Stroke Legislation Impacts Distribution of Certified Stroke Centers in the United States
Background and Purpose—The number of certified primary stroke centers (PSCs) have increased dramatically during the past decade in the United States We aimed to understand the factors affecting PSC distribution in the United States, including the impact of state stroke legislation.
Methods—PSCs certified by national organization or state until December 2013 were searched from available databases. The proportion of PSC among short-term general hospitals in each state was calculated and factors affecting its distribution were analyzed.
Results—By the end of 2013, the proportion of PSC varied from 4% to 100% among the 50 states and District of Columbia. The 18 states that had legislation in designating stroke centers and regulating stroke triage had higher PSC percentages (median, 43%; range, 13%–100%) than the remaining states (median, 13%; range, 4%–75%; P<0.001). State stroke legislation, urbanization, state economic output, and larger hospital size independently increased the likelihood of a hospital to be stroke certified. From 2009 to 2013, states with stroke legislation had greater increase of PSC percentages when compared with the states without legislation (median increase, 16% versus 6%; P=0.0067). Among the 1505 stroke centers, 74% were certified by the Joint Commission, 20% by state, and 6% by other organizations. Stroke centers certified only by state were smaller in size by hospital bed count compared with those certified by the Joint Commission (P<0.001).
Conclusions—State stroke legislation, a generalizable intervention, increased the number of certified stroke centers in the United States, potentially improving accessibility of standardized care for patients with acute ischemic stroke.
In 2000, the Brain Attack Coalition recommended the establishment of primary stroke centers (PSCs) to improve acute stroke care.1 The Joint Commission (JC) began certifying PSCs in 2003. Other organizations, including Healthcare Facilities Accreditation Program and Det Norske Veritas, later developed similar certification programs. In addition, several states developed their own certification programs. The goal of this stroke center certification was to improve stroke care and draw patients with stroke into capable centers. Many states also implemented legislations or policies for emergency medical services (EMS) to route ambulances directly to certified PSCs so that standard acute stroke therapies could be delivered more reliably and rapidly. PSC designation and other efforts seemed to have improved the measurable processes of care such as thrombolysis utilization rates, as well as mortality in participating hospitals.2–4 However, as of 2012, >147 million Americans remained without assured access to hospitals certified to provide standard acute stroke care.5
Geographic disparities in stroke systems of care and mortality have been observed across the nation.5–8 Hospitals in the South and Midwest of United States had lower thrombolytic rates.9,10 The South had higher rates of stroke-related mortality.7 Better understanding of the cause for this disparity of stroke care is needed to develop methods to optimize care throughout the nation. Therefore, we studied the factors that affect the stroke center distribution in the United States. We hypothesized that state stroke legislation improved the number of stroke centers within that state, thereby improving access to acute stroke care.
The list and status of state department of health in implementing stroke center designation and stroke systems of care by December 31, 2013, were obtained from the state and Centers for Disease Control and Prevention’s Web site.11 The lists of PSCs certified by JC, Healthcare Facilities Accreditation Program, Det Norske Veritas, and states were publicly available online at http://www.qualitycheck.org, http://www.hfap.org, http://dnvglhealthcare.com, and state Web site.
The list and characteristics of hospitals were obtained from the 2008 edition of the American Hospital Association Annual Survey Database. This study included only those that were categorized as general hospitals with emergency departments. Rehabilitation, geography, pediatric, long-term care, and federal government hospitals (Veterans and military hospitals) were excluded. For hospitals with multiple sites, the entry in the American Hospital Association Database was used as the unit of 1 hospital. The proportion of PSC-certified hospitals among all hospitals was calculated. The Core Based Statistic Area assignment for each hospital was used as a marker of urbanization and was drawn from the American Hospital Association Database. Core-based statistical areas were defined by the US Office of Management and Budget as metropolitan division (a core with a population of at least 2.5 million with surrounding areas), metropolitan statistical area (urbanized area that has a population of at least 50 000 and surrounding areas), micropolitan statistical area (1 urban cluster that has a population of at least 10 000, but <50 000 and surrounding areas), and rural area.12 We used the region designation by the US Census Bureau as Northeast, Midwest, South, and West.13
The data were analyzed using JMP 10.0 (SAS Inc, Cary, NC). χ2 tests were used to compare proportions and Wilcoxon rank-sum tests were used to compare numeric variables. Logistic regression was used to identify factors independently associated with categorization of a hospital as a PSC. The US Census region (Northeast, South, West, and Midwest) and core-based statistical areas (metropolitan division, metropolitan statistical area, micropolitan statistical area, and rural area) were entered as categorical variables. The average gross domestic product of each state from 2005 to 2013,14 and average All Payers Per Capita State Estimates by State of Residence-Hospital Care (State Health Expenditure) from 2005 to 2009, was separately used to categorize the states into quartiles as an ordinal categorical variable.15 The states with the lowest gross domestic product or Healthcare Expenditure were in the first quartile. Additional variables in the model included the presence of state stroke legislation as a binary variable and logarithmic transformation of hospital bed number.
As of December 2013, there were 1505 certified PSCs in 1574 sites in the United States, representing 32% of 4640 short-term adult general hospitals with emergency departments. Among them, 1118 (74%) PSCs were certified by JC, 34 by Healthcare Facilities Accreditation Program, 53 by Det Norske Veritas, and 300 (20%) by state health departments. The number of PSC-certified hospitals varied widely across the states, ranging from 4% in Wyoming to 100% in Delaware, with an overall median of 25% (Table 1; Figure 1). Delaware, New Mexico, North Dakota, Vermont, and Wyoming had their first stroke centers certified after 2008. The Northeast census region, the region with the densest population in the United States, had the highest proportions of PSC with 59% of hospitals certified, followed by West 32%, South 30%, and 24% in the Midwest (P<0.001 by χ2).
Eighteen states established legislation in designating stroke centers and some of them implemented policies for emergency medical services to route patients with acute stroke to PSC-certified hospitals. The legislation status in each state is shown in Table 1. States with stroke legislation had higher PSC proportion than those states with no stroke legislation (Figure 1). The percentage of PSC certified hospitals in states with no stroke legislation ranged from 4% in Wyoming to 75% in DC, with a median of 13% (Table I in the online-only Data Supplement). In the states with stroke legislation, the percentage of PSCs ranged from 13% in North Dakota to 100% in Delaware, with a median of 43% (P<0.001; Table I in the online-only Data Supplement). When the 50 states and the District of Columbia were stratified into quartiles based on the PSC proportions, the states with stroke legislation mostly fell into the highest quartile (Figure 1). The majority of the states with PSC proportions <25% had no stroke legislation.
From 2009 to 2013, there was a continuous increase in the percentage of PSCs among all hospitals across United States (Figure 2A). States with stroke legislation had greater increase of PSC percentages than states without stroke legislation (median increase, 16% versus 6%; interquartile range [IQR], 11%–28% versus 3%–13%; P=0.0067 in Wilcoxon analysis; Figure 2B). States with stroke legislation also had greater increase of PSCs in absolute numbers than those states without stroke legislation (359 versus 298, data not shown). The increase in some states was limited by already high proportion of PSCs (ceiling effect). Massachusetts passed stroke legislation in 2004 and 97% hospitals were PSC certified by 2009, resulting in near zero increase from 2009 to 2013.
The likelihood of PSC certification is associated with the size of hospitals. Larger size hospitals were more likely to be PSC certified than smaller hospitals (P<0.001 for both bed numbers and admissions volume; Table 2). The PSC hospitals were larger than non-PSCs. The median size of the PSC hospitals was 276 beds (IQR, 176–411) and 13 945 annual admissions (IQR, 8544–20 887), compared with non-PSCs who had a median 61 beds (IQR, 25–124) and 1756 admissions (IQR, 726–4529) (P<0.001 for both bed numbers and admissions volume; Table 2). Within the states with stroke legislation, the size of hospitals certified as PSCs only by the state was smaller than those also certified by the JC. Median bed number in the 300 state-certified hospitals was 204 with 9586 annual admissions (IQR, 117–339 beds; 5749–16 249 admissions), whereas the median bed number in the 581 JC-certified PSCs in these states was 300 with 15 161 annual admissions (IQR, 202–436 beds; 9719–22 018 admissions; P<0.001 for both bed number and admission by Wilcoxon test). There was no difference in the size of JC-certified PSCs between states with stroke legislation and states without state legislation (median 291 beds per 15 088 admissions; IQR, 186–420 beds per 9164–22 046 admissions; P=0.286 and 0.700, respectively).
There was a higher concentration of PSCs in urbanized areas (Table 3). Metropolitan division has the highest proportions of PSCs, followed by metropolitan statistical area. Only 10% hospitals in the micropolitan statistical area were PSC certified and that number dropped to 4% in rural area (P<0.001 by χ2; Table 3).
In multivariable logistic regression analysis, the number of hospital beds, urbanization, gross domestic product, and state stroke legislation were independently associated with stroke center designation (Table 4). State health expenditure was not associated with PSC certification. The Midwest and South census region was associated with a lower likelihood of PSC certification. Although significant on univariable analysis, the Northeast census region was not independently associated with PSC certification after adjustment for the state legislation, hospital size, and urbanization (Table 4).
This study demonstrates the important role of state legislation in the stroke system of care in the United States. State-level stroke center designation, usually accompanied by immediate mandate or future plans to divert patients with acute stroke to designated centers, had a powerful impact on the number of PSCs, and presumably, the availability of acute stroke care that meets basic clinical standards defined by guidelines. The majority of the states with the highest percentage of PSCs among their hospitals had stroke legislation. On the contrary, the majority of the states with the lowest percentages of PSCs among their hospitals did not.
Among the states with stroke legislation, the lower proportion of PSCs in North Dakota and Kentucky may reflect, at least in part, the recent start of the programs. It took 3 to 5 years for the effect of stroke legislation to be observed. Connecticut, Florida, Massachusetts, New Jersey, and New York passed their stroke legislation in 2005 to 2007 and were the top states of PSC proportions. Delaware, a state with only 5 hospitals, had no stroke centers in 2009 when it passed state stroke legislation. All hospitals were PSCs by in 2013. Additional potential explanations for relatively low percentages in North Dakota and Kentucky could be because of lack of an emergency stroke triage policy, which may serve to entice hospitals to become PSCs and the relatively low number of large densely populated cities.
Other factors affecting PSC certification were hospital size, degree of urbanization, and geographic location. Larger hospitals are likely to have greater numbers of patients with stroke and personnel available to develop and maintain stroke certification. Sixty-eight percent of hospitals in Metropolitan Division areas were PSC certified, followed by Metropolitan statistical areas and Micropolitan areas. The higher percentage of PSC hospitals in these areas may relate to competition with other hospitals, the availability of specialty physicians, or prioritization of certifying bodies to more populated areas to maximize impact. These hypotheses should be tested in future studies. Rural hospitals had only 4% of hospitals that were PSCs, suggesting poor access to standard stroke care in these areas. This result was consistent with previous report of disparity in stroke care within the Northeastern United States in rural areas.16 However, a rural 25-bed hospital on an island in Massachusetts was PSC certified using acute telemedicine coverage, and a 243-bed hospital in rural Michigan is among the PSCs, suggesting that hospitals in rural locations or of small size may still become a PSC.
Our data showed an uneven distribution in the proportion of hospitals that were PSCs across the 4 regions as defined by US Census Bureau. Hospitals in the South and Midwest had significantly reduced adjusted odds of being a PSC than hospitals in the Western region of the country. Our findings are consistent with the known geographic variation in utilization of intravenous thrombolysis and stroke fatality rates across the United States.6–9 It is notable that in the Northeast region, Massachusetts, a small and relatively densely populated state in which 97% of hospitals were PSCs, lies next to New Hampshire and Vermont, states in which only 12% and 7% hospitals were PSCs. State stroke legislation in these regions may help improve the proportion of hospitals that are PSCs.
Our study showed that PSCs certified only by states were smaller than PSCs certified by JC in both number of beds and admissions. Hospital volume has been linked to stroke mortality.17 It needs to be further investigated whether the PSCs certified only by the state has the same outcome when compared with the PSCs certified by JC.
A limitation of this study is the precision of data on hospital characteristics and numbers. Some hospitals have several campuses. There are discrepancies in how these hospitals are categorized in the JC and the American Hospital Association databases. The American Hospital Association Annual Survey Database relies on the self-reported hospital type, numbers of bed, and other hospital characteristics. In addition, we excluded federal government hospitals because they serve a special population and may not serve the immediate geographic area for stroke emergencies. We did not identify any federal hospital that was certified as PSC by the JC or Healthcare Facilities Accreditation Program. Only 2 Veterans Administration hospitals in Florida were PSC certified by the state. Another limitation of this cross-sectional study was our focus on a few variables. Other factors that might affect the availability of medical resource, such as population density or age, are not in the scope of this study. The state health expenditure from 2010 to 2013 was not available at this time. This study also did not examine the various components of state legislation and policies. Policy varies among states on designating stroke centers and establishing emergency medical services transport algorithm. This study combined PSCs designated by both the states and the JC. Ongoing study will address the question whether these 2 types of PSCs generate the same outcome for acute stroke treatment. Implementation of a prehospital stroke triage policy may also affect outcome of stroke care.18
In summary, state legislation for stroke center designation and stroke system of care has powerful impact in improving the availability of stroke centers. Many hospitals and regions have the potential to provide standard stroke care if policies are implemented by the state health department. If we were to apply the percentages of PSCs stratified by hospital size derived from states with stroke legislation to the remaining states, we would expect to increase the stroke centers to >1850 hospitals. As emphasized by Larry Goldstein, legislative actions are powerful way to increase access to stroke care.19
Sources of Funding
Dr Man received grant funding from the American Heart Association Mentored Clinical and Population Research Program.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.114.008007/-/DC1.
- Received November 25, 2014.
- Revision received April 23, 2015.
- Accepted April 28, 2015.
- © 2015 American Heart Association, Inc.
- Alberts MJ,
- Latchaw RE,
- Selman WR,
- Shephard T,
- Hadley MN,
- Brass LM,
- et al
- Schwamm LH,
- Fonarow GC,
- Reeves MJ,
- Pan W,
- Frankel MR,
- Smith EE,
- et al
- Song S,
- Saver J.
- Gillum RF,
- Ingram DD.
- Lichtman JH,
- Allen NB,
- Wang Y,
- Watanabe E,
- Jones SB,
- Goldstein LB.
- Kleindorfer D,
- Xu Y,
- Moomaw CJ,
- Khatri P,
- Adeoye O,
- Hornung R.
- Allen NB,
- Kaltenbach L,
- Goldstein LB,
- Olson DM,
- Smith EE,
- Peterson ED,
- et al
- 11.↵Center for Disease Control and Prevention. A Summary of Primary Stroke Center Policy in the United States. http://www.cdc.gov/dhdsp/pubs/docs/primary_stroke_center_report.pdf. Accessed April 10, 2014.
- Spotila JT.
- 13.↵U.S. Census Bureau, U.S. Department of Commerce. Census Division and Census Regions. https://www.census.gov/geo/maps-data/maps/pdfs/reference/us_regdiv.pdf. Accessed April 10, 2014.
- 14.↵Bureau of Economic Analysis, U.S. Department of Commerce. GDP by State. http://www.bea.gov/newsreleases/regional/gdp_state/2014/xls/qgsp0814_current.xls. Accessed April 11, 2015.
- 15.↵Centers for Medicare & Medicaid Services. Health Expenditures by State of Residence, 1991–2009. http://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/NationalHealthExpendData/Downloads/res-tables.pdf. Accessed April 11, 2015.
- Gropen T,
- Magdon-Ismail Z,
- Day D,
- Melluzzo S,
- Schwamm LH
- Goldstein LB.