Blacks have higher mortality rates from aneurysmal subarachnoid hemorrhage (SAH) than Caucasians. The time to treatment for aneurysmal SAH has been found to correlate with mortality and outcome. Therefore, we examined racial differences in the time to treatment of aneurysmal SAH among patients from the Greater Cincinnati area.
We evaluated data from 439 adult aneurysmal SAH patients prospectively identified from May 1997 to August 2001 and July 2002 to March 2005. The primary outcome measure was time to treatment, defined as elapsed time from arrival in the emergency department to aneurysm treatment. A multivariable model was constructed to determine the role of potential variables, including race, on time to treatment for SAH.
In univariate analysis, Caucasian patients were significantly older than black patients (P < 0.0001) and were more likely to be male (P = 0.014), insured (P < 0.0001), and transferred from emergency departments of presentation to other hospitals (P < 0.0001). Black patients were more likely to have anterior circulation aneurysms (P = 0.009) and preexisting hypertension (P < 0.001). In univariate analysis, anterior circulation aneurysms showed a trend toward earlier treatment than posterior circulation aneurysms (P = 0.07). In multivariable models, race was not associated with time to treatment or case-fatality rate. Patients transferred from other facilities were treated more expeditiously than patients who presented directly to the emergency department (P = 0.003), and a history of diabetes mellitus was associated with delay in treatment (P = 0.05).
Race was not associated with time to treatment after aneurysmal SAH in the Greater Cincinnati area. Reducing the increased burden of SAH mortality among blacks must be addressed at the prevention stage.
Several studies have documented racial or ethnic differences in patterns of treatment for different disease entities (19). For example, blacks with myocardial infarction were found to be less likely to receive early reperfusion therapy or other medical therapies such as coronary artery angioplasty and bypass surgery (1,6–8,19). Some studies have found that blacks are less likely to receive tissue-type plasminogen activator for the treatment of acute ischemic stroke (10,14).
Subarachnoid hemorrhage (SAH) secondary to ruptured intracranial aneurysm represents a devastating event that is associated with a high rate of death and complications (26,28). Epidemiological studies have shown that blacks have 57 to 60% higher age-adjusted mortality rates from SAH than non-Hispanic Caucasians and nearly double the incidence rate of SAH (3,11,16,25,27).
Once a patient presents with SAH, steps must be taken to stabilize the patient and prevent further complications such as rebleeding (29). The International Cooperative Study on the Timing of Aneurysm Surgery found that, although early surgery significantly reduced the risk of rebleeding, no significant difference in the overall outcome was identified related to timing of surgery (13). Later studies, however, have shown that early definitive surgery for repair of cerebral aneurysms after SAH not only decreases short-term complications, such as rebleeding and vasospasm, but also improves patient outcomes (4,20,23,26,30). Given the higher mortality rates among blacks compared with Caucasians and the known decrease in mortality rates with earlier treatment, we examined the hypothesis that blacks were treated later than Caucasians after aneurysmal SAH.
This study used data from the Genetic and Environmental Risk Factors for Hemorrhagic Stroke study, an ongoing, population-based, case-control study in the region of Greater Cincinnati and northern Kentucky. All patients in the region who had a potential intracerebral hemorrhage or SAH were identified by surveillance of the area's 16 adult hospital emergency departments (ED) and through screening of hospital discharge diagnoses (International Statistical Classification of Diseases and Related Health Problems-9 code 430–438.9). We analyzed SAH cases from data collected between May 1997 and August 2001 and between July 2002 and March 2005. The gap in data collection between September 2001 and June 2002 was the result of a lapse in study funding.
Patients were eligible for inclusion in this study if they were at least 18 years of age, resided within a 50-mile radius of the University of Cincinnati, had no evidence of trauma or brain tumor as the cause of hemorrhage, and were not already included in the study because of a previous hemorrhage. The Genetic and Environmental Risk Factors for Hemorrhagic Stroke study was approved by the institutional review boards at all participating hospitals.
The Genetic and Environmental Risk Factors for Hemorrhagic Stroke study uses a definition for SAH that was adapted from the Classification of Cerebrovascular Diseases III (1): non-traumatic abrupt onset of severe headache or altered level of consciousness associated with blood in the subarachnoid space, as observed on computed tomographic (CT) scans, magnetic resonance imaging (MRI), or at autopsy, or a clinical history and examination consistent with SAH in which xanthochromia and increased red blood cells are found in the cerebrospinal fluid.
Trained nurse-abstractors recorded detailed data from the medical records of all identified patients, including time of symptom onset, arrival time in the ED where the patient initially sought treatment, initial CT scan time, time of diagnostic angiogram, and time of initiation of the first attempt to repair the aneurysm. The elapsed time from ED arrival to surgical clipping or endovascular coiling was our primary interest.
At the time the data were assembled (July 2005), 827 patients had been confirmed as having SAH, of whom 554 had aneurysms confirmed by angiogram. Of these 554 patients, 50 were excluded because they did not undergo surgery or coiling and 37 were excluded for other reasons, including direct admissions to the hospital, non-aneurysmal sources of the bleed, treatment performed at an institution outside our region, non-diagnostic angiogram, and lack of information from the ED of the transferring hospital. An additional 26 patients were excluded because of undocumented time of ED arrival or time of surgery. One Hispanic patient and one Asian patient were excluded to allow univariate and multivariate comparison between Caucasian and black patients. Thus, 439 patients are included in our analysis.
Demographic, Clinical, and Radiological Variables
In addition to the timing variables mentioned above, we used demographic data from the medical abstracts, such as age, sex, race (self-reported), and insurance status, as well as clinical variables that might affect the time to treatment of an ED patient diagnosed with SAH. The Glasgow Coma Scale was used as the index for severity of SAH. Fisher scale scores, Hunt and Hess scores, and World Federation of Neurological Surgeons scores were not routinely recorded. Ruptured and unruptured intracranial aneurysm locations were subclassified based on their location in the anterior or posterior circulation. Other factors considered in the analysis were performance of a lumbar puncture in the ED; treatment method; interfacility patient transfers; and comorbid conditions, including hypertension, coronary artery disease, diabetes mellitus, and congestive heart failure. In addition, hospitals were categorized as teaching hospitals if there was neurosurgery resident coverage at the facility.
The data were managed and analyzed using SAS versions 8.2 and 9.1 (SAS Institute, Cary, NC). Univariate association between the time to treatment and race, as well other risk factors of interest, was assessed using the two-sample t test. Time to treatment by race was analyzed in a multivariable model using analysis of covariance to adjust for various risk factors. All variables examined were included in the initial model and then backward eliminated in a stepwise fashion retaining variables with a P value of 0.10. Because our specific hypothesis involved the association with race, this variable was forced into the final model.
Of the 439 patients included in the analysis, 345 (78.6%) were Caucasian and 94 (21.4%) were black (Table 1). Caucasian patients were significantly older than black patients (P < 0.001) and were more likely to be male (P = 0.014) (Table 2). Caucasian patients were also more likely to be insured (P < 0.001) and more likely to have been transferred from the ED of presentation to other hospitals that were better equipped to handle SAH patients (P < 0.001). Blacks were more likely to have anterior circulation aneurysms (P = 0.009) and a preexisting diagnosis of hypertension (P = 0.001).
Univariate and multivariate analyses comparing Caucasian patients with black patients did not demonstrate any significant difference in treatment time based on race (Table 3). However, in the univariate analysis, there was a trend toward earlier treatment of anterior circulation aneurysms compared with posterior circulation aneurysms (P = 0.07), and patients who were transferred from other facilities were treated more expeditiously than patients who presented directly to the ED of the treating facility (P = 0.022). In the multivariate analysis of covariance model of ED arrival time to treatment time, hospital transfer predicted shorter treatment time (P = 0.003), and a history of diabetes mellitus was associated with a delay in treatment (P = 0.049). Aneurysm location and race were not significant predictors of time to treatment in the multivariable model (P = 0.074 and 0.274, respectively). There was an interaction between race and hospital transfer variables; however, this interaction was tested and found to have minimal significance. There were no racial differences noted in demographic or clinical characteristics of patients who were diagnosed with ruptured aneurysms but did not undergo aneurysm treatment when compared with those who did undergo endovascular or surgical treatment during hospitalization (P = 0.730).
Finally, we examined 30-day case-fatality rates. In multivariate analysis, a higher risk of death occurred with advancing age (for every 5 yr: odds ratio [OR], 1.2 [1.1–1.3]; P < 0.0001) or a previous history of hypertension (OR, 1.55 [1.08–2.21]; P = 0.016), and a lower risk of death was observed with higher Glasgow Coma Scale score (OR, 0.89 [0.85–0.92]; P < 0.0001) and surgical treatment (OR, 0.32 [0.20–0.49]; P < 0.0001). However, after adjusting for these significant risk factors, there was a strong trend toward a decreased risk of death among blacks compared with Caucasians (OR, 0.61 [0.37–1.01]; P = 0.056).
Variation in treatment has been found to segregate along racial lines in some medical conditions. Given the higher SAH mortality rate among blacks compared with Caucasians and that earlier treatment has been found to lower mortality, we sought to examine whether or not time to treatment differed by race. We demonstrate that a racial bias toward delayed treatment of blacks does not seem to explain the previously recognized excess burden of stroke among blacks in the Cincinnati area (16). It is reassuring to find that, in our region, the care of patients with SAH proceeds with the same urgency regardless of race.
Numerous studies have documented that more blacks die from SAH than Caucasians (mortality rate) (3,10,27). Because the incidence rate of SAH is also higher among blacks than Caucasians, we also examined case-fatality rates, which represent the rate of death among those who have had SAH. We found no significant difference in the case-fatality rates between Caucasians and blacks after accounting for age, sex, and stroke severity (data not shown). This suggests that the higher mortality rates previously reported among blacks is likely secondary to the higher incidence rate. We think our data demonstrate that, at least in the Greater Cincinnati region, there is no significant bias in treatment provided to blacks versus Caucasians with SAH, and that the higher mortality rates previously reported among blacks are secondary to higher incidence rates. These findings suggest that the increased burden of SAH among blacks must be addressed through prevention.
Previous studies have documented excess mortality from all stroke subtypes in blacks when compared with non-Hispanic Caucasians (3,10,27). Ayala et al. (3) examined age-standardized mortality rates for SAH, as well as ischemic stroke and intracerebral hemorrhage, and found higher death rates from SAH in all minority populations compared with non-Hispanic Caucasians. Death rates from SAH in blacks were 50% higher than the rates for non-Hispanic Caucasians. A similar study examined the racial distribution of SAH mortality in Los Angeles County and noted an age-adjusted SAH mortality rate of 3.7 in blacks compared with 1.9 in Caucasians (27). In those younger than 70 years of age, the SAH mortality rate among blacks was 2.2 times that of Caucasians and 1.8 times that of Hispanics and Asians. Case-fatality rates were not reported.
Despite increased mortality rates in blacks after SAH, few studies have investigated clinical factors or treatment outcomes that may account for racial differences in SAH mortality rates. Other studies have suggested that the excess mortality rates are caused by higher incidence of SAH in blacks (15,16,24). Rosen et al. (24) examined the impact of racial differences in demographics, clinical characteristics, in-hospital complications, and outcomes of patients with SAH using data from two randomized controlled trials previously conducted at academic medical centers in the United States. They did not find race to be a prognostic factor affecting outcome after aneurysmal SAH and felt that the excess mortality rate in blacks experiencing SAH is likely owing to a higher incidence of SAH in blacks. We sought to determine whether or not access-to-care issues and racial differences in time to treatment might impact outcome and mortality rates after aneurysmal SAH and have provided evidence that they do not.
The demographic information stratified by race did show a significant racial difference in the prevalence of hypertension (P = 0.001). Black patients were much more likely to have a previous history of hypertension. Hypertension and/or untreated hypertension have been shown to be a major risk factor for aneurysmal SAH (17,33). Kleinpeter and Lehr (17) found an association between hypertension and SAH of cerebral aneurysms in the entire patient group in a study of 141 consecutive patients with SAH (P < 0.0001). Preexisting hypertension has been associated with the development of cerebral infarction after aneurysmal SAH and has been shown to predict mortality after aneurysmal SAH (13,21). Hypertension may thereby affect outcome after aneurysmal SAH. Further studies are necessary to evaluate whether or not racial differences in SAH mortality may be attributable to differences in preexisting illnesses such as hypertension.
Racial differences in aneurysm location have been previously described but the reason for these differences remains unknown (24). We found that black patients were more likely to have anterior circulation aneurysms. Rosen et al. (24) similarly found that blacks and other racial minorities were more likely to have internal carotid artery aneurysms and that Caucasians were more likely to have posterior circulation aneurysms (24). Further studies are warranted to better understand the implications of the observed racial differences in aneurysm location.
Two factors were associated with treatment delays: a previous history of diabetes mellitus and initial evaluation in the ED of the treating hospital. To the best of our knowledge, previous treatment delays among patients with diabetes mellitus have not been reported.
Interestingly, hospital transfer predicted shorter treatment times in the multivariate analysis of covariance model of ED arrival to treatment time. This finding seems counterintuitive, as previous studies have shown that hospital transfers can be time-consuming and may even adversely affect patient outcome (32). It is possible that triage times in suburban hospital EDs may be lower than in urban, high-volume hospital EDs, which might explain why transferred patients were treated more expeditiously in this study (2,5).
Several limitations exist in this study. Our analysis examines “time to treatment” as the time between ED arrival and surgery. We did not consider the time from symptom onset to surgery, which has been found to differ by race (9,18). Times of symptom onset are often poorly documented, especially when the initial symptoms are not witnessed at the time they occurred. It is possible that patients with unknown onset times presented to the ED later than those for whom the onset time was known, which would increase the delay between onset and surgery. Our study did document specific onset times for 78% of the patients, which compares very favorably with other studies that examined symptom onset times (22). No racial differences in the racial distribution of patients with specific versus missing times or in time to presentation were observed (data not shown). Therefore, we do not think that the presence of unknown onset times adversely affected our analysis.
Excluded patients may have introduced additional biases into our analysis. Patients who died early from an aneurysmal SAH and did not have an angiogram to document their ruptured aneurysm were not included. However, because our hypothesis examines the time to treatment, patients who would never have received treatment could be excluded without affecting the analysis. In addition, a small percentage of patients may not have undergone cerebral angiograms secondary to medical comorbidities or contraindications to cerebral angiography. We do not currently have a measure of why patients did not undergo cerebral angiography, but no difference in the rate of treatment of intracranial aneurysm was observed (Table 2).
For the purposes of this study, we also accepted the categorization of different races under the assumption that race is a social construct rather than a biological determinant. We did not raise issues related to the biological basis of these categorical variables or the implication of a fixed racial or ethnic identity in every individual. Problems associated with the use of these racial or ethnic terms in the literature have been previously described (12,31).
In summary, time to treatment does not seem to explain racial differences in outcome after SAH. When examining racial differences in treatment after SAH, case-fatality rates rather than mortality rates should be considered. Based on our findings, reducing the increased burden of SAH mortality among blacks must be addressed at the prevention stage.
This study was supported by a grant from the National Institutes of Health, NINDS R-01 36695.
The authors have provided an important step in analyzing the “why” of racial disparities by focusing specifically on time to treatment as a potential cause of disparities and whether there is a difference in the case-fatality rate for black versus white patients. These two factors did not play a role in their analysis. Although retrospective in nature, this study provides clues that population-based analyses that have pointed to higher mortality rates from subarachnoid hemorrhage (SAH) in the black population may be related to differences in the incidence of the disease in the two populations rather than differences in care once hemorrhage has occurred. This should provide an impetus to focus more on preventative strategies regarding risk factors for aneurysm hemorrhage, such as hypertension, which can be modified to reduce aneurysm mortality.
Eden et al. present intriguing research regarding the latency to treatment for black patients versus white patients presenting with aneurysmal SAH. The authors begin with the premise that blacks have a higher mortality from SAH, and they hypothesize that this worse prognosis might be explained by a delay in treatment. The study demonstrates no difference in the latency to treatment, which was the primary endpoint. As the authors point out, there are several pitfalls and limitations associated with this research. These include potential biological/genetic differences between the races, consideration of common comorbidities among blacks versus whites, and sociological issues such as time from symptom onset to presentation. Because this study has excluded access to care as an explanation for higher mortality from SAH in blacks, future studies can focus on biological and environmental issues that were alluded to in the discussion.
H. Hunt Batjer
The authors have investigated the potential role of timing of treatment as an explanation for the increased mortality of aneurysmal SAH reported in the black population. The authors have used information from the extensive Genetic Environmental Risk Factors for Hemorrhagic Stroke database, which serves the population of the Greater Cincinnati, Ohio and northern Kentucky region. Over the period of the retrospective review, 827 cases of confirmed SAH were identified, of which 439 were included in the study after a significant number of cases were eliminated that did not fit into the study design.
The study demonstrates that the timing of surgical or endovascular intervention did not differ significantly between black and Caucasian patients with aneurysmal SAH, which suggests that other factors are important in determining mortality differences between the two races. The authors found that Caucasian patients were older and more often insured than black patients. They also report that Caucasian patients were more often transferred from the emergency department of presentation to a hospital better equipped to handle SAH patients. The authors do not state how they determined which hospitals were better qualified to manage these patients, but this difference could potentially be related to insurance status. Conversely, black patients were more likely to harbor anterior circulation aneurysms and to have a significantly higher incidence of preexisting hypertension.
Timing of surgical or endovascular treatment of ruptured intracranial aneurysms is a complex issue that is dependent on many factors. Although recent studies have demonstrated some benefits of early intervention, these decisions are influenced by a variety of issues that include, but are not limited to, the time of admission after SAH, the patient's medical and neurological condition, the availability of microsurgical or endovascular expertise, and the location and morphologic features of the aneurysm. It is certainly comforting to learn from this study that race alone did not influence the decision-making process with regard to the timing of intervention. Based on this study, other factors, such as a higher incidence of preexisting hypertension, a higher prevalence of aneurysmal SAH, or differences in aneurysm location among races should be investigated.
Daniel L. Barrow
In this report, the authors examined the hypothesis that the higher mortality rates from SAH that are observed for black patients might be caused by treatment delays. Among a group of 439 adult patients in the Greater Cincinnati, Ohio area, there were some significant differences between Caucasian and black patients: Caucasian patients tended to be older, male, insured, and transferred to tertiary care centers, whereas black patients tended to have hypertension and anterior circulation aneurysms. However, no significant differences were observed in treatment times based on race. It is reassuring to read that there is no racial inequality, physician bias, or inaccessibility of care affecting the management of black patients with ruptured aneurysms. Case-fatality rates were not significantly different between the two races, which leads the authors to conclude that the higher mortality rates among black patients are due to a higher overall incidence and perhaps some adverse effect of hypertension. This contribution will not alter the management of ruptured aneurysm patients, but it does provide some interesting social perspective.
Michael T. Lawton
San Francisco, California
The authors evaluated case-fatality rates of aneurysmal SAH among black and Caucasian patients in the Greater Cincinnati, Ohio area between 1997 and 2005. The results reveal no difference or, if anything, an advantage for blacks. The conclusion is that the study demonstrates that in this population, there is no bias in treatment between blacks and Caucasians. Higher mortality rates are secondary to an increased incidence of SAH but not to differences in management.
There is a problem with the premise of this report. The rationale for the study is the assumption that the death rate from SAH among blacks is higher than that among Caucasians. Unfortunately, the data to support that contention are historic and are based on population studies that investigated death rates before 1994 (1). The assumption that population mortality rates remained unchanged over the next 11 years is not scientifically accurate. Contemporaneous evaluation of population mortality rates and case-fatality rates is necessary. There is no scientific information in this study to suggest that there is an increased burden of SAH mortality among blacks, and it is quite possible that such a difference existed 15 years ago but has already been eliminated by changes in medical practice and access to health care.
Robert A. Solomon
New York, New York