Acute Stroke Care in the US
Results from 4 Pilot Prototypes of the Paul Coverdell National Acute Stroke Registry
Background and Purpose— The Paul Coverdell National Acute Stroke Registry is being developed to improve the quality of acute stroke care. This article describes key features of acute stroke care from 4 prototype registries in Georgia (Ga), Massachusetts (Mass), Michigan (Mich), and Ohio.
Methods— Each prototype developed its own sampling scheme to obtain a representative sample of hospitals. Acute stroke admissions were identified using prospective (Mass, Mich) or retrospective (Ga, Ohio) methods. All prototypes used a common set of case definitions and data elements. Weighted site-specific frequencies were generated for each outcome.
Results— A total of 6867 admissions from 98 hospitals were included; the majority were ischemic strokes (range, 52% to 70%) with transient ischemic attack and intracerebral hemorrhage comprising the bulk of the remainder. Between 19% and 26% of admissions were younger than age 60 years, and between 52% and 58% were female. Black subjects varied from 7.1% (Mich) to 30.6% (Ga). Between 20% and 25% of admissions arrived at the emergency department within 3 hours of onset. Treatment with recombinant tissue plasminogen activator (rtPA) was administered to between 3.0% (Ga) and 8.5% (Mass) of ischemic stroke admissions. Of 118 subjects treated with intravenous rtPA, <20% received it within 60 minutes of arrival. Compliance with secondary prevention practices was poorest for smoking cessation counseling and best for antithrombotics.
Conclusions— A minority of acute stroke patients are treated according to established guidelines. Quality improvement interventions, targeted primarily at the health care systems level, are needed to improve acute stroke care in the United States.
The central goal of the Paul Coverdell National Acute Stroke Registry (PCNASR) is to track delivery of care to hospitalized stroke patients and to guide and monitor improvements in the quality of acute stroke care. The registry’s goals, as well as recommendations concerning its scope and content, have been summarized elsewhere.1 Eight prototype registries were funded to pilot test sampling designs and data collection methods, as well as to test quality improvement interventions.
The aim of this first report is to describe the characteristics of acute stroke admissions and the in-hospital care they received across the 4 prototype registries (Georgia [Ga], Massachusetts [Mass], Michigan [Mich], and Ohio) that completed data collection by the end of 2002. The report includes data on the key prehospital features of acute stroke admissions, acute care treatments (including thrombolysis), and selective stroke care performance indicators related to in-hospital procedures, and secondary prevention measures at discharge. Subsequent analysis of data from these pilot registries, including an assessment of their completeness and data validity, will be critical in the design of the final registry model that will be used to monitor and improve the quality of acute stroke care at the state level.
Materials and Methods
Registry Design and Hospital Selection
Investigators from each prototype were required to develop their own sampling plan to select acute care hospitals that were representative of populations treated in their states. Overall, a total of 98 hospitals were included (12 from Mass, 16 from Mich, 36 from Ohio, and 34 from Ga). All prototypes purposefully selected some hospitals to be in their registry, either because they had an existing relationship with them or because the hospitals were thought to be especially relevant to acute stroke care in the state. These hospitals are referred to as selected with certainty (SWC) and are given a weight of 1.0 in the statistical analysis.2
In the Ga prototype, 8 hospitals were SWC from the most populous county to assure the representation of hospitals with the highest volume of stroke admissions in the state. Fifty-two hospitals were then randomly selected from the remaining statewide hospitals using a sampling fraction of ≈1 to 3. Of these, 9 were excluded because they lacked an emergency department (ED), and of the remaining 43 hospitals, 26 agreed to participate (making a final sample of 34).
In Mass, a purposeful or judgment sampling approach2 was used to obtain a sample of 12 hospitals that were representative of the states demographics and geography. All the hospitals were SWC, having been chosen based on the presence of house staff, current participation in a cardiovascular disease quality-improvement project, and the presence of an opinion leader to champion stroke. None of the hospitals declined to participate. The 12 hospitals were responsible for providing care to ≈25% of all acute stroke admissions statewide.
In Mich, a modified stratified sampling regime was used to obtain a sample of 16 hospitals. First, 8 hospitals were SWC from 4 urban communities that were participating in a community-based stroke project. These hospitals represented larger academic and nonacademic urban institutions. Second, the remaining 114 statewide acute care hospitals that had at least 30 stroke discharges in year 2000 were ranked according to their total number of stroke discharges, and 4 equal-sized strata were created. Two hospitals were then randomly selected, with probability proportional to size, from each of the 4 strata, which in conjunction with the 8 pilot phase (SWC) hospitals resulted in a final sample of 16 hospitals. All of the hospitals agreed to participate. The 16 selected hospitals were responsible for ≈25% of stroke discharges statewide.
The Ohio prototype was also based on a modified stratified sampling design. First, all the hospitals in the Cincinnati and Cleveland metropolitan areas were invited to participate (ie, SWC), and all 12 Cincinnati area hospitals and 14 of the 27 Cleveland area hospitals agreed to do so. Second, the remaining 109 statewide hospitals were ranked according to size (based on the total number of admissions per year) to create 3 strata: >20 000 admissions (large), >10 000 to 20 000 admissions (medium), and 2000 to 10 000 admissions (small). A random sample of hospitals was then drawn from each stratum—specifically, 2 from the 11 large hospitals, 3 from the 21 medium hospitals, and 5 from the 77 small hospitals. The prototype included 36 hospitals in total.
Case Ascertainment and Case Definition
Two registries (Mass and Mich) identified acute stroke admissions using a prospective design, whereby eligible acute stroke admissions were identified in real time based on presenting clinical signs and symptoms. The other 2 states (Ga and Ohio) used a retrospective design, whereby eligible cases were identified after discharge based on specific primary or secondary International Classification of Disease, 9th Revision stroke-related discharge codes (ie, 430, 431, 433 to 436 codes inclusive, and 432.9). These codes have been shown to have high predictive value for acute stroke admissions.3
Regardless of whether a prospective or retrospective design was used, all subjects had to meet the following case definition for an acute stroke admission to be included: (1) present with signs and symptoms consistent with a new onset acute stroke or transient ischemic attack (TIA) that was the principal reason for hospital admission to a registry hospital; and (2) meet 1 of the 7 stroke subtype case definitions, namely ischemic stroke (IS), intracerebral hemorrhage (ICH), subarachnoid hemorrhage, TIA, ischemic stroke of uncertain duration (ISUD), hemorrhagic stroke of uncertain type, or stroke of uncertain type.
These case definitions were adapted from previous work3 and were applied by the trained study personal after review of the completed medical chart. Consultation with a neurologist occurred where necessary. To be included in the registry, TIA cases had to have evidence of stroke-like symptoms on presentation to the ED. For the prospective designs, the ISUD case definition was used when the duration of clinical signs was unknown for subjects who had an ischemic event (it was therefore impossible to differentiate between IS and TIA4). Because there is no ICD-9 code for ISUD, and because information on the duration of stroke symptoms during hospitalization was typically lacking in medical records, retrospective registries did not use the ISUD case definition.
Human subject approval was obtained from each hospital’s institutional review board before beginning data collection. Most institutional review boards granted the project exempt status because the registry’s primary purpose was quality improvement, and no direct patient contact was involved. All sites collected the same set of core data elements, which included information on demographics, prehospital events (including stroke onset and use of emergency medical services), ED triage and workup (including brain imaging and thrombolytic treatment), medical history, in-hospital diagnostic and treatment procedures, in-hospital complications, discharge information, and use of secondary prevention measures at discharge. Because the registry was unable to obtain individual patient consent, it was not possible to collect follow-up outcomes data after discharge.
Sites using a retrospective design obtained data through chart abstraction using trained medical abstractors. At sites using a prospective design, trained hospital staff collected data using a combination of real-time data collection (for prehospital events and ED workup) and chart abstraction. All case ascertainment occurred between October 2001 and November 2002; the time period of data collection was 2 months for Ohio, 3 months for Ga and Mass, and 6 months for Mich. All consecutive acute stroke admissions that occurred during these time periods were included in each registry. Each site undertook their own data quality-assessment by re-abstracting a sample of charts. In addition, Research Triangle Institute, which served as an independent auditor to Centers for Disease Control, re-abstracted a sample of cases from each prototype using a team of trained medical abstractors.5
Definition of Stroke Onset Time
Every admission included in the registry had a stroke onset time recorded to the highest level of accuracy the available information would support. Time of stroke onset was documented as either specific (ie, witnessed), estimated (ie, onset estimable within a 6-hour window), or unknown. For estimated times, the onset was taken as the earliest hour within the 6-hour time period. For subjects who awoke with stroke symptoms, onset time was estimated as the time the subject went to sleep. For subjects who had an unknown onset time, the date the subject was last known to be normal was collected. This information along with the ED arrival date was used to estimate onset time as <24 hours, 24 to <48 hours, ≥48 hours, or unknown.
Recombinant Tissue Plasminogen Activator Treatment
Detailed information regarding treatment with recombinant tissue plasminogen activator (rtPA) was recorded, including treatment route (ie, intravenous [IV] only, intra-arterial [IA] only, or IV/IA combination); date and time treatment began (ie, needle time); and complications caused by symptomatic ICH. The eligible population were limited to those with ischemic stroke (ie, IS or ISUD). Two rtPA treatment rates were calculated, one for all treated cases regardless of route and one for IV-only–treated cases. The IV-only treatment group was limited to cases whose treatment was initiated at a Coverdell registry hospital; patients whose treatment began at a nonregistry hospital but were then transferred to a registry hospital were excluded. Complication rates of symptomatic ICH caused by rtPA treatment (defined as clinical neurological deterioration within 36 hours of treatment with evidence of intracranial bleeding on brain imaging) were calculated for all rtPA cases and for IV-only–treated cases. Finally, 3 critical time intervals: onset time to ED arrival, ED arrival to treatment initiation (ie, door-to-needle time), and onset to treatment initiation (ie, onset-to-needle time) were calculated for IV-only–treated subjects.
In-Hospital Procedures, Treatments, and Medications at Discharge
Calculation of stroke care performance indicators requires the relevant eligible population be defined according to stroke subtype, discharge status, and the presence of documented contraindications. A lipid profile was defined as measurement of any lipid panel component (ie, total cholesterol and/or low-density lipoprotein or high-density lipoprotein subfractions and/or triglycerides) during hospitalization. The relevant population was limited to subjects with a diagnosis of IS, ISUD, or TIA, consistent with current evidence-based guidelines.6,7 Smoking cessation counseling was defined as any documented discussion, including referral to counseling services or discharge with smoking cessation medications. Antithrombotic medications at discharge included any antiplatelet or anticoagulant medication listed on the discharge instructions. Use of anticoagulant medications at discharge in atrial fibrillation patients included warfarin or other anticoagulation medications listed on the discharge instructions.
A statistical approach consistent with the descriptive objectives of this report was used throughout. To generalize findings to the state as a whole, site-specific proportions (with 95% confidence intervals) were weighted to adjust for each prototype’s sampling design using SUDAAN, a statistical package for complex survey designs. Hospital weights were generated as the inverse of the sampling probability within each stratum, whereas SWC hospitals were given a weight of 1.0.2 Because the Mass prototype included only SWC hospitals, the analysis of their data are essentially unweighted and can be regarded as expressing variability in the data over time. Data from the Ga prototype were also adjusted for hospital nonresponse.
A total of 6867 acute stroke admissions were included in the 4 prototype registries. The distribution of stroke subtypes varied across the 4 sites; between 52.1% (Mass) and 69.5% (Ohio) of admissions had a final stroke diagnosis of IS, whereas between 10.3% (Ohio) and 17.7% (Mass) had a hemorrhagic stroke (ie, ICH or subarachnoid hemorrhage) (Table 1). The proportion of TIA admissions varied from 16.6% (Mass) to 23.4% (Mich). The age distribution also varied across the 4 prototypes; the median age was lowest in Ga (68 years) and highest in Ohio (71 years), as did the proportion of females (range 52.1% [Mass] to 58% [Ohio]). As expected, the proportion of black subjects varied considerably across the 4 sites; Ga had the highest proportion (30.6%) and Mich the lowest (7.1%). In all 4 prototypes, Hispanic status was not documented in the great majority of admissions.
A summary of stroke onset time information is provided in Table 2. There was marked variability in the recording of onset times across the 4 sites. For example, almost half the subjects in the Mass registry had specific onset times recorded, which was much higher than any of the other 3 sites; the proportion of subjects with no onset times documented was much higher in Mich (42%) than anywhere else. Less than 10% of admissions arrived at the ED within 1 hour of onset (range, 5.6% [Ga] to 9.5% [Mass]), whereas only ≈ 20% to 25% arrived within 3 hours (range, 19.0% [Mich] to 26.5% [Mass]).
Across the 4 prototypes, a total of 177 subjects were treated with rtPA (IV, IA, or IV/IA) among 4280 eligible subjects (defined as those with a final diagnosis of IS or ISUD) (Table 3). Site-specific overall rtPA treatment rates varied from 3.0% in Ga to 8.5% in Mass. A total of 118 subjects had IV-only rtPA treatment that was initiated in a Coverdell registry hospital; site-specific IV-only rtPA treatment rates varied from 2.0% in Ohio to 6.3% in Mass. A total of 27 subjects (from Mass, Mich, and Ohio) received IV treatment that was initiated outside a Coverdell registry hospital, whereas 32 cases (from all 4 sites) received either IA or IV/IA combined treatment (Table 3).
Three critical time intervals for the 118 IV-only rtPA–treated subjects are summarized in Table 4 (ie, onset to ED arrival, ED arrival to needle, and onset to needle). Between 45.8% (Ohio) and 67.6% (Mich) of treated subjects arrived at the ED within 1 hour of symptom onset. Only a minority of patients (range, 10.8% [Ohio] to 19.6% [Mich]) received treatment within the recommended 1-hour “door-to-needle” time interval,8 with the majority receiving treatment between 1 and 2 hours after ED arrival. The onset to needle time interval was longer than 3 hours for 16 subjects and was unknown for another 3.
Of the 118 IV-only rtPA–treated cases, 5 had complications of symptomatic ICH; the site-specific complication rates were 1 of 19 (5.3%) in Ga, 2 of 45 (4.4%) in Mass, and 2 of 33 (6.1%) in Mich, and 0 of 21 (0.0%) in Ohio. When all 177 rtPA-treated cases were considered, 9 cases had symptomatic ICH; the site-specific complication rates were 1 of 22 (4.5%) in Ga, 5 of 60 (8.3%) in Mass, 2 of 53 (3.8%) in Mich, and 1 of 42 (2.4%) in Ohio.
The proportion of subjects documented to have undergone dysphagia screening during hospitalization was reasonably consistent across sites, varying from 38.5% in Ga to 50.7% in Mass (Table 5). Lipid profiles were checked during hospitalization in a minority of eligible subjects (range, 28.4% [Ohio] to 39.4% [Mich]). Compliance with smoking cessation interventions was very poor in all prototypes, varying between only 16.5% in Ohio to just 34.1% in Mich. In terms of discharge medications, use of antithrombotics among ischemic stroke patients was universally high (range, 87.7% [Ohio] to 97.7% [Mich]), whereas among 544 patients who had atrial fibrillation, the rate of anticoagulation use varied quite widely (range, 64.1% [Ohio] to 91% [Mich]).
The findings of this report give a comprehensive overview of onset-to-arrival times for acute stroke patients and makes for sobering reading. Less than 10% of acute stroke admissions arrived at the ED within 1 hour of stroke onset, whereas less than one-quarter arrived within 3 hours. Stroke onset was also poorly documented; only between one-third and one-half of acute stroke admissions had a specific stroke onset time documented, whereas onset information was not available for 20% to 40%. These findings point to the need for continued mass public education to increase awareness and recognition of early warning signs, and the importance of seeking emergency medical care.8–10 Direct comparisons between our onset time data and previous studies are made difficult because of the marked variability in methodological approaches. However, in a systematic review of prehospital and in-hospital delay based on 48 reports, Evenson et al11 reported that the median time between symptom onset and ED arrival ranged between 3 and 6 hours, and that between 30% and 60% of patients arrive at the ED within 3 hours of onset. Because we accounted for all subjects in our estimate (including those with unknown onset times), our data provide a conservative but more realistic view of onset-to-arrival times. The variability in the recording of onset time across the 4 sites probably stems mostly from differences in the interpretation of definitions for specific and estimated onset times. The distinction between these 2 hinges on the presence of a witnessed event that implies onset time is known precisely. Developing a common approach to onset time documentation would do much to improve the completeness and accuracy of stroke onset time information recorded in future registries.11
Overall, between 3.0% and 8.5% of ischemic stroke patients were treated with rtPA in these pilot registries. These rtPA treatment rates are consistent with those reported in the other community-based or multi-hospital reports,12–20 in which rtPA treatment rates among all ischemic stroke admissions have varied from 1.6%15 to 9%,16 with an average of ≈3% to 4%. Similarly, with the exception of 1 study,14 post-treatment symptomatic ICH rates have ranged between 5% and 7% in these reports.12–20 In our data, symptomatic ICH complication rates were similar or even lower than these previous reports, giving further assurance that rtPA is being used with a high degree of safety in more community-based settings.
Although the majority of IV rtPA–treated cases in this registry were treated within the recommended 3-hour treatment window, only a minority (between 11% and 20%) actually received the drug within the recommended 1-hour door-to-needle time.8 These results imply that most patients would have needed to arrive at the hospital within 1 hour of stroke onset to receive rtPA treatment. This is indeed borne out by the data; more than half of the IV rtPA–treated cases arrived at the hospital within 1 hour of onset. These observations further illustrate the inherent limitations of the current situation in which <10% of all admissions arrived at the hospital within 1 hour of onset. Although the need to decrease onset-to-arrival time for suspect stroke victims is clear, these data also emphasize the need for quality improvement projects focused on reducing in-hospital delays in the emergent evaluation of acute stroke admissions.21,22 Efforts to reduce in-hospital delays would likely have the greatest immediate impact on increasing rtPA treatment rates.
This registry documented that in-hospital dysphagia screening, lipid testing, and smoking cessation counseling were all greatly underused. Antithrombotic and anticoagulant therapies were used at encouragingly high levels, although there was more site-to-site variation in the use of anticoagulants (range 64% to 91%). Our findings are in general agreement with those of more recent reviews,23,24 as well as the latest findings of the Canadian Stroke Registry.25 We found the use of anticoagulants among atrial fibrillation patients was higher at the 2 prospective sites (Mass and Mich); it is possible that this method of data collection allowed for the more accurate identification of the eligible subpopulation (thereby increasing the overall compliance with this measure). Finally, despite the fact that direct comparisons between different hospital-based stroke registries are complicated by variations in selection criteria and case definitions, the overall distribution of demographic and stroke subtype characteristics are in general agreement with those of other hospital-based registries.25–30
Many factors contribute to the difficulties in identifying stroke cases and measuring stroke outcomes, including variation in case definitions, the clinical spectrum of stroke, and the accuracy of diagnostic codes.31 These challenges result in the need to identify stroke patients in a consistent, systematic manner using hospital or population-based stroke registries. The design of these prototypes attempted to address many of these concerns. First, prototypes collected data that they considered to be representative of acute stroke care in the state as a whole, an approach that led to the use of weighted statistical estimates. Second, prototypes used similar methods for case ascertainment (whether prospective or retrospective) and applied the same case definitions and data elements.
There are several limitations associated with this report. Despite the use of standardized data definitions, it is clear that variations in the interpretation of some definitions occurred. Data accuracy is obviously limited by the quality and completeness of the medical records. The use of the purposeful or judgment sampling approach in Mass resulted in data that could not be weighted in the analysis. Although the generalizability of the Mass data could be questioned, it is important to note that many of their finding are in the same range of the other 3 prototypes. The inclusion of TIA patients in the registry may be controversial; however, the investigators believe it to be justified given the fact that TIA subjects had to have stroke signs and symptoms on presentation (and could therefore initially be considered eligible for acute stroke interventions). Perhaps the greatest limitation of the current prototypes, however, is their inability to collect outcomes data after discharge. The nature of stroke and the known benefits of rtPA therapy suggest that the collection of outcomes data be given priority as the PCNASR undergoes further development. Finally, whereas this report provides the first descriptive findings from the PCNASR registries, much more work, including an assessment of completeness of case ascertainment, diagnostic accuracy of case definitions, validity and reproducibility of data collection methods, and the development of criteria to develop valid sampling designs with appropriate statistical analyses, all need to be completed to provide definitive guidelines to the design, conduct, and analysis of the final registry.
These initial results of the PCNASR represent an important step in measuring and improving the quality of acute stroke care in the United States. Despite the differences in the design on the 4 prototype, we believe that there is a remarkable level on consistency in many of the findings across the states. Quality stroke care is being provided on a consistent manner for some performance measures (eg, antithrombotics at discharge); however, treatment gaps and state-to-state variability were found in several measures (eg, documentation of stroke onset, screening for dysphagia, and smoking cessation counseling). Ongoing endeavors to define performance measures32,33 and incorporate them into common national standards for stroke-related quality insurance programs, such as those sponsored by Joint Commission on Accreditation of Health Care Organizations34 and American Stroke Association,35 will result in a national registry that will be able to identify, enable, and monitor improvements in hospital-based acute stroke care with tangible benefits to stroke patients and their families.
Funding for the Paul Coverdell National Acute Stroke Registry was provided by a cooperative agreement from the US Centers for Disease Control and Prevention (CDC). CDC staff participated in developing common data collection tools and methods, and in the manuscript review process. None of the authors have conflicts of interests to declare. In addition, the authors would like to acknowledge the contributions of the following investigators, study staff, and funding source in each state:
Georgia (The Paul Coverdell Georgia Stroke Registry Pilot Prototype): Principal Investigator: Michael R. Frankel, MD; Co-Principal Investigator: Vicki Hertzberg, PhD; Project Coordinator: Kerrie Krompf, BS and staff, Rodney Presley, PhD; Michelle Manzo, MPH.
This study was supported by US Centers for Disease Control and Prevention Cooperative Agreement No. (U50/CCU420275-01).
Massachusetts (The Massachusetts Prototype for the Paul Coverdell National Acute Stroke Registry): MassPRO: Kenneth A. LaBresh, MD, FACC (Co-Director); Massachusetts General Hospital/Harvard Medical School, Lee H. Schwamm, MD (Co-Director); MassPRO staff: Patricia J. Lambert, RN, BSN, CPHQ (Project Coordinator), James Freccero, MPH (Statistical Analysis). American Heart Association-Northeast Affiliate staff: Denise P. Normandin (Program Manager, Quality Initiatives), Janet Prvu, MS (Operation Stroke Program Director).
This study was supported by US Centers for Disease Control and Prevention Cooperative Agreement No. (U50/CCU120238-01).
Michigan (MASCOTS—Michigan Acute Stroke Care Overview and Treatment Surveillance System): Principle investigator: Mathew J Reeves PhD; Co-investigators: Gretchen Birbeck MD, Susan Hickenbottom MD, Brad Jacobs MD, Rashmi Kothari MD; Staff: Andrew Mullard (database management and statistical analysis), Sue Wehner (project manager), and Yingzi Deng (statistical analysis).
This study was supported by US Centers for Disease Control and Prevention Cooperative Agreement No. (U50/CCU520272-01).
Ohio (Ohio Prototype for the Paul Coverdell Acute Stroke Registry): Principal Investigator: Joseph P. Broderick, MD; Northern Ohio Co-Investigator: Irene Katzan, MD, MS; Southern Ohio Co-Investigators: Brett Kissela, MD, Dawn Kleindorfer, MD; Project Coordinators: Laura Sauerbeck, RN, MS (year 1), Rosie Miller, RN (year 2); Study Coordinator Northern Ohio: Charlene Kolz, RN; Biostatistician: Richard W. Hornung, DrPH, Charles J. Moomaw. Additional contributors: Kyle Allen, DO, Frank Bright, MS, Mark Carrozza, MA, Randall Cebul, MD, Anthony Furlan, MD, Joseph Hanna, MD, Gary Houser (The Stroke Group), Eric Hixson, MBA, Dennis Landis, MD, Deborah Nadzam, PhD, RN, Warren Selman, MD, Andrew Slivka, MD, D. Michael Waggoner, MD.
This study was supported by US Centers for Disease Control and Prevention Cooperative Agreement No. (U50/CCU520278-01).
The authors would also like to acknowledge all of the staff from the following hospitals who participated in the registries:
Georgia: The Paul Coverdell Georgia Stroke Registry Pilot Prototype would like to thank the hospitals and their staff who agreed to participate in the prototype on a confidential basis.
Massachusetts: Baystate Medical Center-Springfield; Berkshire Medical Center-Pittsfield; Beth Israel Deaconess Medical Center-Boston; Boston Medical Center-Boston; Brigham & Women’s Hospital-Boston; Caritas Carney Hospital-Dorchester; Faulkner Hospital-Jamaica Plain; Lahey Clinic Medical Center-Burlington; Martha’s Vineyard Hospital-Oak Bluffs; St. Elizabeth’s Medical Center-Brighton; Tufts-New England Medical Center-Boston.
Michigan: Spectrum Health Systems, Grand Rapids; St. Joseph Mercy Hospital, Ann Arbor; University of Michigan Hospital, Ann Arbor; Borgess Medical Center, Kalamazoo; Sparrow Health Systems, Lansing; Ingham Regional Medical Center, Lansing; Detroit Receiving Hospital; Henry Ford Wyandotte Hospital; St. Joseph Mercy of Macomb; Northern Michigan Regional Health System, Petoskey; St. Mary’s Hospital, Saginaw; Bronson Methodist Hospital, Kalamazoo; Harper University Hospital, Detroit; Alpena General Hospital; St. Joseph Health Systems, Tawas.
Ohio: Northern Ohio Regional Sites (Cleveland area): Cleveland Clinic Health System (Cleveland Clinic Foundation, Euclid Hospital, Hillcrest Hospital, Huron Hospital, South Pointe Hospital, Fairview Hospital; Lakewood Hospital; Lutheran Hospital; Marymount Hospital); MetroHealth Medical Center; Southwest General Health Center; University Hospitals of Cleveland; University Hospitals Health System Geauga Regional Hospital.
Southern Ohio Regional Sites (Greater Cincinnati area): Deaconess Hospital; The Health Alliance Hospitals (The Christ Hospital, The Jewish Hospital, The University Hospital); The Mercy Health Partners (Anderson Mercy Hospital, Clermont Mercy Hospital, Mercy Fairfield, Mercy Franciscan Mount Airy, and Mercy Franciscan Western Hills); Tri-Health (Bethesda North, Good Samaritan Hospital); Veterans Affairs Medical Center.
Other Ohio Sites: Barnesville Hospital Association, Barnesville; Cuyahoga Falls General Hospital, Cuyahoga Falls; Fairfield Medical Center, Lancaster; Genesis Healthcare System, Zanesville; Humility of Mary Health Partners, Youngstown, Warren, Boardman; Joint Township District Memorial Hospital, St. Marys; MedCentral Health System, Mansfield; Ohio State University Medical Center, Columbus; Pomerene Hospital, Millersburg; Riverside Methodist Hospital, Columbus.
- Received October 13, 2004.
- Accepted December 7, 2004.
Wattigney WA, Croft JB, Mensah GA, Alberts MJ, Shephard TJ, Gorelick PB, Nilasena DS, Hess DC, Walker MD, Hanley DF Jr, Shwayder P, Girgus M, Neff LJ, Williams JE, LaBarthe DR, Collins JL; Paul Coverdell National Acute Stroke Registry. Establishing data elements for the Paul Coverdell National Acute Stroke Registry. Part I: proceedings of an expert panel. Stroke. 2003; 34: 151–156.
Lohr SL. Sampling: Design and Analysis. Pacific Grove, Calif: Duxbury Press; 1999.
Broderick J, Brott T, Kothari R, Miller R, Khoury J, Pancioli A, Gebel J, Mills D, Minneci L, Shukla R. The Greater Cincinnati/Northern Kentucky Stroke Study: preliminary first-ever total incidence rates of stroke among blacks. Stroke. 1998; 29: 415–421.
Evaluation of prototypes for the Paul Coverdell National Acute Stroke Registry: Summary of audit activities and findings from the first four prototypes. Final report to CDC. Research Triangle Institute; 2003.
Marler J, Jones P, Emr M, eds. Proceedings of a National Symposium on Rapid Identification and Treatment of Acute Stroke. Arlington, VA: National Institute of Neurological Disorders and Stroke; 1997.
Am Heart Association. Stroke warning signs and risk factors. 2001. Available at: http://www.americanheart.org/presenter.jhtml?identifier=3053. Accessed April 18, 2005.
National Institute of Neurological Disorders and Stroke (NINDS). Know stroke. Know the signs. Act in time. May 2001. Available at: http://www.ninds.nih.gov/disorders/stroke/knowstroke.htm. Accessed April 18, 2005.
Chiu D, Krieger D, Villar-Cordova C, Kasner SE, Morgenstern LB, Bratina PL, Yatsu FM, Grotta JC. Intravenous tissue plasminogen activator for acute ischemic stroke. Stroke. 1998; 29: 18–22.
Wang DZ, Rose JA, Honings DS, Garwacki DJ, Milbrandt JC. Treating acute stroke patients with intravenous rt-PA—the OSF stroke network experience. Stroke. 2000; 31: 77–81.
Reed SD, Cramer SC, Blough DK, Meyer K, Jarvik JG.. Treatment with tissues plasminogen activator and inpatient mortality rates for patients with ischemic stroke treated in community hospitals. Stroke. 2001; 32: 1832–1840.
Heuschmann PU, Berger K, Misslewitz B, Hermanek P, Leffmann C, Adelmann M, Buecker-Nott HJ, Rother J, Neundoerfer B, Kolominsky-Rabas PL; German Stroke Registers Study Group; Competence Net Stroke. Frequency of thrombolytic therapy in patients with acute ischemic stroke and the risk of in-hospital mortality: The German Stroke Registers Study Group. Stroke. 2003; 34: 1106–1113.
Katzan IL, Hammer MD, Furlan AJ, Hixson ED, Nadzam DM; Cleveland Clinic Health System Stroke Quality Improvement Team. Quality improvement and tissue-type plasminogen activator for acute ischemic stroke: Cleveland update. Stroke. 2003; 34: 799–800.
Kleindorfer D, Kissela B, Schneider A, Woo D, Khoury J, Miller R, Alwell K, Gebel J, Szaflarski J, Pancioli A, Jauch E, Moomaw C, Shukla R, Broderick JP; Neuroscience Institute. Eligibility for recombinant tissue plasminogen activator in acute ischemic stroke. Stroke. 2004; 35: e27–e29.
NINDS rt-PA Stroke Study Group. A systems approach to immediate evaluation and management of hyperacute stroke: experience at eight centers and implications for community practice and patient care. Stroke. 1997; 28: 1530–1540.
Holloway RG, Benesch C, Rush SR. Stroke prevention: narrowing the evidence-practice gap. Neurology. 2000; 54: 1899–1906.
Kapral MK, Laupacis A, Phillips SJ, Silver FL, Hill MD, Fang J, Richards J, Tu JV; Investigators of the Registry of the Canadian Stroke Network. Stroke care delivery in institutions participating in the Registry of the Canadian Stroke Network. Stroke. 2004; 35: 1756–1762.
Bogousslavsky J, Van Melle G, Regli F. The Lausanne Stroke Registry: analysis of 1,000 consecutive patients with first stroke. Stroke. 1988; 19: 1083–1092.
Zweifler RM, Lyden PD, Taft B, Kelly N, Rothrock JF. Impact of race and ethnicity on ischemic stroke. Stroke. 1995; 26: 245–248.
Di Carlo A, Lamassa M, Baldereschi M, Pracucci G, Basile AM, Wolfe CD, Giroud M, Rudd A, Ghetti A, Inzitari D; European BIOMED Study of Stroke Care Group. Sex differences in the clinical presentation, resource use and 3-month outcomes of acute stroke in Europe: data from a multicenter multinational hospital-based registry. Stroke. 2003; 34: 1114–1119.
Bonita R, Broad JB, Anderson NE, Beaglehole R. Approaches to the problems of measuring the incidence of stroke: the Auckland Stroke Study 1991–1992. Int J Epidemiol. 1995; 24: 535–542.
Measuring and improving quality of care: report from the American Heart Association and American College of Cardiology first scientific forum on assessment of healthcare quality on cardiovascular disease and stroke. Stroke. 2000; 31: 1002–1012.
Holloway RG, Vickrey BG, Benesch C, Hinchey JA, Bieber J; National Expert Stroke Panel. Development of performance measures of acute ischemic stroke. Stroke. 2001; 32: 2058–2074.
Joint Commission on Accreditation of Healthcare Organizations (JCAHO). Primary Stroke Center Certification Program. Available at: http://www.strokeassociation.org/presenter.jhtml?identifier=3016808. Accessed April 18, 2005.
American Stroke Association. Get with the guidelines—stroke. Available at: http://www.strokeassociation.org/presenter.jhtml?identifier=3002728. Accessed April 18, 2005.