Clinical Outcomes of Primary Subarachnoid Hemorrhage: An Exploratory Cohort Study from Sudan

Abstract

Background

Although Subarachnoid Hemorrhage (SAH) is an emergency condition, its epidemiology and prognosis remain poorly understood in Africa. We aim to explore the clinical presentations, outcomes, and potential mortality predictors of primary SAH patients within 3 weeks of hospitalization in a tertiary hospital in Sudan.

Methods

We prospectively studied 40 SAH patients over 5 months, with 3 weeks of follow-up for the symptomatology, signs, Glasgow coma scale (GCS), CT scan findings, and outcomes. The fatal outcome group was defined as dying within 3 weeks.

Results

The mean age was 53.5 years (SD, 6.9; range, 41–65), and 62.5% were women. One-third (30.0%) were smokers, 37.5% were hypertensive, two-thirds (62.5%) had elevated blood pressure on admission, 37.5% had >24 hours delayed presentation, and 15% had missed SAH diagnosis. The most common presenting symptoms were headache and neck pain/stiffness, while seizures were reported in 12.5%. Approximately one-quarter of patients (22.5%) had large-sized Computed Tomography scan hemorrhage, and 40.0% had moderate size. In-hospital mortality rate was 40.0% (16/40); and 87.5% of them passed away within the first week. Compared to survivors, fatal outcome patients had significantly higher rates of smoking (50.0%), hypertension (68.8%), elevated presenting blood pressure (93.8%), delayed diagnosis (56.2%), large hemorrhage (56.2%), lower GCS scores at presentation, and cerebral rebleeding (P < 0.05 for each). The primary causes of death were the direct effect of the primary hemorrhage (43.8%), rebleeding (31.3%), and delayed cerebral infarction (12.5%).

Conclusions

SAH is associated with a high in-hospital mortality rate in this cohort of Sudanese SAH patients due to modifiable factors such as delayed diagnosis, hypertension, and smoking. Strategies toward minimizing these factors are recommended.

Introduction

Subarachnoid hemorrhage (SAH) is a stroke associated with high mortality and morbidity, especially if the diagnosis is missed. SAH contributes to 5–6% of all strokes. ¹ The overall incidence of SAH is estimated to be approximately nine per 100,000 person-years, affecting women more than men. ² The emergency physician needs to have a high index of suspicion for SAH possible diagnosis. ³ Different risk factors were associated with increased SAH incidence, including ruptured saccular aneurysms, arteriovenous malformations (AVM) trauma, familial risk, hypertension, smoking, alcohol consumption, anticoagulant therapy, and impaired coagulation mechanisms or bleeding into various tumors.^4-7 Other rare risk factors include polycystic kidney disease, female gender, and sympathomimetic drugs such as cocaine abuse.^2,7-9

Subarachnoid hemorrhage symptoms are characterized by a sudden onset and could include severe thunderclap headache, loss of consciousness, neck pain or stiffness, nausea and vomiting, seizure, photophobia, double vision, blind spots, or temporary vision loss in one eye.^10,11 SAH can be devastating with high morbidity and mortality rates. ¹¹ The presence of free blood can lead to cerebral spasm, one of the most common complications for patients who suffer from SAH.^12,13 Rebleeding is another major complication of aneurysmal SAH associated with high mortality rates, mainly in the first 12 hours after the hemorrhage.^7,14,15 Increased Intracranial Pressure (ICP) after SAH may also lead to brain tissue crushing against the bony skull and herniation. ¹⁶ In addition, hydrocephalus, seizures, cardiac complications, and electrolyte disturbances are recognized complications of SAH.^7,11,17,18 These complications are associated with an increased possibility of poor outcomes among SAH patients. ¹¹ Therefore, early and proper diagnosis and management of SAH are recommended to achieve patients’ stability and deal with bleeding and SAH complications. ⁴

Non-contrast Computed Tomography (CT) scan of the brain was found to be of high sensitivity (98.7%) with a 95% confidence interval (CI) of 97.1%–99.4% and high specificity (.999, 95% CI, .993–1.0) within 6 hours of headache onset. ¹⁹ If the imaging is completed after the six-hour timeframe, the sensitivity of CT drops to 85.7%. ²⁰ Lumbar puncture is the following diagnostic tool that should be discussed with certain patients, such as those with indeterminate or negative non-contrast head CT scan, delayed presentation time to CT scan study, or severe anemia, and in the cases with limited CT scan interpretation.^20,21 Elevated red blood cell counts, presence of bilirubin, and xanthochromia in CSF are significant indicators of SAH. ²¹ Lastly, CT Angiography (CTA), Digital Subtraction Angiography (DSA), and further investigations could be utilized to characterize the hemorrhage’s cause and source. ²² High power MRI is not widely used as it simulates the standard angiography in the detection rate. ²³

Since 1933, there are more than 40 grading systems have been proposed for patients with cerebral aneurysms. Hunt and Hess, Fisher, the World Federation of Neurological Surgeons (WFNS) and its modified scales, and Glasgow Coma Scale (GCS) are the most used SAH grading scales.^24,25 In 1968, Hunt and Hess established a SAH severity scale based on the presentation symptoms.^24,25 After two decades, the Fisher scale was created in 1980 to classify SAH based on the appearance findings of the CT scan.^24,25 In 1988, the WFNS grading scale was developed to include the patient’s GCS and focal neurological deficits to scale the symptoms severity.^24,25 In 1998, Ogilvy and Carter ²⁶ created a comprehensive classification system to help predict outcomes and gauge therapy for SAH patients. In 2001, Claassen modified the Fisher grading scale, including the additive risk of intraventricular hemorrhage.^24,25 Recently, the WFNS CVD and T Committee and the Japan Neurosurgical Society proposed a modified WFNS (m-WFNS) scale, which depends on GCS scores with ignorance of the presence of neurologic deficits.^27-29 This scale was developed to avoid the substantial variability in applying the original WFNS scale attributed to the ambiguity in recognizing the presence of neurologic deficits.^27-29

Despite the improvements in diagnostic tools, grading scales, and interventional and medical treatments of SAH neurological and systemic complications, ³⁰ SAH is still associated with very high mortality and morbidity rates. ¹¹ Previous studies, including population-based studies, systematic reviews, and meta-analyses, reported a wide variation in case-fatality rates of SAH as ranging between 8% and up to 67%.^31-34 Lantigua et al. investigated the prevalence and risk factors of in-hospital mortality among 1200 SAH patients enrolled in the Columbia University SAH Outcomes Project. ³³ The authors reported in-hospital mortality of 18%. They indicated that the age, loss of consciousness, GCS score, large aneurysm size, Modified Fisher Scale score, and Acute Physiology and Chronic Health Evaluation II (APACHE II) physiologic sub-score at admission were risk factors of mortality. ³³ Importantly, all these studies were conducted in developed countries lacking studies from low-income developing regions such as Africa. Nieuwkamp et al. indicated regional differences in SAH case-fatality rates as they varied from 8.3% to 66.7% between studies and decreased .8% per year with 11.8% lower in Japan than it was in New Zealand, Australia, USA, and Europe. ³² In comparison, the lowest case-fatality rate was reported in Japan (26.7%). ³² Previous studies indicated that the incidence rate of SAH was more among African Americans than Caucasian Americans.^35,36

The data regarding SAH patients outside Western countries is scarce, especially in Africa. The limited available reports from African countries indicated higher mortality rates resulting from SAH than high-income developed countries. The reported 30-day mortality rate due to aneurysmal SAH in Kenya was 26.6%, while the SAH fatality rate increased to 44.4% in Nigeria.^37,38 Thus, this exploratory study aimed to assess the clinical presentations and 3-week in-hospital course and outcomes of SAH patients in a tertiary hospital in Sudan, an African low-income developing country with limited healthcare resources. Also, we aimed in this study to evaluate the potential predictors of in-hospital mortality associated with SAH patients in our cohort. Up to our knowledge, this study is the first to investigate SAH patients’ characteristics in Sudan.

Methods

Study Design and Setting

A prospective, cohort, hospital-based study was conducted over 5 months, between May 2013 and September 2013, including conducting direct interviews with the SAH patients, taking history, performing physical examinations, ordering radiological images, laboratory investigations, and investigating the hospitalization course and outcomes. The participants were followed up for 3 weeks, the minimum duration of Intensive Care Unit (ICU) hospitalization for SAH patients in our hospital. Most deaths due to SAH occur within the first 3 weeks, primarily due to rebleeding.^39-42 This study was carried out in Omdurman teaching hospital, the central hospital in Omdurman city, Khartoum state, Sudan, and the second-largest hospital in the country. The estimated population coverage in the catchment area is two million people. The emergency department receives 1200–1500 patients per day. The bed capacity is 460 beds for medical and surgical patients. There is a CT scan in the hospital around 500 m away from the emergency department.

Adult Sudanese patients with documented primary SAH during the study period, as mentioned early, were invited to participate in this study, and they were followed up for 3 weeks. Primary SAH was defined as spontaneous bleeding in the subarachnoid space that was not due to extension from subdural, intraventricular, or intraparenchymal hemorrhage.^43,44 Both patients with and those without a documented aneurysm were included in this study. Exclusion criteria include non-Sudanese patients, those under 18 years old, and patients with a history of previous SAH and/or vascular malformation. Also, those who refused to participate in this study or sign the consent were excluded from this study. Out of 53 screened SAH patients, the number of consecutive patients who fulfilled the inclusion and exclusion criteria during the allocated period and were included in this study was 40 patients.

This study was ethically approved and secured by the research ethics committees of Omdurman Teaching Hospital and Sudan Medical Specialization Board (SMSB 2013, Internal Medicine MD N28). Written informed consent was obtained from each patient or their legally authorized representative. This study was conducted following the 1975 Helsinki declaration, as revised in 2008 and later amendments or comparable ethical standards.

Management Protocol in the Hospital

The SAH patients were managed by neurology, neurosurgery, and ICU healthcare workers’ teams. The SAH management protocol in our hospital includes admission to the ICU; conducting a non-contrast head CT scan to confirm the diagnosis of SAH on admission; and initiation of aggressive resuscitation, including blood access placement, administration of parenteral fluids to achieve fluid balance, control of blood pressure (BP), and conducting mechanical ventilation and intubation if the patient becomes unable to protect the airway or develops respiratory distress. Also, nimodipine therapy was initiated on admission. Although the benefits of dexamethasone in SAH are still unproven and controversial, by convention and after getting the patients informed consents, we administered dexamethasone injections to decrease ICP according to some evidence in literature.^45,46 Some patients received mannitol and antiepileptic medications (AEDs) as needed.

An external ventricular drain (EVD) was performed in certain patients with poor clinical grades on admission, acute neurological deterioration, or progressive ventricular enlargement. This EVD was used to remove CSF or blood that can cause increased ICP. ⁴⁷ Patients with SAH associated with elevated ICP and brain shift causing secondary brain compression were not taken to the operative room for hematoma evacuation or aneurysm surgical clipping due to affordability issues. ²¹

Data Collection and Used Grading Scales

Demographic data, medical history, comorbidities, and clinical characteristics at SAH onset were obtained shortly after admission using direct communications with the patients and their family members and by revising the patients’ medical records. The demographic data, including age, gender, and smoking status, were collected. The symptoms, clinical signs, and vital signs of presentation, and detailed neurological examination findings at admission were documented. According to the National Institute of Health (NIH) categories, the BP was defined high when it is >140/100 on admission.

The neurological status of each patient was assessed using the GCS and m-WFNS grading scales.^27-29 The GCS score was calculated for each patient at the time of presentation, and then the m-WFNS scores were assumed and documented at the time of patient presentation. The GCS is the most universally recognized and accepted scale for grading neurological conditions, including SAH.^28,29 It measures the neurological function based on 3 axes, including performing eye movements, speaking, and moving the body. Thus, it requires 3 gradings and one summation to produce a total comprehensive GCS score for each patient. The GCS score can range from 3, representing an entirely unresponsive patient, to 15 indicating a fully responsive person. In our study, after calculating GCS scores, the patients were categorized into four grades, based on the breakpoints of 15; 14 and 11; 10 and 6; 5 and 3. ²⁹ We also utilized the m-WFNS, a new validated scale for categorizing patients with SAH into 5 grades based on the GCS scores and regardless of their neurological defects. ²⁷ The 5 grades included Grade I, which is the favorable grade with GCS of 15; Grade II defined by GCS of 14; Grade III equals 13 on GCS; Grade IV includes GCS scores of 7–12; and grade V which is the worst grade with GCS scores of 3–6 ²⁷.

The primary radiological method for SAH diagnosis in all patients was the non-contrast brain CT on admission. The radiological findings at the time of presentation were documented. The hemorrhage size was then graded according to the radiologist’s report and the patient non-contrast CT scan appearance as either small, moderate, or large hemorrhages based on Hyperdense SAH grading scale.^48,49 The previous researchers created 3 categories to describe the grades of hyperdensity of the basal cisterns and subarachnoid space with the thickness and extent of hemorrhage (Table 1).^48,49

Table 1.

Hyperdense subarachnoid hemorrhage (SAH) grading scale as modified from Mohsen et al. ⁴⁸

Hyperdensity Grade	Description
Small (Grade I)	A localized thin layer of blood (<2 mm) in the basal cisterns on one side or in one Sylvian fissure only, without diffuse deposition of blood in the subarachnoid space
Moderate (Grade II)	A localized thick layer of blood (≥2 mm) in the basal cisterns, Sylvian fissure, or interhemispheric fissure without diffuse deposition in the subarachnoid space; or a diffuse deposition without a thick layer
Large (Grade III)	Diffuse deposition of blood in the subarachnoid space and basal cisterns + basal cisterns filled with a thick layer of blood (≥2 mm) on at least one side

A daily review of each patient hospital course was conducted for 3 weeks to document their symptoms, clinical signs, vital signs, GCS, complications, events, radiological findings, clinical improvement or deterioration, and hospitalization outcomes.

Hospitalization Outcome Assessment

In-hospital mortality after 3 weeks of hospitalization was considered the primary outcome to allow for uniform evaluation of patients as this is the minimum hospital stay length for SAH patients in our hospital, and most SAH deaths occur within the first 3 weeks. Thus, after 3 weeks of hospitalization, living patients were named “survival outcome group,” while those deceased were named “fatal outcome group.” The patients' progress details were critically followed up during this period. The causes of death were divided into 5 categories: direct effect of the primary hemorrhage, rebleeding, delayed cerebral ischemia (DCI), cerebral edema, and other medical complications (e.g., cardiac arrhythmia, multisystem organ failure, or pulmonary embolism, etc.). ³³ The rebleeding diagnosis was determined based on the appearance of new hemorrhage in the patient’s CT scan associated with their neurological status deterioration. ¹¹ The DCI was defined as cerebral infarction and neurological deterioration when the cause was suspected to be vasospasm after careful exclusion of other causes.^33,50

Data Analysis

Descriptive and inferential statistical analyses of data were performed using the IBM Statistical Package for the Social Sciences (SPSS), Windows version 24. The data were described using frequencies and percentages for categorical variables, while continuous variables, including age, were presented as mean (Standard Deviation, SD). Univariate analyses were conducted to determine the potential predictors of in-hospital mortality among SAH patients. Thus, a chi-square test or Fisher’s exact test was used to examine the differences in the categorical variables between the fatal and survival outcome groups. In contrast, a student T-test or one-way ANOVA test compared the means of continuous variables between the two groups. A P-value of less than .05 was considered significant.

Results

A total of 40 patients with primary SAH were included in this study. Their mean (SD) age was 53.45 (6.93) years and ranged from 41 to 65 years, with more than half of patients (n = 25, 62.5%) being women. After 3 weeks of hospitalization, 16 (40.0%) of patients died. Of those, 14 (87.5%) passed away in the first week, while two (12.5%) died during the second week; however, no one died during the third week. The participants’ symptoms, signs, presentation times, and hemorrhage grades are summarized in Table 2.

Table 2.

The clinical characteristics, time of presentation, and hemorrhage size on the non-contrast brain CT scan at the admission date of the participants, and the differences between survival and fatal outcomes groups.

Variables	Frequency (%)
	All cohort, n = 40	Outcome post three weeks of hospitalization
		Survival group, n = 24 (60%)	Fatal group, n = 16 (40%)
Smoking status
Current smoker	12 (30.0%)	4 (16.7)	8 (50.0%)
Non-smoker	28 (70.0)	20 (83.3)	8 (50.0%)
Hypertension comorbidity
Yes	15 (37.5)	4 (16.7)	11 (68.8)
No	25 (62.5)	20 (83.3)	5 (31.2)
Presenting symptoms on admission
Headache	36 (90.0)	23 (95.8)	13 (81.3)
Neck pain/stiffness	29 (72.5)	18 (75.0)	11 (68.8)
Fatigue/General weakness	20 (50.0)	8 (33.3)	12 (75.0)
Blurred vision	16 (40.0)	8 (33.3)	8 (50.0)
Photophobia	14 (35.0)	6 (25.0)	8 (50.0)
Hemiparesis	8 (20.0)	2 (8.3)	6 (37.5)
Irritability	7 (17.5)	0 (.0)	7 (43.8)
Diplopia	0 (.0)	0 (.0)	0 (.0)
Clinical signs on admission
Meningeal signs	26 (65.0)	12 (50.0)	14 (87.5)
Sensory or motor disturbances	11 (27.5)	2 (8.3)	9 (56.3)
Sub-hyaloid hemorrhage	6 (15.0)	1 (4.2)	5 (31.3)
Seizures	5 (12.5)	1 (4.2)	4 (25.0)
Muscles’ weakness	3 (7.5)	0 (.0)	3 (18.8)
Ptosis (unilateral)	1 (2.5)	0 (.0)	1 (6.3)
Third cranial nerve palsy	1 (2.5)	0 (.0)	1 (6.3)
Time of presentation
Within 24 hours of symptoms onset	25 (62.5)	18 (75.0)	7 (43.8)
>24 hours of symptoms onset	15 (37.5)	6 (25.0)	9 (56.2)
The hemorrhage grade on brain CT scan a
Small (Grade I)	15 (37.5)	15 (62.5)	0 (.0)
Moderate (Grade II)	16 (40.0)	9 (37.5)	7 (43.8)
Large (Grade III)	9 (22.5)	0 (.0)	9 (56.2)

^aThe classification of the hemorrhage grade was based on the Hyperdense subarachnoid hemorrhage grading scale as modified from Mohsen et al. ⁴⁸

Almost one-third of SAH patients (n = 12, 30.0%) were current smokers, and 15 (37.5%) participants had hypertension. However, at the time of presentation, 25 (62.5%) had elevated BP. The fatal outcome group had a higher percentage of smoking (n = 8, 50.0%) than the survival outcome group (n = 4, 16.7%) (P = 0.024). The past medical history of hypertension was reported in 11 (68.8%) patients of the fatal outcome group, while four (16.7%) patients in the survival outcome group had hypertension (P < 0.001). On admission, the BP was high in 15 (93.8%) of the fatal outcome group versus (vs.) 10 (41.7%) of the survival outcome group (P < 0.001).

Among all cohorts, the most frequently reported symptoms at presentation were headache (90.0%), neck pain/stiffness (72.5%), and fatigue/generalized weakness (50.0%), followed by blurred vision (40.0%) and photophobia (35.0%). In contrast, the least reported presenting symptoms were hemiparesis (20.0%) and irritability (17.5%), while no patients reported diplopia at the time of presentation. Although no statistically significant differences were noted in headache, neck pain/stiffness, blurred vision, and photophobia prevalence estimates at the time of admission between the fatal and survival outcome groups (P = 0.132, P = 0.665, P = 0.292, P = 0.104; respectively), there were statistically significant differences in other presenting symptoms. Compared to the survival outcome group, the fatal outcome patients reported higher prevalence estimates of fatigue/generalized weakness (33.3% vs 75.0%, P = 0.001), hemiparesis (8.3% vs 37.5%, P = 0.024), and irritability (zero vs 43.8%, P < 0.001); respectively (Table 2).

At the presentation time, the most common clinical signs were meningeal signs in the form of nuchal rigidity and positive Kernig’s sign (65.0%). About one-quarter of the SAH patients (27.5%) had sensory or motor disturbances and 15.0% had a sub-hyaloid hemorrhage. Seizures were reported in 5 patients (12.5%), while unilateral ptosis and cranial neuropathy in the form of third nerve palsy occurred in one patient (2.5%) for each. The fatal outcome group revealed significantly higher rates of SAH clinical signs on the day of admission than the survival outcome ones; these signs included the meningeal signs (87.5% vs 50.0%, P = 0.015); sensory or motor disturbances (56.3% vs 8.3%, P < 0.001); sub-hyaloid hemorrhage (31.5% vs 4.2%, P = 0.019); clinical seizures (25.0% vs 4.2%, P = 0.041), and muscles weakness (18.8% vs zero, P = 0.027); respectively. Of notice, all patients with presenting symptoms of muscle weakness, unilateral ptosis, and third cranial nerve palsy died (Table 2).

Among all cohorts, 25 (62.5%) patients presented within 24 hours of symptom onset, while the remaining 15 (37.5%) had delayed presentation to the hospital after 24 hours of symptoms onset. However, 6 out of the 15 delayed presenters (40.0%) visited the emergency room 1 day before with mild headaches and no abnormal clinical signs, and thus, they were sent home without admission, i.e., missed diagnosis. Approximately half of the fatal outcome patients (n = 7, 43.8%) presented within 24 hours of symptoms appearance, while 18 (75.0%) patients of the survival outcome group presented on day one (P = 0.046) (Table 2).

Regarding hemorrhagic grade/size on the primary non-contrast brain CT scan at the time of presentation, approximately one-quarter of the patients (22.5%) had a large-sized hemorrhage, 40% had a moderate size, and 37.5% had a small bleeding size. The hemorrhage was small in 62.5% of the survival group, moderate in 37.5%, and no patient among this group had a large hemorrhage. While 56.2% of the fatal outcome group had a large hemorrhage, 43.8% had a moderate hematoma, and no patients had a small SAH (P < 0.001) (Table 2).

Table 3 represents the differences in presenting grading scales’ scores between survival and fatal patients. At presentation; the GCS was 15/15 in half (50.0%) of patients, 11–14 in 15.0%, 6–10 in 27.5%, and 3–5 in 7.5% of patients. Also, the m-WFNS grading scale at the presentation of patients showed that half (50.0%) of them had grade I, and more than one-quarter (27.5%) scored grade IV. Grades II, III, and V were recorded among 7.5%, 5.0%, and 10.0% of patients, respectively. Further analysis revealed statistically significant differences between survival and fatal patients in grading scales’ scores distribution. Approximately 67.0% of the patients among the survival outcome group were fully responsive persons with GCS scores of 15/15, 20.8% had GCS scores of 11–14, 12.5% had 6–10 scores, and no patient had a GCS score of ≤5. In contrast to what has been found on admission in the survival outcome group, the GCS of 15/15 was observed in 25.0% of the fatal outcome group, 6.3% of fatal patients scored 11–14 GCS scores, 50.0% scored 6–10, and about 19.0% had GCS scores of 3–5 (P < 0.001). Also, the highest grades of m-WFNS (grades IV and V) were observed among 75.0% of the fatal outcome group, while among the survival outcome group, 12.5% had grade IV and no one had a grade V (P = 0.001).

Table 3.

The scores distribution of subarachnoid hemorrhage grading scales at the time of admission among cohort, and the scores differences between survival and fatal outcomes groups.

Variables	Frequency (%)			P-value
	All cohort, n = 40	Outcome post three weeks of hospitalization
		Survival group, n = 24 (60%)	Fatal group, n = 16 (40%)
Glasgow coma scale (GCS) on admission
GCS (3–5)	3 (7.5)	0 (0.0)	3 (18.8)	<0.001
GCS (6–10)	11 (27.5)	3 (12.5)	8 (50.0)
GCS (11–14)	6 (15.0)	5 (20.8)	1 (6.3)
GCS (15/15)	20 (50.0)	16 (66.7)	4 (25.0)
Modified World Federation of Neurological Surgeons (m-WFNS) grading scale at the time of presentation
Grade I	20 (50.0)	16 (66.7)	4 (25.0)	0.001
Grade II	3 (7.5)	3 (12.5)	0 (.0)
Grade III	2 (5.0)	2 (8.3)	0 (.0)
Grade IV	11 (27.5)	3 (12.5)	8 (50.0)
Grade V	4 (10.0)	0 (.0)	4 (25.0)

The size (grade) of hemorrhage was positively correlated with GCS scores among the study participants, as shown in Figure 1. All SAH patients with GCS scores of ≤5 had large (grade III) SAH, while only two patients with GCS scores ≥11 (7.7%) had such hemorrhage size (P = .003).

Figure 1.

Distribution of Glasgow coma scale (GCS) scores and hemorrhage size/grade among the study participants (n = 40) (P = 0.003).

The cerebral rebleeding was developed among 5 (31.3%) patients of the fatal outcome group, compared to two (8.4%) patients among the survival outcome group developed a such complication (P = 0.034). Although the fatal outcome group patients had a higher percentage of DCI development than survival outcome ones, the DCI was not statistically different between the two groups (12.5% vs 8.4%, respectively; P = 0.333). In contrast, cerebral edema was significantly more common among the survival outcome group (n = 6, 25.0%) than fatal outcome group (n = 1, 6.25%) (P = 0.011). Of notice, most of these events (80%) occur among late presenter patients, including five out of 6 rebleeding events (83.4%), 3 out of four DCI events (75.0%), and four out of seven cerebral edemata (57.1%) were developed among SAH presented after 24 hours of symptoms onset. Overall, among the fatal outcome group, the most common adjudicated primary causes of death were the direct effect of the primary hemorrhage (43.8%) and rebleeding (31.3%), followed by DCI from vasospasm (12.5%), cerebral edema (6.25%), and other medical complications (6.25%) (Figure 2).

Figure 2.

Primary adjudicated causes of death and neurological devastation lead to withdrawal of support. *Other medical complications indicated cardiac arrhythmia.

At the end of the follow-up period, 3 weeks later, the headache was still present in five out of the survival patients (20.8%) and neck pain/stiffness in 3 (12.5%). However, the hemiparesis, irritability, sensory or motor disturbances, and meningeal signs had equal proportions of 4.2% for each, as each symptom presented in one patient. At the end of the 3 weeks, no patient complained of blurred vision, photophobia, diplopia, seizures, ptosis, or cranial neuropathy. Also, at the end of the follow-up period, the GCS score was 15/15 in 21 (87.5%) patients of the survival outcome group, while 3 (12.5%) patients had GCS scores of 11–14, and no patients had GCS scores of ≤10. Besides, all the survival patients had normal BP at the end of the follow-up.

Discussion

In Africa, the epidemiology, risk factors, management, and prognosis of SAH remain poorly investigated in light of limited studies from this region of the world. In this contemporary, prospective, follow-up study of 40 patients diagnosed with primary SAH from Sudan, the in-hospital mortality rate within a 3-week of ICU hospitalization in a resource-poor country was as high as 40%, out of which 87.5% happened during the first-week post-SAH. This rate represents the upper end of the range of 18–50% previously reported post-SAH and stroke mortality rates in the literature.^{32,33,37,38,51-61} Most SAH patients were women and older adults, and about one-third were smokers and hypertensive. Smoking, history of hypertension, elevated BP at presentation, delayed or missed SAH diagnosis, large hemorrhage size, low GCS scores, high m-WFNS grades, and cerebral rebleeding were significant potential predictors for mortality among SAH patients (P < 0.05 for each). Among the presenting symptoms, patients with fatigue/general weakness, hemiparesis, irritability, meningeal signs, sensory or motor disturbances, sub-hyaloid hemorrhage, and muscles weakness at emergency presentation and those with SAH clinical signs were more likely to die within the first 3 weeks of ICU hospitalization.

Hypertension and smoking are identified as the most critical risk factors of SAH.^7,11,62,63 In our study, 37.5% were known to have hypertension before SAH, and 62.5% of patients presented with high BP. This finding is partially compatible with the observations of a previous study on 46 SAH patients that showed 63.7% of patients presented with elevated BP, and only half of them were known hypertensive. ⁶⁴ In a systematic review, hypertension was reported as a risk factor for SAH. ²⁴ In another Japanese study enrolling 247 adults, hypertension was a significant risk factor for aneurysmal SAH. ⁶⁵ Moreover, smoking, pre-existing hypertension, and elevated BP at presentation can destabilize cerebral aneurysms' growth and contribute to a fatal outcome which has been noticed in our results and reported by many previous studies.^66-68 A recent meta-analysis conducted by Etminan et al. showed a decline in global SAH incidence rates by 2.4% with every percentage point decrease in smoking prevalence and by 7.1% with every millimeter of mercury decrease in systolic BP. ⁶⁹

Headache, neck pain or stiffness, fatigue, and weakness were the most common presenting symptoms among our cohort in this study. This finding is concordant with the literature that reported the sudden headache as a cardinal symptom of SAH.^11,70-74 The SAH-associated headache is often diffuse and described by the patients as the worsening and most severe headache they have ever experienced in their life. A previous study of 364 patients with aneurysmal SAH reported the thunderclap headache as a primary warning symptom of SAH and described it as a severe, unusual, and sudden headache. ⁷⁰ Also, in a prospective study on 148 patients who had a new, sudden, and severe headache, 25% were found to have SAH. ⁷⁵ Neck pain, stiffness, and meningeal signs are other common presentations of SAH attributed to the inflammatory reactions to the hemorrhage in the subarachnoid space. The presence of neck pain or stiffness increases the likelihood of SAH, as pointed out in a meta-analysis describing the diagnostic accuracy of clinical history among SAH patients presenting to the ER. ⁷⁶ In that review, neck stiffness was identified as a negative predictor for SAH. ⁷⁶ Most SAH patients presented with headache and neck pain/stiffness in our study regardless of their outcomes. In the survival outcome group, the headache was present in 96% and neck pain/stiffness in 75% versus 81% and 69%, respectively, in the fatal outcome group. This may reflect the significant symptomatology of SAH and should raise general and medical awareness at the emergency level. Neck pain/stiffness usually takes hours to develop and may not be reported in those with minor hemorrhage and unconscious SAH patients.^71,77 Although neck pain and stiffness are widespread among our patients, their absence could not exclude SAH diagnosis, such as in patients with thunderclap headaches. However, headache may be presented as an isolated symptom or associated with other symptoms, including stiff neck, vomiting, motor weakness, blurred vision, photophobia, brief loss of consciousness, or focal neurological deficits (including cranial nerve palsies).^10,11

Seizures are not uncommon CNS complications of SAH; they could occur in up to 26% of SAH patients. ¹⁸ In our study, 5 patients (12.5%) experienced clinical seizures at presentation, and most of them (80.0%) had a fatal outcome. Hence, clinical seizures at the onset of SAH appears to be a significant predictor of poor outcome. Our findings are concordant with previous studies that indicate seizures occurrence in one of every 14 SAH patients (approximately 7%) and its prediction of poor outcome.^72,78-80 Several risk factors for early seizures development include young age, hypertension, ruptured intracranial aneurysms, aneurysm in a middle cerebral artery, poor neurological grade, hemorrhage thickness, rebleeding, and cerebral infarction.^7,11,81 There is a piece of low-quality evidence with weak recommendations for a short course of anticonvulsant prophylaxis use, such as in patients who had seizures after presentation. ³⁰

It is worth mentioning that the in-hospital mortality within the first 3 weeks of ICU hospitalization in this study was 40%. This mortality rate is almost similar to what has been found in a nationwide observational study in Europe and Asia, Monica stroke study, which reported a 30-day mortality of 42%. ⁸² Also, it approximates the mortality rate of SAH in Burkina Faso (37.3%), ⁷⁴ and Nigeria (44.4%), ³⁸ and the mortality rate reported by a nationwide Danish study (38%). ⁸³ However, this mortality rate was higher than reported in Kenya (26.6–30.4%),^37,84 Saudi Arabia (15%), ⁸⁵ Morocco (18%), ⁸⁶ Kashmir (20.38%), ⁸⁷ Martinique (24%), ⁸⁸ and developed countries.^33,89-92 In our study, the first week had the highest mortality rate, with 87.5% of fatal outcome group patients dying within this week. These results supported by the cognitive and functional outcome after SAH estimations study that 10–15% of patients die before reaching the hospital, approximately 25% of patients die within 24 hours, with or without medical attention, and hospitalized patients have an average mortality rate of 40% in the first month which reasonably matches our results. ⁹³ In that study, rebleeding as a significant complication carries a mortality rate of 80% during the first 10 days, which may explain the high mortality rate during week one in our study. ⁹³ The primary causes of death among our cohort were the direct effect of the primary hemorrhage and rebleeding.

There are variations in SAH incidence and mortality rates based on a regional level, and there are limited population-based data on SAH patients for most African populations. ³² Most available studies from Africa either discussed SAH specifically or, in general, with other stroke subtypes. ⁹⁴ For example, the available data from Nigeria, a neighboring African country, suggests that SAH represents 3.4–13.6% of the stroke subtypes.^38,94-96 The primary reported risk factor for SAH in Nigeria was hypertension, as observed in about 80% of SAH patients, while smoking has not been reported as a significant risk factor for SAH in Nigeria.^94,97 Also, the authors indicated that SAH is more prevalent among males and young people below the age of 40 years.^94,97 These data disagree with the generally accepted views about SAH risk factors. However, these data were collected and published before decades, thus do not reflect the current medical status of SAH in Nigeria as the vital paper on SAH from Nigeria was published in 1970.^94,97

A more recent study from Nigeria, prospectively conducted by Alkali et al. on 272 acute stroke patients between January 2010 and June 2012, showed SAH among 3.4% of the studies patients using brain CT and MRI scans. ³⁸ The case-fatality rate at 30 days was 18.8% for all stroke subtypes, and it was the highest for SAH patients with a 44.4% fatality rate. ³⁸ The authors reported hypertension, obesity, diabetes, hyperlipidemia, atrial fibrillation, and cigarette smoking as important stroke risk factors. ³⁸ However, most of these risk factors were discovered in many patients only after a stroke which is an alarming finding. ³⁸ The most recent study from Nigeria was conducted by Okon et al. between November 1, 2010, and October 31, 2011, and published on 2014. ⁹⁸ The overall SAH incidence rate was 4.1 (95% CI 2.6-6.3), and the SAH case-fatality rate was 6.51 per 100,000 per year. ⁹⁸ The incidence rate of SAH was significantly greater in women than in men. ⁹⁸

A previous study from another African country, Kenya, on 58 patients with aneurysmal SAH reported that the mean age of patients at presentation was 50.9 years and the most commonly affected age groups were 41–60 years. ⁸⁴ Also, the authors indicated a female predominance among patients with aneurysms of the posterior communicating artery and internal carotid artery. ⁸⁴ These findings are concordant with our results as the mean age of our patients was 53.4 years and ranged from 41 to 65 years, with more than half of the patients being female patients. The incidence of SAH is higher in women than men, which may be due to hormonal status and mainly develop in persons between 40 and 60 years of age. ¹¹

Of significance issue, undiagnosed hypertension is a significant problem and a critical hidden risk factor for SAH in Africa.^{94,96,97,99-101} This issue was evident in our results, as 62.5% of our patients had elevated BP at the time of presentation, while only 37.5% were known to be hypertensive. Also, SAH is often misdiagnosed, and most cases remain untreated due to a lack of neuroimaging and limited access to vascular neurosurgeons and neurosurgical facilities in African countries.^102,103 In our study, we could not classify SAH patients into aneurysmal and non-aneurysmal bleeding due to the lack of diagnostic tools in our institute. This fact highlights the limitation of diagnostic tools to identify an aneurysm and the lack of access to interventional neurovascular procedures in low and middle-income African countries compared to high-income countries. ¹⁰⁴

In this study, 62.5% of patients presented within 24 hours of symptom onset, while the remaining 37.5% of patients had mild symptoms that delayed presentation to the hospital. This finding is consistent with previous study findings to determine the association between presentation time and clinical outcome in SAH, approximately one-quarter of 721 patients presented after 24 hours of symptoms onset. ¹⁰⁵ However, the authors did not approve the presentation time as an independent risk factor for poor outcome. ¹⁰⁵ In our study, the delayed presentation was significantly more common among fatal patients than survival ones; this finding could be attributed to the higher rates of rebleeding, DCI, and cerebral edema events among late presenter patients. Several reviews of SAH cases showed that miss- and delayed-diagnoses could happen in 23–37% of cases with 6 days delay in SAH diagnosis.^3,106,107 Another study showed that in 60% of patients, the diagnosis was delayed for (1–2 days) due to mild symptoms and late presentation, but the remaining 40% had a very severe headache at the onset of SAH. ¹⁰⁸ This difference is likely due to the relatively small sample size in the latter study.

This study showed that 6 patients had actually presented to the emergency room 1 day before, with only mild headaches, no abnormal clinical signs, and sent home, that is, missed diagnosis. The missed or delayed diagnosis rate varies from 12% to 51%.^{106,107,109-114} This variation could be attributed to the differences in the methods and definitions to identify misdiagnoses due to the lack of a standard method to identify missed SAH. ¹¹⁵ In our study, the missed diagnosis rate represents 15.0% of the overall study group and 40.0% of those presented after 24 hours which is a high percentage. However, the missed diagnosis percentage was lower in another study, including 1507 patients, and reported 5.4% miss-diagnosed cases. ¹¹⁵ The missed SAH diagnosis and, thus, delay in treatment could explain the high mortality rates and risk of rebleeding in our study.^109,116 Administrative data could be a valuable tool for identifying misdiagnosis rates among SAH patients, as has been done previously for complications, ¹¹⁷ thus applying effective interventions to reduce these rates.

Our findings indicate that large hematoma on brain CT scan at the time of presentation is a potential negative predictor as all SAH patients with grade III hemorrhage, based on the Hyperdense SAH grading scale, were belonged to the fatal outcome group. Since the first description of aneurysmal SAH on CT scans, several reports showed that the increased amount of blood had been associated with higher mortality rates, vasospasm, and cerebral infarction.^48,49,118 This has impacted the limited available resources to care for such patients, which may contribute to the poor prognosis of patients. After 3 weeks, at the end of the follow-up period, no patient had visual complaints in the survival outcome group, while blurring vision, photophobia, and irritability were still present in up to 50% of the fatal outcome group patients. That indicates that long persistence of visual symptoms and irritability is likely associated with poor outcomes. This is due to their association with hemorrhage size, which again emerges as an essential negative predictor for the outcome.

On admission, almost two-thirds of the survival outcome group patients had a GCS score of 15/15 and scored grade I on m-WFNS, while only one-quarter of the fatal outcome group had these scores. Also, the hemorrhage size on brain CT scans was positively correlated with GCS scores among the study participants. Concentrating on the GCS is essential for the classification of severity and possible outcome. Overall, this is in keeping with the international Cooperative aneurysm study that reviewed 3521 patient driving data from 68 countries that stated that conscious level is one of the two most important factors to predict the outcome in which GCS is used universally scale. ⁹³ Moreover, another study, including four clinical trials that enrolled 3567 patients and addressed prognostic factors of SAH outcome, concluded that most of these non-modifiable factors are present on admission, but a substantial part is developing after admission. ¹¹⁹

This study’s strengths included its novelty as it is one of the first studies that obtained insights into SAH patients in Africa and examined the risk factors for in-hospital mortality in a developing African country. Thus, we tried to fill the gap of literature regarding a such disease outside Western countries. Also, this study focused on the factors and reasons that might explain high mortality rates among SAH patients in Sudan.

However, this study faced some limitations, such as being a single-center-based study conducted over a few months, and the follow-up period for outcome observation was relatively short, over 3 weeks. The financial issues and healthcare work responsibilities were the main reasons for short study conduction and follow-up periods. However, the notice of such patients' rapid and high in-hospital mortality rates was the foundation for this article’s idea to define the possible poor prognostic factors in the limited health setting of a resource-poor country with lack of SAH neurosurgical interventions either due to lack of service in the particular hospital or the financial cost. Regarding outcome labeling being a poor outcome, it would have been better to define another factor for a fatal outcome like hemiplegia, dysphasia, or adding disability score plus reassessment of the score after 3 weeks. Using hyperdense SAH grading scale for radiological classification of hemorrhage grade on brain CT scan is another limitation for this study as it had low sensitivity (71%) and specificity (64%), as well as poor positive and negative predictive capabilities (63%, 73%; respectively) to predict cerebral vasospasm. ⁴⁹ However, it is a good predictor for DCI after SAH as it occurred in 62.5% of patients with large (grade III) hemorrhage, 33–3% of those with moderate (grade II), and none of the other grades with fewer amounts of blood. ⁴⁸ Thus, more accurate radiological scales are recommended, such as Fisher and modified Fisher scales. The other essential limitations in this study included the small sample size and some patients being managed in the general ward due to limited resources.

Also, this study did not classify the included SAH patients into aneurysmal and non-aneurysmal SAH, and the study lacks data about aneurysm location and size due to limited resources in our center. CTA and DSA of the brain should be performed for each patient to look for abnormalities of the intracranial blood vessels such as an aneurysm, arteriovenous malformation, or arterial dissection. However, this study aimed to hint at the presentation, clinical course, and outcomes of SAH patients in an African low-income developing country with limited health services. In addition, we could not determine missed SAH patients who attended the emergency room but died without re-admission. Thus, both the numbers of missed SAH and the associated mortality rates may be underestimated. Also, we could not obtain data regarding other factors that may have explained the high mortality rates, such as physicians’ experiences, availability of consultants, and other resources such as diagnostic technologies. These and other potential explanatory mechanisms require further analyses.

Conclusions and Recommendations

This study provides the landmark for the epidemiology, clinical course, risk factors, management, and prognosis of SAH patients in an African low-income country. Our findings show a high in-hospital mortality rate of SAH in in this cohort of Sudanese patients as 40% of patients died within 3 weeks, and 87.5% died in the first week. Most SAH patients were women and older adults, and one-third were smokers and hypertensive. The potential risk factors for in-hospital mortality included smoking, history of hypertension, elevated BP at presentation, delayed or missed SAH diagnosis, large bleeding on the brain CT brain at admission, low GCS scores, high m-WFNS grades, cerebral rebleeding, and presence of SAH’s clinical signs. Most of these factors can be modifiable such as hypertension, smoking, delayed and missed diagnosis, and aneurysm rebleeding. Thus, applying effective interventions and prevention strategies directed toward minimizing smoking, cardiovascular diseases, brain injury, and cerebral rebleeding could be the best promise for reducing the incidence and mortality rates of SAH in African countries.

Considering the limited access and availability of diagnostic and treatment tools for aneurysmal SAH in the African neurosurgical sectors, providing accurate diagnostic tools and modern management modalities such as aneurysm repair and endovascular procedures are recommended. The prevalence rate of endovascular treatment is low in African centers (3.2%) compared to developed countries (45%) because of its high cost and shortage of qualified specialists.^120,121 Thus, we impulse stakeholders and local governments in African countries to invest in the acquisition of necessary equipment, the infrastructure of the neurosurgical centers, and the capacity building and to review the training curricula to buildup qualified endovascular specialties in order to achieve the goal of quality and timely intervention access for all SAH patients. These initiatives are not only to the African governments and local health policymakers but also to the World Health Organization (WHO) to achieve universal health coverage. Moreover, the SAH case morbidity should be given more attention in future studies because of its importance for the patient perspective and the high costs from the loss of productive life-years and the long-term care of disabled SAH patients. More studies are required to better comprehensively evaluate the epidemiology, clinical course, morbidity and mortality rates, and economic burden of SAH in Africa at regional and national levels, thus understanding the determinants for regional differences.

ORCID iDs

Abdel-Hameed Al-Mistarehi https://orcid.org/0000-0003-4713-8536

Yousef S. Khader https://orcid.org/0000-0002-7830-6857