Key PointsQuestion
How does hospitals’ likelihood of adopting stroke care certification differ between historically underserved and general communities, and how does it differ by communities’ degree of segregation?
Findings
In this cohort study including all general acute nonfederal hospitals in the continental US between 2009 and 2019, hospitals in low-income communities and rural hospitals were less likely to adopt any level of stroke care certification relative to urban hospitals. Black, racially segregated communities had a higher likelihood of adopting stroke certification than non–Black, racially segregated communities, but after taking into account hospital and population size, likelihood of stroke certification adoption for Black, racially segregated communities was lower than for their counterparts.
Meaning
Access to new stroke-certified hospitals for patients differs by community characteristics of segregation by income and race.
Importance
Stroke centers are associated with better outcomes. There is substantial literature surrounding disparities in stroke outcomes for underserved populations. However, the existing literature has focused primarily on discrimination at the individual or institutional level, and studies of structural discrimination in stroke care are scant.
Objective
To examine differences in hospitals’ likelihood of adopting stroke care certification between historically underserved and general communities.
Design, Setting, and Participants
This study combined a data set of hospital stroke certification from all general acute nonfederal hospitals in the continental US from January 1, 2009, to December 31, 2019, with national, hospital, and census data to define historically underserved communities by racial and ethnic composition, income distribution, and rurality. For all categories except rurality, communities were categorized by the composition and degree of segregation of each characteristic. Cox proportional hazard models were then estimated to compare the hazard of adopting stroke care certification between historically underserved and general communities, adjusting for population size and hospital bed capacity. Data were analyzed from June 2021 to April 2022.
Main Outcomes and Measures
Hospitals’ likelihood of adopting stroke care certification.
Results
A total of 4984 hospitals were included. From 2009 to 2019, the total number of hospitals with stroke certification grew from 961 to 1763. Hospitals serving Black, racially segregated communities had the highest hazard of adopting stroke care certification (hazard ratio [HR], 1.67; 95% CI, 1.41-1.97) in models not accounting for population size, but their hazard was 26% lower than among those serving non–Black, racially segregated communities (HR, 0.74; 95% CI, 0.62-0.89) in models controlling for population and hospital size. Adoption hazard was lower in low-income communities compared with high-income communities, regardless of their level of economic segregation, and rural hospitals were much less likely to adopt any level of stroke care certification relative to urban hospitals (HR, 0.43; 95% CI, 0.35-0.51).
Conclusions and Relevance
In this analysis of stroke certification adoption across acute care hospitals in the US from 2009 to 2019, hospitals in low-income and rural communities had a lower likelihood of receiving stroke certification than hospitals in general communities. Hospitals operating in Black, racially segregated communities had the highest likelihood of adopting stroke care, but because these communities had the largest population, patients in these communities had the lowest likelihood of access to stroke-certified hospitals when the model controlled for population size. These findings provide empirical evidence that the provision of acute neurological services is structurally inequitable across historically underserved communities.
While treatment innovations over the past 2 decades have reduced disability and death for patients with stroke,1-10 the benefits from these scientific advancements have not accrued evenly across different populations, especially for historically underserved groups. For example, Black patients are less likely to receive intravenous thrombolysis with alteplase and mechanical thrombectomy compared with White patients, and wealthy patients are about 1.5-fold more likely to receive thrombectomy compared with patients from the poorest zip codes.11-13 Black and low-income patients with stroke also have a much lower likelihood of receiving endovascular mechanical thrombectomy.14-18
While disparities in stroke outcomes are partially due to disparities in stroke incidence,19 another potential reason for these widening disparities is the built environment of health care supply and geographic distribution of services, which has contributed to access and treatment inequities for historically underserved populations.17,18 In 2004, the Joint Commission, in partnership with the American Heart Association (AHA), began efforts to standardize stroke care by offering to certify hospitals meeting their stroke center requirements; interested hospitals could seek certification at their own expense.20,21 Certified stroke centers are more likely to provide timely treatment,22,23 with decreased mortality24-27 and higher quality of care28 for patients with acute stroke. Currently, the US Centers for Medicare & Medicaid Services (CMS) uses several accreditation organizations to certify hospitals into different levels, ranging from acute stroke-ready hospitals (ASRHs) to primary stroke centers (PSCs), thrombectomy-capable stroke centers (TSCs), and comprehensive stroke centers (CSCs). However, cross-sectional studies reveal that certain populations have less access to advanced stroke care than others.29,30 Given that racial and ethnic minority, low-income, and rural patients already have higher baseline risk of stroke31-37 and experience greater stroke severity at onset,38-41 this disparity in access to care could be a double hit for patients with stroke in these communities.
The existing literature has focused primarily on discrimination at the individual or institutional level. Studies of structural discrimination—the ways societies foster discrimination through mutually reinforcing inequitable systems that may not be intentionally designed but still produce inequity—in stroke care are scant.42-48 There is some evidence that stroke centers are preferentially located in higher-income areas20; that hospitals in lower-income areas with lower profit margins are less likely to be certified49; and that nonurban areas with a higher proportion of American Indian, uninsured, or low-income residents tend to be located farther away from a CSC.50 However, little is known about how expansion of specialized stroke facilities differs by degrees of community segregation, an important factor in understanding structural discrimination.
This study specifically answers the call for increased research regarding structural discrimination and inequity44,51,52 by quantifying differences in stroke certification adoption between hospitals serving historically underserved and general communities, where historically underserved status is defined by the share of underserved populations and the degree of segregation.53,54 Developing a better understanding of structural disparities underlying stroke care differences could inform administrative and policy changes that affect the geographical distribution of care within the stroke care system.55 We hypothesize that communities with higher proportions of White and non-Hispanic residents and a higher degree of racial and ethnic segregation, as well as those with higher income and greater income inequality, are more likely to attain higher levels of stroke center certification over time, which could contribute to decreased access to advances in stroke care among historically underserved populations.
The study universe of this cohort study included all general acute nonfederal hospitals in the continental US between January 1, 2009, and December 31, 2019, that were reported either in stroke certification data, AHA annual surveys, or Healthcare Cost Report Information data set. Federal hospitals, including those run by the Veterans Administration, were excluded. We identified levels of stroke certification from January 2009 to December 2019 based on data from the CMS deeming authorities: the Joint Commission, DNV, Healthcare Facilities Accreditation Program, and Center for Improvement in Healthcare Quality. We supplemented the national certification programs’ data with stroke hospitals that were certified by state health departments between 2009 and 2013, which Uchino et al shared.56 We updated the supplemental state data to 2019 assuming hospitals did not lose certification from states. We used AHA annual surveys and Healthcare Cost Report Information data set to capture the remaining general acute hospitals that did not have stroke certification and obtain additional facility data for all hospitals. This study was approved through the UCSF Human Research Protection Program and followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
The primary community information to define the historically underserved status of each hospital’s hospital service area (HSA) was based on the US Census and the Social Determinants of Health Database from the Agency for Healthcare Research and Quality.57 An HSA is a collection of zip codes whose residents receive most of their hospitalizations from hospitals in that area.58
Defining Hospital Stroke Care Capacity
A hospital was labeled as an ASRH, PSC, TSC, or CSC based on data from the certification programs. Care level designations expanded over time, starting with PSC (since 2008), and expanding to CSC (since 2012, quarter 3), ASRH (since 2015, quarter 3), and TSC (since 2018).
Defining HSAs as Historically Underserved
Our definition of historically underserved communities was based on the National Institutes of Health US health disparity populations, which include Black, Hispanic, socioeconomically disadvantaged, and rural populations.59 We focused on 4 dimensions: race (Black vs non-Black), ethnicity (Hispanic vs non-Hispanic), income (low income vs high income), and rurality (rural vs urban). For the first 3 dimensions, we used both the level of composition (ie, the predominance of each group) and degree of segregation to identify historically underserved communities. For example, an area might have a small Black population but is highly segregated.
The eMethods in the Supplement contains details regarding the method used to identify share of historically underserved population and degree of segregation of each community. For each socioeconomic dimension, we categorized HSAs into 4 mutually exclusive categories. For the race dimension, the categories were predominantly non–Black, racially segregated (reference population); predominantly non–Black, racially integrated; predominantly Black, racially integrated; and predominantly Black, racially segregated. likewise for the ethnicity dimension. For the income dimension, the categories were predominantly high income, economically segregated (reference population); predominantly high income, economically integrated; predominantly low income, economically integrated; and predominantly low income, economically segregated.
Following prior literature on medical technology or service adoption, we estimated a hospital’s adoption of any stroke care (regardless of level) using a survival analysis framework.49,60 The study included 44 quarterly intervals (from January 1, 2009, to December 31, 2019) during which a hospital could acquire stroke certification, where the interval captured all hospitals that were already stroke-certified as of 2009 quarter 1. To estimate the relationship between a hospital’s HSA socioeconomic status and the hazard that it would obtain stroke care certification, we used Cox proportional hazard models.61,62
Our main model first estimated unadjusted hazard ratios (HRs) comparing the hazard of adopting stroke care across the 4 categories of HSAs for each dimension, without adjusting for area or hospital characteristics. In model 1A, we focused on racial composition; in model 1B, on ethnic composition; in model 1C, on income distribution; and in model 1D, on rurality.
As secondary analyses, we estimated adjusted HRs for each dimension, taking population and hospital size into account. Controlling for population size is important in our context because emergency department capacity constraints have been associated with poorer quality of care and outcomes for stroke,63,64 and other time-sensitive conditions, such as acute myocardial infarction,65-67 sepsis,68 and trauma.69,70 For example, the same-sized hospital covering 1000 vs 1 million residents means fewer residents have to compete for access. Similar logic applies to controlling for hospital bed capacity. While critically ill patients with stroke may be pushed to the front of the line for receiving care, other factors, such as intensive care unit beds, critical care physicians, nurses, equipment, and other shared resources that are commonly used for other conditions (eg, computed tomography scanners for trauma patients, complex medical patients, and patients with stroke), may be in short supply. In hospitals with capacity constraints, these resources are stretched thin, negatively affecting quality of care for patients with stroke.63,64 Models 2A to 2D allow us to examine the overall differences in per-capita stroke care capacity among the 4 types of communities. We sought to determine how much of the difference in geographic access estimated in model 2 could be explained by underlying hospital characteristics. Therefore, we also added individual hospitals’ organizational characteristics, such as hospital ownership (not-for-profit [reference population], for-profit, and government), teaching hospital status (defined as resident-to-bed ratio greater than 0.25), whether a hospital was part of a system, and mean occupancy rate, to model 3. Data were analyzed using Stata version 17 (StataCorp). Statistical significance was set at P < .05, and all P values were 2-tailed.
Figure 1 shows the growth of stroke-certified hospitals between 2009 and 2019. During this period, the total number of hospitals with stroke certification grew from 961 to 1763 (of 4948 hospitals). Most were PSCs, the first certification type introduced in 2008, increasing from 961 in 2009 to 1363 in 2019. CSC certification was introduced in the third quarter of 2012 and expanded from 61 centers in 2012 to 254 in 2019. Additionally, there were 18 and 45 TSC centers in 2018 and 2019, respectively, given that TSC designation was introduced in 2018. ASRH certification was introduced in the third quarter of 2015 to encourage certification of rural hospitals. By 2019, 101 hospitals had obtained ASRH certification, of which 27 were rural and 74 were urban.
Table 1 shows the total sample of 4984 hospitals, stratified according to levels of stroke certification. A total of 3390 hospitals (68.0%) served non–Black, racially integrated communities; 486 (9.8%) served non–Black, racially segregated communities; 610 (12.2%) served Black, racially integrated communities; and 498 (10.0%) served Black, segregated communities. For income distribution, 2252 hospitals (45.2%) were located in high-income, economically integrated HSAs; 477 (9.6%) in high-income, economically segregated HSAs; 1268 (25.4%) in low-income, economically integrated HSAs; and 987 (19.8%) in low-income, economically segregated HSAs. A total of 918 of 1124 government-owned hospitals (81.7%) did not have a stroke center compared with 1485 of 2863 not-for-profit hospitals (51.9%) and 508 of 823 for-profit hospitals (61.7%). The mean (SD) population size of each HSA was 300 824 (603 995), with 1941 hospitals (38.9%) designated as rural hospitals per CMS.
Figure 2 shows the estimated HRs and 95% CIs from models 1 and 2 (full results are reported in eTable 1 in the Supplement). Figure 2A captures the HRs across the 4 types of HSAs based on race. Without controlling for population and hospital size, hospitals in predominantly Black, racially segregated HSAs were 1.67-fold more likely to adopt stroke care of any level relative to predominantly non–Black, racially segregated HSAs (HR, 1.67; 95% CI, 1.41-1.97). However, Black, racially segregated HSAs tended to cluster in areas with larger population sizes (average population of 1.29 million) compared with non–Black, racially segregated HSAs (average population of 307 768). After adjusting for population and hospital bed size (Figure 2B), the likelihood of adopting stroke care among hospitals serving Black, racially segregated communities was significantly lower than among those serving non–Black, racially segregated communities (HR, 0.74; 95% CI, 0.62-0.89) (Table 2). In other words, on a per-capita basis, a hospital serving a predominantly Black, racially segregated community was 26% less likely to adopt stroke certification of any level than a hospital in a predominantly non–Black, racially segregated community. The other 3 types of communities had comparable population-adjusted adoption rates.
Along the ethnicity dimension, we observed the same pattern: hospitals in predominantly Hispanic, ethnically segregated HSAs (average population of 1.52 million) were 1.22-fold more likely to adopt stroke care than predominantly non–Hispanic, ethnically segregated HSAs (average population of 704 398; HR, 1.22; 95% CI, 1.01-1.47). However, after accounting for population, hospitals in predominantly non–Hispanic communities, regardless of ethnic segregation, did not have significant differences in the likelihood of stroke care adoption.
Along the income dimension, the results did not reverse direction between models 1 and 2, since income segregation did not cluster in large HSAs like racial and ethnic segregation did. In our main model, hospitals serving high-income areas, regardless of income segregation, had higher likelihoods of adopting any level of stroke care compared with hospitals serving low-income, economically integrated areas (HR, 0.23; 95% CI, 0.20-0.27) and low-income, economically segregated areas (HR, 0.29; 95% CI, 0.24-0.34). Finally, rural hospitals were much less likely to adopt any level of stroke care relative to urban hospitals (HR, 0.10; 95% CI, 0.09-0.12).
Table 2 also provides the results of model 3 (full results in eTable 1 in the Supplement), which controls for population and hospital capacity and also accounts for hospital-specific characteristics. Because hospital locations are not random (eg, government hospitals are more likely than for-profit hospitals to be in rural areas), comparisons between models 2 and 3 provide insight into how much of the adoption hazard differences across HSAs can be explained by the types of hospitals that choose to operate in specific types of communities. Model 3 shows that when we take into account hospitals’ organizational characteristics, hospitals in historically underserved HSAs continue to have a lower adoption hazard of stroke certification by race, income, and rurality dimensions.
We further stratified our analysis by urban and rural hospitals, since rural hospitals face different challenges—specifically, rural disadvantages stem less from the decision to certify than a lack of local hospitals to certify.71 eTable 2 in the Supplement shows results from urban hospitals were similar to Table 2, but rural hospitals serving high-income, economically segregated communities were 3-fold more likely to adopt stroke care capacity than low-income, economically segregated communities (HR, 3.03; 95% CI, 1.59-5.56) or integrated communities (HR, 3.13; 95% CI, 1.64-5.88).
Finally, we conducted a further sensitivity analysis that restricted our analysis to an alternative stroke care capacity definition, where a hospital was defined to have stroke care capacity if it received stroke certification from national certification programs (eTable 3 in the Supplement). Our conclusions remained similar, except that hospitals in HSAs with a high Hispanic population (regardless of degree of segregation) had a lower hazard of seeking stroke certification.
Our analysis of stroke certification across acute care hospitals in the US from 2009 to 2019 paints a complicated picture of historically underserved communities’ access to hospitals with stroke certification. Our main model shows that Black, racially segregated communities experienced the highest likelihood of adopting stroke care, as did high-income, economically segregated communities. Stratified analyses between rural and urban hospitals showed similar patterns. Our secondary analyses in model 2 with population adjustment suggest that, on a per-patient basis, access to stroke-certified hospitals is less available in Black, racially segregated communities (ie, a stroke-certified hospital’s potential patient population base is much larger in those communities than in non–Black, racially segregated communities). In other words, while patients in Black, racially segregated communities have easier geographic access to stroke care relative to other communities, they may not be able to actually use this specialty care owing to resource constraints, since the same level of stroke care supply must accommodate a much higher level of stroke care demand in those segregated communities. Model 3 is valuable to gauge whether such structural inequities disappear if we assume the distribution of hospitals is uniform across HSAs. The coefficient changes from model 1 to model 3 do not reflect a nonrobust result; rather, they identify important mechanisms, such as population distributions of communities by race, through which structural racism and discrimination could be masked.
Prior studies reveal seemingly conflicting results regarding stroke center access disparities for vulnerable patients. One study using a prospective, longitudinal national cohort did not identify racial disparities in access to primary stroke centers.72 The same group later showed that a higher proportion of non-White patients than White patients had access to a stroke center within 60 minutes.29 However, while that study reflected geographic access from a patient perspective (ie, distance to the nearest stroke center), it did not control for population size. Our study, therefore, builds on that important work by comparing models with and without controlling for population and hospital capacity so that we can see more clearly how likely communities are to adopt stroke certification. This is important since racial and ethnic minority groups tend to be crowded in urban cities with high population density.
Our study adds 2 important insights to the existing literature. First, we examine residential segregation, which is an important component of defining historically underserved communities and, to our knowledge, has not been studied in stroke. Second, we show another mechanism of structural disparity in addition to specialty services closures: for certain communities, such as low-income and rural communities, specialized services are not even an option to begin with.
To be sure, geographic access is not the only solution for improving access to care. One study in the surgical literature has shown, for example, that even when Black patients live closer to higher-quality hospitals, they tend to receive care at lower-quality hospitals.73 In addition, emergency medical systems transfer patterns and regionalization of stroke care systems have evolved significantly over the past decade, which may mitigate these findings. Nevertheless, disparities cannot be eliminated if hospitals or specialized services are not physically present in certain communities. The National Institute of Minority Health and Health Disparities has identified that access to health care services and technology may be a specific and intervenable mechanism by which historically underserved communities benefit differently from the general population.74
There are several potential implications of this work. Recent work by Himmelstein and Himmelstein75 shows that both Black-serving and Hispanic-serving hospitals have fewer resources than hospitals that are neither Black-serving nor Hispanic-serving. Our results suggest that it might be necessary to incentivize hospitals operating in underserved communities to seek stroke certification or to entice hospitals with higher propensity to adopt stroke care to operate in such communities so access at the per-patient level becomes more equitable. Identification of barriers to certification could help shed light on potential policy interventions. Prior literature supports the idea that increasing investment in hospital capital, which can stimulate economic conditions for hospitals to locate in these areas, could be a potential remedy.75-77 For economically underserved and rural communities, specific initiatives that encourage stroke certification, potentially with interim provision of mobile stroke units, may also be necessary.
Another potential real-world application of our findings is revision of stroke center definitional requirements or implementation of certificate-of-need regulations that purposefully include population-based equity measures for health disparity populations. State stroke legislation has been shown to be effective in increasing the number of stroke centers.56 This could be applied to policies involving telestroke or even mobile stroke units, with particular attention to the needs of historically underserved populations, including reforms to legal and regulatory changes regarding licensing, credentialing, reimbursement, and liability addressing this specific modality of delivering health care.78-80 Variations in state legislation and policy governing other health care services, such as freestanding emergency departments, have been associated with a marked difference in distribution of services across populations.81,82 At the federal level, CMS grants deeming authority to private organizations to certify stroke centers. Currently, these decisions are based solely on hospital capabilities; however, it might be prudent for these certification bodies to integrate community need as a factor in certification decisions. Other potential next steps are additional research on more efficient use of existing systems of care, models of networked stroke systems, and cost-utility analysis of certifying all hospitals vs other models.
This study has several important limitations. First, because there is no central repository for stroke certification data from all certification organizations and states, data collected are not uniform and will contain errors. However, this should only result in bias if there is a differential pattern in reporting such that hospitals serving historically underserved communities are missing at different rates than hospitals serving general communities, which we do not expect. In addition, our sensitivity analysis limiting hospitals to only those who received their stroke certification from private certification bodies resulted in similar findings (eTable 3 in the Supplement).
Second, while we controlled for overall population in parts of our analysis, we did not have granular information to control for differential demand for stroke service by subpopulations. Given that Black individuals are 50% more likely to have a stroke and that stroke incidence is also significantly higher in Hispanic and low-income patients,31,83-87 our estimated magnitude of disparity in stroke adoption rate between historically underserved and general HSAs is conservative.
Third, there is overlap between the historically underserved communities across the 4 dimensions. We estimated all models separately for each dimension because including all dimensions would result in an overcontrolled model, yielding regression results that do not reflect reality and are unhelpful for policy discussions. For example, the interpretation of the HR between a Black, racially segregated community and non–Black, racially segregated community from a model that included both racial and income dimensions would take on a meaning that would assume the 2 types of communities have the same income distribution, which is unrepresentative of reality. Similarly, performing regressions that assume hospitals in disparate communities are the same negates the reality that the hospitals that serve general communities do not exist in historically underserved communities. We did, however, include hospital characteristics in model 3.
We examined patterns of adoption of stroke certification in hospitals across communities in the US and found that hospitals in low-income (regardless of segregation status) and rural areas were much less likely to adopt stroke care compared with hospitals in high-income and urban communities, respectively. In addition, while hospitals operating in Black, racially segregated communities had the highest likelihood of adopting stroke care, access to stroke-certified hospitals was less available in these Black, racially segregated communities after adjusting for population size. Other literature has shown that stroke-certified hospitals provide higher-quality stroke care; our findings suggest that structural inequities in stroke care may be an important consideration in eliminating stroke disparities for vulnerable populations.
Accepted for Publication: April 28, 2022.
Published Online: June 27, 2022. doi:10.1001/jamaneurol.2022.1621
Corresponding Author: Renee Y. Hsia, MD, MSc, Department of Emergency Medicine, University of California, San Francisco, 1001 Potrero Ave, Box 1377, Bldg 5, Ste 6A, San Francisco, CA 94110 (renee.hsia@ucsf.edu).
Author Contributions: Dr Shen had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Shen, Hsia.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: All authors.
Critical revision of the manuscript for important intellectual content: Shen, Hsia.
Statistical analysis: Shen, Sarkar.
Obtained funding: Shen, Hsia.
Administrative, technical, or material support: Sarkar, Hsia.
Study supervision: Shen.
Conflict of Interest Disclosures: Drs Shen and Hsia have received grants from the National Institute of Aging and the National Heart, Lung, and Blood Institute. No other disclosures were reported.
Funding/Support: This project was supported by the Pilot Project Award from the National Bureau of Economic Research Center for Aging and Health Research funded by the National Institute on Aging grant number P30AG012810; as well as National Center for Advancing Translational Sciences, National Institutes of Health, through UCSF Clinical & Translational Science Institute grant number KL2 TR001870.
Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: We are indebted to all the contributors of data who made this project possible, including the Joint Commission, DNV, Healthcare Facilities Accreditation Program, Center for Improvement in Healthcare Quality, Ken Uchino, MD (Cleveland Clinic, Cleveland, Ohio), and Kori Zachrison, MD, MSc (Massachusetts General Hospital, Boston). We also thank Judy Hahn, PhD, MA (University of California, San Francisco), and Stefany Zagorov, BA (University of California, San Francisco), for their editorial assistance. Dr Hahn and Ms Zagorov were compensated for their work.
5.Powers
WJ, Rabinstein
AA, Ackerson
T,
et al. Guidelines for the early management of patients with acute ischemic stroke: 2019 update to the 2018 Guidelines for the Early Management of Acute Ischemic Stroke: a guideline for healthcare professionals from the American Heart Association/American Stroke Association.
Stroke. 2019;50(12):e344-e418. doi:
10.1161/STR.0000000000000211PubMedGoogle ScholarCrossref 6.Albers
GW, Lansberg
MG, Brown
S,
et al; AURORA Investigators. Assessment of optimal patient selection for endovascular thrombectomy beyond 6 hours after symptom onset: a pooled analysis of the AURORA database.
JAMA Neurol. 2021;78(9):1064-1071. doi:
10.1001/jamaneurol.2021.2319PubMedGoogle ScholarCrossref 7.Thomalla
G, Cheng
B, Ebinger
M,
et al; STIR and VISTA Imaging Investigators. DWI-FLAIR mismatch for the identification of patients with acute ischaemic stroke within 4·5 h of symptom onset (PRE-FLAIR): a multicentre observational study.
Lancet Neurol. 2011;10(11):978-986. doi:
10.1016/S1474-4422(11)70192-2PubMedGoogle ScholarCrossref 9.Campbell
BCV, Ma
H, Ringleb
PA,
et al; EXTEND, ECASS-4, and EPITHET Investigators. Extending thrombolysis to 4·5-9 h and wake-up stroke using perfusion imaging: a systematic review and meta-analysis of individual patient data.
Lancet. 2019;394(10193):139-147. doi:
10.1016/S0140-6736(19)31053-0PubMedGoogle ScholarCrossref 15.Esenwa
C, Lekoubou
A, Bishu
KG, Small
K, Liberman
A, Ovbiagele
B. Racial differences in mechanical thrombectomy utilization for ischemic stroke in the United States.
Ethn Dis. 2020;30(1):91-96. doi:
10.18865/ed.30.1.91PubMedGoogle ScholarCrossref 22.McKinney
JS, Cheng
JQ, Rybinnik
I, Kostis
JB; Myocardial Infarction Data Acquisition System (MIDAS 22) Study Group. Comprehensive stroke centers may be associated with improved survival in hemorrhagic stroke.
J Am Heart Assoc. 2015;4(5):e001448. doi:
10.1161/JAHA.114.001448PubMedGoogle ScholarCrossref 26.Lichtman
JH, Jones
SB, Leifheit-Limson
EC, Wang
Y, Goldstein
LB. 30-Day mortality and readmission after hemorrhagic stroke among Medicare beneficiaries in Joint Commission primary stroke center-certified and noncertified hospitals.
Stroke. 2011;42(12):3387-3391. doi:
10.1161/STROKEAHA.111.622613PubMedGoogle ScholarCrossref 28.Johnson
AM, Goldstein
LB, Bennett
P, O’Brien
EC, Rosamond
WD; investigators of the Registry of the North Carolina Stroke Care Collaborative. Compliance with acute stroke care quality measures in hospitals with and without primary stroke center certification: the North Carolina Stroke Care Collaborative.
J Am Heart Assoc. 2014;3(2):e000423. doi:
10.1161/JAHA.113.000423PubMedGoogle ScholarCrossref 39.Kerr
GD, Slavin
H, Clark
D, Coupar
F, Langhorne
P, Stott
DJ. Do vascular risk factors explain the association between socioeconomic status and stroke incidence: a meta-analysis.
Cerebrovasc Dis. 2011;31(1):57-63. doi:
10.1159/000320855PubMedGoogle ScholarCrossref 42.Havranek
EP, Mujahid
MS, Barr
DA,
et al; American Heart Association Council on Quality of Care and Outcomes Research, Council on Epidemiology and Prevention, Council on Cardiovascular and Stroke Nursing, Council on Lifestyle and Cardiometabolic Health, and Stroke Council. Social determinants of risk and outcomes for cardiovascular disease: a scientific statement from the American Heart Association.
Circulation. 2015;132(9):873-898. doi:
10.1161/CIR.0000000000000228PubMedGoogle ScholarCrossref 43.Krieger
N. Discrimination and health inequities. In: Berkman
LF, Kawachi
I, Glymour
MM, eds. Social Epidemiology. 2nd ed. Oxford University Press; 2014:1-96.
45.Cruz-Flores
S, Rabinstein
A, Biller
J,
et al; American Heart Association Stroke Council; Council on Cardiovascular Nursing; Council on Epidemiology and Prevention; Council on Quality of Care and Outcomes Research. Racial-ethnic disparities in stroke care: the American experience: a statement for healthcare professionals from the American Heart Association/American Stroke Association.
Stroke. 2011;42(7):2091-2116. doi:
10.1161/STR.0b013e3182213e24PubMedGoogle ScholarCrossref 48.Skolarus
LE, Sharrief
A, Gardener
H, Jenkins
C, Boden-Albala
B. Considerations in addressing social determinants of health to reduce racial/ethnic disparities in stroke outcomes in the United States.
Stroke. 2020;51(11):3433-3439. doi:
10.1161/STROKEAHA.120.030426PubMedGoogle ScholarCrossref 51.Churchwell
K, Elkind
MSV, Benjamin
RM,
et al; American Heart Association. Call to action: structural racism as a fundamental driver of health disparities: a Presidential Advisory from the American Heart Association.
Circulation. 2020;142(24):e454-e468. doi:
10.1161/CIR.0000000000000936PubMedGoogle ScholarCrossref 54. Designation of medically underserved populations and health professional shortage areas.
Fed Regist. 2008;73(41):11232-11281.
Google Scholar 60.Cox
DR. Regression models and life tables.
J R Stat Soc [Ser A]. 1972;34:187-220.
Google Scholar 69.Jo
S, Jeong
T, Jin
YH, Lee
JB, Yoon
J, Park
B. ED crowding is associated with inpatient mortality among critically ill patients admitted via the ED: post hoc analysis from a retrospective study.
Am J Emerg Med. 2015;33(12):1725-1731. doi:
10.1016/j.ajem.2015.08.004PubMedGoogle ScholarCrossref 70.Wu
D, Zhou
X, Ye
L, Gan
J, Zhang
M. Emergency department crowding and the performance of damage control resuscitation in major trauma patients with hemorrhagic shock.
Acad Emerg Med. 2015;22(8):915-921. doi:
10.1111/acem.12726PubMedGoogle ScholarCrossref 79.Kulcsar
M, Gilchrist
S, George
MG. Improving stroke outcomes in rural areas through telestroke programs: an examination of barriers, facilitators, and state policies.
Telemed J E Health. 2014;20(1):3-10. doi:
10.1089/tmj.2013.0048PubMedGoogle ScholarCrossref 81.Gutierrez
C, Lindor
RA, Baker
O, Cutler
D, Schuur
JD. State regulation of freestanding emergency departments varies widely, affecting location, growth, and services provided.
Health Aff (Millwood). 2016;35(10):1857-1866. doi:
10.1377/hlthaff.2016.0412PubMedGoogle ScholarCrossref 84.Howard
VJ. Reasons underlying racial differences in stroke incidence and mortality.
Stroke. 2013;44(6)(suppl 1):S126-S128.
PubMedGoogle Scholar 85.Ashley
C, Berry
SD. The association between race and stroke prevalence in a patient cohort in Mississippi.
Perspect Health Inf Manag. 2020;18(winter):1i.
PubMedGoogle Scholar