Characteristics of an “Ill-Defined“ Diagnosis for Stroke

    the Cardiovascular Health Branch, Division of Adult and Community Health, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, Ga.

    Abstract

    Background and Purpose— Rapid and accurate evaluation of stroke subtypes is crucial for optimal treatment and outcomes. This study assessed factors associated with the likelihood of an "ill-defined" diagnosis for stroke hospitalizations.

    Methods— We examined all hospital claims for stroke among Medicare beneficiaries aged 65 years in 2000. Stroke subtypes included hemorrhagic (International Classification of Diseases, Ninth Revision, Clinical Modification codes 430 to 432), ischemic (433 to 434), ill-defined (436 to 437), and late effects of cerebrovascular disease (438).

    Results— Among 445 452 hospital claims for stroke, 65.3% were ischemic, 20.9% were ill defined, 11.9% were hemorrhagic, and 1.9% were late effects of cerebrovascular disease. After controlling for age, women (odds ratio [OR],1.30; 95% CI, 1.28 to 1.32), blacks (OR, 1.31; 95% CI, 1.28 to 1.33), and Hispanics (OR, 1.27; 95% CI, 1.20 to 1.34) were more likely to receive a discharge diagnosis of ill defined compared with men and whites, respectively. Differences in age, sex, emergency room presentation, and evidence of diagnostic procedures accounted for some but not all racial disparities. In 14 states, ill-defined strokes constituted 25% of all stroke diagnoses.

    Conclusion— The high proportion of stroke patients who receive an ill-defined diagnosis on discharge suggests a continued need for improvements in early response and prompt evaluation of strokes. Findings of geographic, gender, and racial disparities in ill-defined stroke diagnosis warrant further investigation. Reimbursement practices and public health efforts that promote hospital stroke policies are critical to improve disease reporting as well as clinical outcomes.

    Key Words: ethnicity  hospitalization  Medicare  race  stroke

    Introduction

    Stroke, the third leading cause of death1 and a major cause of serious disability in the United States,2 is estimated to have cost $56.8 billion, including $13.2 billion for nursing home care, in 2005.3 Although death rates have declined since 1900,4 hospitalizations for stroke and the proportion of stroke patients transferred to short- and long-term care facilities have increased in the past decade.5 Of >445 000 Medicare hospital claims submitted in 2000 for short-term stays for stroke, 20.9% resulted in transfers to skilled nursing facilities and 19.6% in transfers to other care facilities.6 Because severe stroke cases are more likely to be discharged to a skilled nursing facility,7–9 use of care facilities will likely increase as the US population continues to age.

    The chance of surviving an acute stroke with minimal disability could be increased by rapid recognition and reaction of bystanders and patients to the warning signs, early transport to an emergency room (ER) equipped and staffed to handle stroke cases, and rapid evaluation, diagnosis, and treatment. During the past decade, recommendations for the care of acute stroke patient have evolved with considerable advances in diagnosis and treatment.10–16 However, optimizing treatment continues to depend on rapid evaluation and accurate diagnosis of the specific stroke subtype. We assessed the distribution of stroke subtypes in Medicare hospital claims for stroke among adults 65 years of age in 2000 to determine differences in the proportion of stroke hospitalizations with an ill-defined discharge diagnosis for groups defined by age, race/ethnicity, sex, and state residency. We then examined possible explanations for differences between these groups in terms of ER presentation, evidence of reported or billed diagnostic procedures, other conditions, and hospital outcomes.

    Methods

    For this cross-sectional study, we obtained all Medicare Part A claims from the Centers for Medicare and Medicaid Services. Claims included age, sex, race/ethnicity, state of residence, type of admission, admission diagnosis, 10 discharge diagnosis fields, 6 procedure fields, bill indicators, discharge outcome, and diagnosis-related group (DRG). There were 445 452 hospital claims in 2000 submitted for persons admitted to short-stay hospitals with a principal (first-listed) discharge diagnosis of stroke, which included International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes for hemorrhagic stroke (430 to 432), ischemic stroke (433 to 434), ill-defined stroke (436 to 437), and late effects of cerebrovascular disease (438), which includes neurological deficits that persist after the initial onset of the cerebrovascular event. A discharge diagnosis for transient ischemic attack (TIA; ICD-9-CM code 435) was not included in this study.

    For all hospital claims, we examined the distribution of stroke subtypes for patients defined by age (65 to 74, 75 to 84, and 85 years), sex, race/ethnicity (white, black, Hispanic, and other), state of residence, ER admission (ER versus all other types of referrals and transfers), and admission diagnosis. Diagnosis at admission included ill-defined stroke (ICD-9-CM code 436 to 437), other specific stroke subtypes (430 to 434 or 436 to 438), TIA (435), and all other admission diagnoses. We also compared the percentages of these and other selected characteristics (secondary diagnoses, procedures, DRG, and discharge characteristics) among subtypes. Secondary diagnoses (second through tenth ICD-9-CM discharge codes) included diabetes (250.0 to 250.9), atrial fibrillation (427.3 or 427.3 to 427.2), and hypertension (401 to 405). Up to 6 procedures reported on the hospital claim may have included those specific to the evaluation and diagnosis of stroke subtypes such as computerized tomography (CT) of the head (ICD-9-CM 87.03), arteriography of cerebral arteries (88.41), diagnostic ultrasound of the head and neck (88.71), or MRI of the brain and brain stem (88.91). Bill indicators included those for a CT scan or for other unspecified imaging techniques. Discharge outcomes included discharge to home, a skilled nursing facility or another facility (intermediate care, short-term care, or other type of facility), death during the hospital stay, or other outcome (eg, left against medical advice or unknown outcome).

    To assess the association of selected characteristics with a diagnosis of ill-defined stroke rather than any other subtype, we obtained odds ratios (ORs) and 95% CIs for each characteristic from a logistic regression model that included age and then from a multivariable logistic regression model that included age, race/ethnicity, sex, admission through an ER, and evidence of diagnostic procedures (CT, diagnostic ultrasound, arteriography, or MRI) as covariates. Because ER presentation was highly associated with receiving an ill-defined diagnosis, we assessed the association of selected characteristics with ER presentation for any stroke in separate multivariate logistic regression models. In a final analysis assessing correlates for an ill-defined diagnosis, we restricted the study to patients admitted only through the ER.

    Results

    In 2000, stroke subtypes included 65.3% ischemic, 20.9% ill defined, 11.9% hemorrhagic (intracerebral and subarachnoid), and 1.9% late effects of cerebrovascular disease (Table 1). The proportion with a discharge diagnosis of ischemic decreased among successive age groups, whereas the proportions with a diagnosis of either ill-defined or hemorrhagic stroke increased with age. Women were more likely than men and blacks and Hispanics were more likely than whites to receive ill-defined diagnoses. Patients admitted through the ER were more likely to have discharge diagnoses of either ill-defined or hemorrhagic stroke and less likely to be diagnosed with ischemic stroke than patients referred or transferred from other sources (private physicians, clinics, other hospitals, skilled nursing facilities, or other facilities). Patients with an admitting diagnosis of ill-defined stroke were more likely to have the same diagnosis at discharge than patients with admission diagnoses of specific subtypes, TIA, or any other diagnosis. States with lower proportions of stroke hospital claims reported with an ill-defined diagnosis on discharge had much higher proportions with an ischemic diagnosis (Table 2). There were 14 states with 25% of all stroke hospital claims diagnosed as ill defined on discharge; states in the northeastern United States had lower proportions of strokes diagnosed as ill defined.

    We examined characteristics of patients with each stroke subtype (Table 3). Patients with a discharge diagnosis of ill-defined stroke compared with those with ischemic or hemorrhagic stroke were more likely to be 85 years of age, a woman, or black. In addition, stroke patients with a discharge diagnosis of ill defined were more likely than those with an ischemic diagnosis to have ER presentation and an admitting diagnosis of ill-defined stroke. A report of hypertension was less common among those with ill-defined compared with ischemic diagnosis. Regardless of discharge diagnosis, the proportion with a hospital claim bill indicator for CT was high, although the reported use of diagnostic imaging procedures such as CT was low (Table 3). For example, report of a CT procedure varied between 5.5% and 10.2% among subtypes, whereas submission of a bill indicator for CT varied between 55.3% and 89.2%. Patients with an ill-defined diagnosis were less likely than those with an ischemic diagnosis to have procedure codes reported for arteriography, diagnostic ultrasound, or MRI but more likely to have bill indicators for CT and procedure codes reported for CT. In addition, stroke patients with an ill-defined diagnosis were more likely to be discharged to skilled nursing or other care facilities than those with an ischemic diagnosis. Patients with a hemorrhagic diagnosis had the highest in-hospital mortality and lowest discharge to home than patients with ill-defined or ischemic diagnoses. Ill-defined diagnosis at discharge was more likely to have DRG 14 (specific cerebrovascular, including intracranial hemorrhage and cerebral infarction) than other subtypes, mainly because 29% of ischemic and 20% of hemorrhagic diagnoses had a DRG associated with surgery.

    The proportion of stroke patients with ill-defined diagnoses increased with age (P<0.05; Table 4). After controlling for age, an ill-defined diagnosis was significantly more likely (P<0.05) among women than men (OR, 1.30; 95% CI, 1.28 to 1.32) and among blacks (OR, 1.31; 95% CI, 1.28 to1.33) and Hispanics (OR, 1.27; 95% CI, 1.20 to 1.34) than whites. An ill-defined diagnosis at discharge was also associated (P<0.05) with ER presentation, an ill-defined diagnosis at admission, presence of a bill indicator or procedure report for CT, absence of reported procedures for brain MRI or arteriography of cerebral arteries, a secondary diagnosis of diabetes, absence of secondary diagnosis codes for atrial fibrillation and hypertension, hospital survival, and discharge to a skilled nursing or other care facility. Multivariable logistic regression showed that differences in age, sex, ER presentation, and evidence of diagnostic procedures accounted for much of the higher proportion of ill-defined diagnoses among blacks and Hispanics compared with whites (Table 4). However, significant disparities in receiving an ill-defined diagnosis still remained for those aged 85 years, women, blacks, and Hispanics.

    Only 59.6% of stroke hospitalizations had an ER presentation; however, there was an ER presentation for 70.8% of hospitalizations with an ill-defined diagnosis compared with 56.6% of those with other subtype diagnoses. We examined differences for stroke cases defined by ER presentation. Among all stroke hospitalizations, ER presentation increased with age (52.4% for 65 to 74, 59.1% for 75 to 84, and 70.3% for 85 years; P<0.01), was higher for women than men (62.6% versus 55.6%, respectively; P<0.01 after age adjustment), and higher for blacks and Hispanics than whites (70.1% and 70.9% versus 58.1%, respectively; P<0.01 after age adjustment). However, ER presentation was lower in the 14 states with 25% of stroke hospitalizations having an ill-defined diagnosis compared with all other states (54.7% versus 60.6%, respectively; P<0.01 after adjustment for age, race, and sex). Among all stroke hospitalizations, those with ER presentations compared with those referred/transferred, respectively, were more likely to have an ill-defined diagnosis on hospital admission (39.9% versus 20.7%); bill indicators for CT (92.7% versus 44.4%); reported procedure codes for CT of head (10.1% versus 4.8%), diagnostic ultrasound of head and neck (3.9% versus 2.4%), or brain MRI (4.5% versus 2.6%); and DRG 14 for specific cardiovascular disease (84.8% versus 43.1%). All comparisons were significant at P<0.01 after age adjustment.

    In final multivariable logistic regression analyses of the 265 441 stroke cases who were admitted only through the ER, a greater likelihood of an ill-defined diagnosis remained for blacks versus whites but was no longer statistically significant (OR, 1.03; 95% CI, 1.00 to 1.06); significant differences remained between Hispanics and whites (OR, 1.16; 95% CI, 1.09 to 1.24) and between women and men (OR, 1.13; 95% CI, 1.11 to 1.15). A bill indicator for CT was associated with a lowered likelihood of having an ill-defined diagnosis (OR, 0.70; 95% CI, 0.68 to 0.73). Procedure code reported for CT of head, diagnostic ultrasound of head and neck, arteriography of cerebral arteries, and brain MRI were also associated with lower likelihoods of receiving an ill-defined diagnosis (P<0.01) among hospitalizations with an ER presentation.

    Discussion

    In 2000, 1 in 5 Medicare hospital claims for stroke involved an ill-defined diagnosis at discharge, a disturbing proportion given the need for an accurate diagnosis to guide evaluation and treatment. The likelihood of receiving an ill-defined diagnosis increased with age and was higher among women, blacks, and Hispanics, findings that could raise issues of relative quality of care. Of the 14 states with high proportions (25%) of ill-defined diagnoses, 5 were southeastern states. Overall procedure reporting for CT of the head and brain MRI were low among Medicare stroke patients in 2000, although the majority of these Medicare claims had bill indicators for CT scans. An ill-defined diagnosis was less likely with a reported procedure code for MRI but more likely with a bill indicator for CT scan. ER presentation and an admission diagnosis of ill-defined stroke were highly associated with a discharge diagnosis of ill defined. However, variations in evidence of diagnostic procedures and ER presentation seemed to account for some but not all of the differences between groups defined by age, race, and sex.

    ER presentation was highly associated with both the likelihood of receiving an ill-defined diagnosis and with greater evidence of diagnostic procedures. We found no evidence to suggest that ER physicians discriminated in procedure use between these groups. ER presentation increased with age and was higher among women, blacks, and Hispanics. Thus, it is possible that other factors related to reaching the ER in time for an accurate evaluation may have influenced the likelihood of receiving an ill-defined diagnosis. Blacks and Hispanics may have lower recognition of the major stroke signs than whites;17–19 blacks may also have greater delay times and be less likely to arrive at the ER via ambulance in some communities.20 Women may also experience longer delay times to hospital treatment for acute stroke than men.21 Although this explanation is speculative, greater public health efforts are needed to increase awareness in the general population about stroke signs and the need for prompt emergency transport. Further investigation of other explanations for these disparities in subtype diagnosis is also warranted.

    Whether these findings represent systematic under-reporting of specific diagnoses and procedures in these population subgroups and states also remains to be investigated. DRG codes show that patients receiving surgical interventions were less likely to have ill-defined diagnoses, suggesting that financial incentives or reimbursement maximization may impact reporting specific subtypes on Medicare hospital claims. Certainly, Medicare reimbursement has been relatively low for physician evaluation and treatment of acute stroke; reimbursement for DRG 14 has been $5800.22 Geographic patterns in excess reporting of ill-defined strokes in contrast to specific subtypes may reflect either regional patterns in financial incentives to report specific subtypes or regional differences in medical opinion regarding definitions for specific subtypes. However, there are unlikely to be incentives to differentially diagnose ill-defined strokes among certain races or women. Because we examined data from the Medicare population 65 years of age, it is also unclear whether this is a reporting issue only among the elderly or whether disparities exist or are even greater at younger ages.

    The findings of this report are subject to several limitations. First, these administrative records do not distinguish between first or recurrent strokes; there is no information to determine whether first strokes were more or less likely to receive ill-defined diagnoses. Admitting diagnosis that specifies the ICD-9-CM code at the time of hospital admission was associated with an ill-defined diagnosis on discharge; however, there is no information regarding whether this diagnosis was given before or after performing any diagnostic procedures. In addition, Medicare Part A hospital claims records do not provide any information on medical management or medical therapy; hospital characteristics (eg, academic versus community, urban versus rural, number of beds, or average number of stroke patients); availability of diagnostic technology, stroke unit, or neurologists; or indicators for the attention to risk factor control goals or quality improvement initiatives of the hospital system. Such information would be helpful to determine whether patterns reflect access to care and other quality of care issues. The major limitation of hospital claims records is the inability to determine accuracy of physician or administrative reporting of stroke subtypes or procedure use. Without information on results of clinical evaluations, it is impossible to validate reported subtype diagnoses in administrative hospital claims. Finally, the discrepancy between bill indicators for CT and CT procedure codes suggests under-reporting of this and other diagnostic procedures on hospital claims. Despite these limitations, Medicare claims currently constitute the only available source of state-specific and national data that allow assessment of subtype distributions.

    Our objective is not to focus on the validity or accuracy of diagnosis of specific subtypes such as ischemic or hemorrhagic stroke but is concerned with reporting the high proportion of stroke patients receiving an ill-defined diagnosis. Ill-defined stroke, the operational definition of ICD-9-CM codes 436 and 437, includes apoplexy, cerebral seizure, cerebral atherosclerosis, other generalized ischemic cerebrovascular disease, cerebral aneurysm, and several other specific but rare cerebrovascular diseases. In recent surveillance reports of hospital and mortality data,5,23 ill-defined stroke (codes 436 to 437) was aggregated with ischemic stroke (codes 433 to 434). This practice may be based solely on anecdotes from clinicians that most ill-defined strokes are actually ischemic or on studies assessing ICD-9-CM coding algorithms for ischemic stroke. Various studies that estimated the sensitivity of stroke codes in hospital discharge data reported that results vary depending on the ICD-9-CM algorithm used, particularly for ischemic stroke.24–28 Although ICD-9-CM codes show relative agreement with medical record reviews for stroke,25,26 it is recognized that there may be some inaccuracies in the classification of patients with ischemic stroke.29 Efforts to eliminate the underdiagnosis of specific subtypes and improve the reporting of subtypes on hospital records must be systematic to obtain a more credible estimate of the true distribution of ischemic and hemorrhagic strokes in the United States, particularly in high-risk racial and ethnic populations.

    To our knowledge, this is the first study to examine closely the characteristics of stroke patients with an ill-defined diagnosis. More attention could potentially reduce numbers of individuals receiving such a diagnosis. Accurate, prompt evaluation of patients admitted with suspected stroke and timely classification of stroke subtype should lead to improved treatment and reduced disability. Improved treatment could increase costs savings associated with skilled nursing and rehabilitation facilities and improve quality of life.30,31 System-wide adherence to prevention, evaluation, and treatment guidelines10–16,32 is critical to prevent strokes and provide quality of care to stroke patients. Federal agencies and other organizations have made stroke care an urgent priority and have taken steps toward improving acute stroke care. These actions include developing the Paul Coverdell National Acute Stroke Registry to monitor and improve quality of acute care,33 establishing standardized performance measures,34 making recommendations for systems of stroke care,35 and developing hospital-based programs to implement guidelines for stroke patients.36 State-specific assessments may be warranted to determine availability of diagnostic equipment and ER stroke teams with expertise in diagnosis and treatment. Additional efforts and mechanisms to give physicians feedback may be needed to promote health system adherence to evidence-based guidelines for evaluating and diagnosing stroke subtypes. Efforts are also needed in some hospital systems to improve medical reporting of stroke-related procedures and diagnoses on hospital records and Medicare claims. Removing financial barriers in Medicare reimbursement may also motivate system-wide changes in the appropriate evaluation and treatment of stroke patients. Ideally, programs such as the Paul Coverdell National Acute Stroke Registry, the American Heart Association Get With the Guidelines—Stroke, and the Joint Commission on Accreditation of Healthcare Organizations will assist clinicians in their efforts improve the quality of care provided to acute stroke patients.

    References

    Kochanek KD, Smith BL. Deaths: preliminary data for 2002. Natl Vital Stat Rep. 2004; 52: 1–47.

    Centers for Disease Control and Prevention. Prevalence of disabilities and associated health conditions among adults—United States, 1999. Morb Mortal Wkly Rep. 2001; 50: 120–125.

    American Heart Association. Heart disease and stroke statistics—2005 update. Dallas, Tex: American Heart Association; 2005.

    Centers for Disease Control and Prevention. Decline in deaths from heart disease and stroke—United States, 1990–1999. Morb Mortal Wkly Rep. 1999; 48: 649–656.

    Fang J, Alderman MH. Trend of stroke hospitalization, United States, 1988–1997. Stroke. 2001; 32: 2221–2226.

    Centers for Disease Control and Prevention. Hospitalizations for stroke among adults aged 65 years—United States, 2000. Morb Mortal Wkly Rep. 2003; 52: 586–589.

    Lai S-M, Alter M, Friday G, Lai S-L, Sobel E. Disposition after acute stroke: who is not sent home from hospital Neuroepidemiology. 1998; 17: 21–29.

    Chen Q, Kane RL, Finch MD. The cost effectiveness of post-acute care for elderly Medicare beneficiaries. Inquiry. 2000–2001; 37: 359–375.

    Kane RL, Lin WC, Blewett LA. Geographic variation in the use of post-acute care. Health Serv Res. 2002; 37: 667–682.

    Adams HP Jr, Brott TG, Crowell RM, Furlan AJ, Gomez CR, Grotta J, Helgason CM, Marler JR, Woolson RF, Zivin JA. Guidelines for the management of patients with acute ischemic stroke. A statement for healthcare professionals from a special writing group of the Stroke Council, American Heart Association. Circulation. 1994; 90: 1588–1601.

    Broderick JP, Adams HP, Barsan W, Feinberg W, Feldmann E, Grotta J, Kase C, Krieger D, Mayberg M, Tilley B, Zabramski JM, Zuccarello M. Guidelines for the management of spontaneous intracerebral hemorrhage. A statement for healthcare professionals from a special writing group of the Stroke Council, American Heart Association. Stroke. 1999; 30: 905–915.

    Bederson JB, Awad IA, Wiebers DO, Piepgras D, Haley EC Jr, Brott T, Hademenos G, Chyatte D, Rosenwasser R, Caroselli C. Recommendations for the management of patients with unruptured intracranial aneurysms. A statement for healthcare professionals from the Stroke Council of the American Heart Association. Circulation. 2000; 102: 2300– 2308.

    Coull BM, Williams LS, Goldstein LB, Meschia JF, Heitzman D, Chaturvedi S, Johnston KC, Starkman S, Morganstern LB, Wilterdink JL, Levine SR, Saver JI; Joint Stroke Guideline Development Committee of the American Academy of Neurology; American Stroke Association. Anticoagulants and antiplatelet agents in acute ischemic stroke: report of the Joint Stroke Guideline Development Committee of the American Academy of Neurology and the American Stroke Association (a division of the Am Heart Association). Stroke. 2002; 33: 1934–1942.

    Johnston SC, Higashida RT, Barrow DL, Caplan LR, Dion JE, Hademenos G, Hopkins LN, Molyneux A, Rosenwasser RH, Vinuela F, Wilson CB; Committee on Cerebrovascular Imaging of the American Heart Association Council on Cardiovascular Radiology. Recommendations for the endovascular treatment of intracranial aneurysms: a statement for healthcare professionals from the Committee on Cerebrovascular Imaging of the American Heart Association Council on Cardiovascular Radiology. Stroke. 2002; 33: 2536–2544.

    Latchaw RE, Yonas H, Hunter GJ, Yuh WT, Ueda T, Sorensen AG, Sunshine JL, Biller J, Wechsler L, Higashida R, Hademenos G; Council on Cardiovascular Radiology of the American Heart Association. Guidelines and recommendations for perfusion imaging in cerebral ischemia. A scientific statement for healthcare professionals by the writing group on perfusion imaging, from the Council on Cardiovascular Radiology of the American Heart Association. Stroke. 2003; 34: 1084–1104.

    Adams HP Jr, Adams RJ, Brott T, del Zoppo GJ, Furlan A, Goldstein LB, Grubb RL, Higashida R, Kidwell C, Kwiatkowski TG, Marler JR, Hademenos GJ; Stroke Council of the American Stroke Association. Guidelines for the early management of patients with ischemic stroke. A scientific statement from the Stroke Council of the American Stroke Association. Stroke. 2003; 34: 1056–1083.

    Greenlund KJ, Neff LJ, Zheng ZJ, Keenan NL, Giles WH, Ayala CA, Croft JB, Mensah GA. Low public recognition of major stroke symptoms. Am J Prev Med. 2003; 25: 315–319.

    Schneider AT, Pancioli AM, Khoury JC, Rademacher E, Tuchfarber A, Miller R, Woo D, Kissela B, Broderick JP. Trends in community knowledge of the warning signs and risk factors for stroke. J Am Med Assoc. 2003; 289: 343–346.

    Gordon DL. The Mississippi stroke education consortium: a state-based template to promote stroke awareness, prevention, and emergency treatment. Neuroepidemiology. 2000; 19: 1–22.

    Lacy CR, Suh D-C, Bueno M, Kostis JB; S.T.R.O.K.E. Collaborative Study Group. Delay in presentation and evaluation for acute stroke: stroke time registry for outcomes knowledge and epidemiology (S.T.R.O.K.E.). Stroke. 2001111; 32: 63–69.

    Menon SC, Pandey DK, Morgenstern LB. Critical factors determining access to acute stroke care. Neurology. 1998; 51: 427–432.

    Kleindorfer D, Hill MD, Woo D, Tomsick T, Pancioli A, Kissela B, Demchuk AM, Losiewicz D, Jauch E, Schneider A, Ringer A, Kanter D, Broderick JP. A description of Canadian and United States physician reimbursement for thrombolytic therapy administration in acute ischemic stroke. Stroke. 2005; 36: 682–687.

    Ayala C, Croft JB, Greenlund KJ, Keenan NL, Donehoo RS, Malarcher AM, Mensah GA. Sex differences in US mortality rates for stroke and stroke subtypes by race/ethnicity and age, 1995–1998. Stroke. 2002; 33: 1197–1201.

    Leibson CL, Naessens JM, Brown RD, Whisnant JP. Accuracy of hospital discharge abstracts for identifying stroke. Stroke. 1994; 25: 2348–2355.

    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 and total incidence rates of stroke among blacks. Stroke. 1998; 29: 415–421.

    Goldstein LB. Accuracy of ICD-9-CM coding for the identification of patients with acute ischemic stroke: effect of modifier codes. Stroke. 1998; 29: 1602–1604.

    Reker DM, Rosen AK, Hoenig H, Berlowitz DR, Laughlin J, Anderson L, Marshall CR, Rittman M. The hazards of stroke case selection using administrative data. Med Care. 2002; 40: 96–104.

    Collins TC, Petersen NJ, Menke TJ, Souchek J, Foster W, Ashton CM. Short-term, intermediate-term, and long-term mortality in patients hospitalized for stroke. J Clin Epidemiol. 2003; 56: 81–87.

    Benesch C, Witter DM Jr, Wilder AL, Duncan PW, Samsa GP, Matchar DB. Inaccuracy of the International Classification of Diseases (ICD-9-CM) in identifying the diagnosis of ischemic cerebrovascular disease. Neurology. 1997; 49: 660–664.

    Fagan SC, Morgenstern LB, Petitta A, Ward RE, Tilley BC, Marler JR, Levine SR, Broderick JP, Kwiatkowski TG, Frankel M, Brott TG, Walker MD. Cost-effectiveness of tissue plasminogen activator for acute ischemic stroke: NINDS rt-PA stroke study. Neurology. 1998; 50: 883–890.

    Kennedy J, Newcommon NJ, Cole-Haskayne A, et al. Organized stroke care increases the numbers and the speed with which patients with acute ischemic stroke are thrombolyzed. Stroke. 2002; 33: 354.

    Pearson TA, Blair SN, Daniels SR, Eckel RH, Fair JM, Fortmann SP, Franklin BA, Goldstein LB, Greenland P, Grundy SM, Hong Y, Miller NH, Lauer RM, Ockene IS, Sacco RL, Sallis JF Jr, Smith SC Jr, Stone NJ, Taubert KA. AHA guidelines for primary prevention of cardiovascular disease and stroke: 2002 update: consensus panel guide to comprehensive risk reduction for adult patients without coronary or other atherosclerotic vascular diseases. Circulation. 2002; 106: 388–391.

    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 1: proceedings of an expert panel. Stroke. 2003; 34: 151–156.

    Joint Commission on Accreditation of Healthcare Organizations. Primary Stroke Centers. Available at: http://www.jcaho.org/dscc/psc/pm_stroke.htm. Accessed March 14, 2005.

    Schwamm LH, Pancioli A, Acker JE III, Goldstein LB, Zorowitz RD, Shephard TJ, Moyer P, Gorman M, Johnston SC, Duncan PW, Gorelick P, Frank J, Stranne SK, Smith R, Federspiel W, Horton KB, Magnis E, Adams RJ; American Stroke Association’s Task Force on the Development of Stroke Systems. Recommendations for the establishment of stroke systems of care: recommendations from the American Stroke Association’s Task Force on the Development of Stroke Systems. Stroke. 2005; 36: 690–703.

    Smaha LA, American Heart Association. The American Heart Association Get with the Guidelines program. Am Heart J. 2004; 148 (5 suppl): S46–S48.