SAH appeared similar among men and women, the higher incidence of aneurysmal SAH explained the higher incidence of all SAHs in women seen in our and other studies [1].

Our current understanding of management of SAH patients is based upon data generated from aneurysmal SAH patients [14]. Non-aneurysmal SAH patients have been excluded from several large clinical trials including randomized clinical trials of nimodipine [23–25], nicardipine [26], magnesium sulfate [27, 28], tirilazad mesylate [29–31], and clazosentan [32, 33]. We found no case fatality but 2 of 5 patients (40 %) had severe disability following non-aneurysmal SAH. The absence of any case fatality in our study maybe attributable to perimesencephalic location of SAH and/or absence of angiographic vasospasm requiring treatment in non-aneurysmal SAH patients. Death or severe disability can be seen in 20 % of the patients with non-aneurysmal SAH [15] with high incidence of in-hospital adverse events [14]. In a combined analysis of case series and 18 previous reports [34], the overall rate of vasospasm was 16.3 % (95 % CI 12.3–21.1) and cerebral ischemia was 6.8 % (95 % CI 5.0–9.1) in non-aneurysmal SAH patients. In patients with diffuse SAH, the rates of vasospasm and cerebral ischemia were 31.3 % (95 % CI 21.6–43.9) and 9.7 % (95 % CI 7.0–13.0), respectively, Therefore, exclusion of such patients from clinical trials and studies of therapeutic interventions prevents adequate reduction in death and disability associated with SAH [5].

Our largest limitation was small number of events which was a consequence of restricting the denominator population to two counties and one hospital. Such a restriction resulted in high accuracy of the diagnosis of non-aneurysmal SAH and avoidance of variations in management between hospitals. However, the precision of estimates is lower than that seen with a larger number of events. The relatively low precision of estimate due to large contribution of single events in the analysis can exaggerate the incidence and disability estimates associated with non-aneurysmal SAH. The small number of events also reduced our ability to detect differences in incidences between men and women and other subgroups. There was a difference in time periods between US census and case ascertainment but the estimated 1.0 and 2.0 % population increases between April 1, 2010 and July 1, 2013 in Stearns and Benton counties was too small to affect calculation of incidences [35]. Nonetheless, the high incidence and burden of non-aneurysmal SAH were highlighted by the results of our study. The unique predisposition of men to non-aneurysmal SAH requires further study.