Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Comparing Hospitals on Stroke Care: The Need to Account for Stroke Severity

Comparing Hospitals on Stroke Care: The Need to Account for Stroke Severity Stroke is a leading cause of morbidity and mortality in the United States and worldwide and is associated with enormous health care expenditures.1 Approximately 800 000 new or recurrent strokes occur annually in the United States, and of these, about 87% are ischemic cerebral infarctions.1 Since 1996, when the use of intravenous tissue plasminogen activator was approved by the US Food and Drug Administration (FDA) for the treatment of patients with acute ischemic stroke within 3 hours of symptom onset, there has been a sea change in the approach to identification and management of stroke patients to improve their outcomes. Parallel to these advances, there has been an equally important move toward a systems-based approach to stroke care. The very short time window needed for acute therapies to reverse brain injury has inspired several national and statewide initiatives to improve hospital care of the stroke patient. The Brain Attack Coalition criteria for primary stroke centers, first recommended in 2000 and revised in 2011, have formed the basis for Joint Commission certification for stroke center status.2 Additional certification for comprehensive stroke centers begins in July 2012.3 These efforts are rooted in a developing evidence base demonstrating that organized, systems-based care, informed by guidelines and quality assurance efforts, can improve outcomes. There is evidence, for example, that patients with acute stroke are more likely to survive, return home, and regain independence if treated in hospital units specializing in the care of patients with stroke.4 The restructuring of health insurance in the United States, with the focus on increasing efficiency, improving outcomes, and providing value to the public, has stimulated efforts to provide valid measures of health care quality. One such measure includes assessment and ranking of hospital performance in the care of commonly encountered and significant illnesses, of which stroke is a good example. Approaches to the measurement of such outcomes, however, are potentially fraught with biases and other complexities. Although the implications of such measures are substantial, both for the individual hospital and the health care system as a whole, targeted research about health care–related outcomes is a relatively new field and optimal analytic approaches are still being developed. In this issue of JAMA, Fonarow and colleagues5 evaluate the influence of including or excluding stroke severity in prognostic stroke outcome models in a large Medicare insurance database. The authors used data from almost 128 000 patients with ischemic stroke from 782 Get With The Guidelines–Stroke participating hospitals. For all patients included in the analysis, information on the severity of the stroke was available in the form of the National Institutes of Health Stroke Scale (NIHSS) score. The primary outcome was 30-day mortality, and prognostic models were evaluated for their overall model discrimination as well as for differences in rankings of the hospital performance when including or excluding information on stroke severity. The main prognostic model included information on age, sex, prior stroke or transient ischemic attack, and a large number of comorbid conditions and was compared with a model that additionally included information on stroke severity. All statistical measures of model performance showed that the model including stroke severity was superior, indicating that stroke severity substantially improved the prediction of 30-day mortality above and beyond other clinical predictors. Considered at the level of the individual patient, this result does not seem surprising: a patient who has a relatively more severe stroke is at increased risk of death. Moreover, when the prediction models were used to classify and rank hospitals based on 30-day mortality, the model including stroke severity demonstrated substantially more accurate classification and substantially changed hospital rankings. Of the 782 participating hospitals, the absolute change of the median hospital rank position was 79 places. More than half (58%) of hospitals first classified as having higher than expected mortality were reclassified to having the expected mortality rate after incorporating severity into the model. Further, among hospitals in the top fifth percentile, 41% (16 of 39) were newly identified when using the model including stroke severity compared with a model without. Overall, when considering hospitals ranked in the top 20% and bottom 20%, close to one-third of 294 hospitals would have been reclassified. These results suggest the importance of carefully assessing case mix at different hospitals before classifying them as high or low performing. Several aspects should be borne in mind when interpreting the results of the study by Fonarow et al. From a methodological point of view, it is important to consider the setup and interplay of variables included in prognostic models. In their analysis, Fonarow et al included information on stroke severity as a single covariate in the model. Whether interactions between stroke severity and other comorbid conditions would further improve the prognostic model performance remains unclear. In models evaluating the association between acute treatment and mortality after ischemic stroke, the results can substantially differ if the underlying interrelationship between treatment and stroke severity is ignored,6 which may be part of the reason thrombolysis in patients with ischemic stroke has been associated with increased risk of mortality in observational studies.7 In some cases, particularly when there are significant interactions with other patient characteristics, it may be better to stratify patients by stroke severity at baseline, rather than adjust for it in a prediction model. For studies evaluating mortality after stroke, it also may be of importance to consider the number of patients with acute ischemic stroke who are treated at an individual hospital, as this factor influences prediction of mortality after stroke.8 The effects of a stroke depend strongly on its location: a small area of necrosis in a critical brain location can cause a large deficit. The NIHSS score is weighted toward deficits in language and cognition; thus, patients with left hemispheric stroke will consistently score worse on the NIHSS than patients with right hemispheric stroke, even when stroke volume is identical. The clinical outcomes, however, are no worse or better among patients with left hemispheric stroke.9 Stroke severity also may change substantially within a few hours or days after symptom onset, and the performance of the NIHSS depends on the time from stroke.10 Such considerations, however, would likely have biased the results of this study toward finding less influence of stroke severity, further emphasizing the importance of including at least some measure of stroke severity. Further studies may better refine the use and timing of measures of stroke severity. Another consideration is that data on stroke severity for this analysis were collected from hospitals involved with the Get With The Guidelines–Stroke program. These hospitals all have an interest in stroke, and hospital personnel have been trained to collect the NIHSS score. At other hospitals, stroke severity may not be routinely collected or may be collected less accurately. It is also plausible that some centers might systematically overestimate the severity of their stroke patients if severity became an important measure for ranking the hospitals. In addition, mortality should be regarded as only one important measure of quality of care after stroke. The effect of ischemic stroke often manifests more as disability than mortality, and many patients are left disabled but alive. At 6 months, 30% of stroke survivors are unable to walk without assistance, 26% are dependent on others for activities of daily living, 35% have symptoms of depression, and 26% are institutionalized.1 Even though mortality is relatively easy to measure, it may not fully capture the ability of the hospital to improve the functional outcomes of patients with stroke. Incorporation of patient-centered outcomes may be particularly important for a disease like stroke. Despite these considerations, the results of the study by Fonarow and colleagues5 clearly highlight the importance of incorporating information on stroke severity when conducting health outcomes research in stroke. Excluding this information will lead to incorrect ranking of hospital performance by failing to consider that hospitals care for different patient populations. The influence of stroke severity on these outcome measures, moreover, seems different from that of measures of severity in other conditions. For other cardiovascular diseases, risk adjustment using demographic characteristics and claims-derived comorbid conditions may sufficiently account for the underlying case mix.11 Ischemic stroke is a much more heterogeneous condition than ischemic heart disease and is characterized by multiple subtypes, etiologies, and diverse outcomes. The assumption that what is true of myocardial infarction is also true of stroke, therefore, is flawed, as the present data underscore. The particular characteristics of stroke have to be taken into consideration by clinicians, insurance companies, and policy makers when comparing disease-specific health outcomes. Back to top Article Information Corresponding Author: Tobias Kurth, MD, ScD, INSERM U708—Neuroepidemiology, University Bordeaux Segalen, 146 rue Leo Saignat, Case 11, 33076 Bordeaux CEDEX, France (tobias.kurth@univ-bordeaux.fr). Conflict of Interest Disclosures: Both authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Kurth reported having received investigator-initiated research funding from the French National Research Agency, the US National Institutes of Health (NIH), the Migraine Research Foundation, and the Parkinson's Disease Foundation and having received honoraria from the BMJ for editorial services; from Allergan, the American Academy of Neurology, and Merck for educational lectures; and from MAP Pharmaceutical for contributing to a scientific advisory panel. Dr Elkind reported receiving research support from diaDexus, Bristol-Myers Squibb/Sanofi Pharmaceuticals Partnership, and the NIH; serving on an event adjudication committee for Jarvik Heart; having received compensation for participation in litigation on behalf of Novartis, Organon, and GlaxoSmithKline; and receiving compensation from the American Academy of Neurology for service as Resident and Fellow Section Editor for the journal Neurology. Editorials represent the opinions of the authors and JAMA and not those of the American Medical Association. References 1. Roger VL, Go AS, Lloyd-Jones DM, et al; American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Heart disease and stroke statistics: 2012 update: a report from the American Heart Association. Circulation. 2012;125(1):e2-e22022179539PubMedGoogle ScholarCrossref 2. Alberts MJ, Latchaw RE, Jagoda A, et al; Brain Attack Coalition. Revised and updated recommendations for the establishment of primary stroke centers: a summary statement from the Brain Attack Coalition. Stroke. 2011;42(9):2651-266521868727PubMedGoogle ScholarCrossref 3. Alberts MJ, Latchaw RE, Selman WR, et al; Brain Attack Coalition. Recommendations for comprehensive stroke centers: a consensus statement from the Brain Attack Coalition. Stroke. 2005;36(7):1597-161615961715PubMedGoogle ScholarCrossref 4. Govan L, Weir CJ, Langhorne P.for the Stroke Unit Trialists' Collaboration. Organized inpatient (stroke unit) care for stroke. Stroke. 2008;39:2402-240318556580PubMedGoogle ScholarCrossref 5. Fonarow GC, Pan W, Saver JL, et al. Comparison of 30-day mortality models for profiling hospital performance in acute ischemic stroke with vs without adjustment for stroke severity. JAMA. 2012;308(3):joc120053257-264Google ScholarCrossref 6. Kurth T, Walker AM, Glynn RJ, et al. Results of multivariable logistic regression, propensity matching, propensity adjustment, and propensity-based weighting under conditions of nonuniform effect. Am J Epidemiol. 2006;163(3):262-27016371515PubMedGoogle ScholarCrossref 7. Dubinsky R, Lai SM. Mortality of stroke patients treated with thrombolysis: analysis of nationwide inpatient sample. Neurology. 2006;66(11):1742-174416769953PubMedGoogle ScholarCrossref 8. Heuschmann PU, Kolominsky-Rabas PL, Roether J, et al; German Stroke Registers Study Group. Predictors of in-hospital mortality in patients with acute ischemic stroke treated with thrombolytic therapy. JAMA. 2004;292(15):1831-183815494580PubMedGoogle ScholarCrossref 9. Elkind MS, Prabhakaran S, Pittman J, Koroshetz W, Jacoby M, Johnston KC.GAIN Americas Investigators. Sex as a predictor of outcomes in patients treated with thrombolysis for acute stroke. Neurology. 2007;68(11):842-84817353472PubMedGoogle ScholarCrossref 10. Olavarría VV, Delgado I, Hoppe A, et al. Validity of the NIHSS in predicting arterial occlusion in cerebral infarction is time-dependent. Neurology. 2011;76(1):62-6821205696PubMedGoogle ScholarCrossref 11. Krumholz HM, Normand SL. Public reporting of 30-day mortality for patients hospitalized with acute myocardial infarction and heart failure. Circulation. 2008;118(13):1394-139718725492PubMedGoogle ScholarCrossref http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png JAMA American Medical Association

Comparing Hospitals on Stroke Care: The Need to Account for Stroke Severity

Kurth, Tobias; Elkind, Mitchell S. V.
JAMA , Volume 308 (3) – Jul 18, 2012
Download PDF
3 pages

Loading next page...
 
/lp/american-medical-association/comparing-hospitals-on-stroke-care-the-need-to-account-for-stroke-40Ck3iqb7I

References (13)

Publisher
American Medical Association
Copyright
Copyright © 2012 American Medical Association. All Rights Reserved.
ISSN
0098-7484
eISSN
1538-3598
DOI
10.1001/jama.2012.8448
Publisher site
See Article on Publisher Site

Abstract

Stroke is a leading cause of morbidity and mortality in the United States and worldwide and is associated with enormous health care expenditures.1 Approximately 800 000 new or recurrent strokes occur annually in the United States, and of these, about 87% are ischemic cerebral infarctions.1 Since 1996, when the use of intravenous tissue plasminogen activator was approved by the US Food and Drug Administration (FDA) for the treatment of patients with acute ischemic stroke within 3 hours of symptom onset, there has been a sea change in the approach to identification and management of stroke patients to improve their outcomes. Parallel to these advances, there has been an equally important move toward a systems-based approach to stroke care. The very short time window needed for acute therapies to reverse brain injury has inspired several national and statewide initiatives to improve hospital care of the stroke patient. The Brain Attack Coalition criteria for primary stroke centers, first recommended in 2000 and revised in 2011, have formed the basis for Joint Commission certification for stroke center status.2 Additional certification for comprehensive stroke centers begins in July 2012.3 These efforts are rooted in a developing evidence base demonstrating that organized, systems-based care, informed by guidelines and quality assurance efforts, can improve outcomes. There is evidence, for example, that patients with acute stroke are more likely to survive, return home, and regain independence if treated in hospital units specializing in the care of patients with stroke.4 The restructuring of health insurance in the United States, with the focus on increasing efficiency, improving outcomes, and providing value to the public, has stimulated efforts to provide valid measures of health care quality. One such measure includes assessment and ranking of hospital performance in the care of commonly encountered and significant illnesses, of which stroke is a good example. Approaches to the measurement of such outcomes, however, are potentially fraught with biases and other complexities. Although the implications of such measures are substantial, both for the individual hospital and the health care system as a whole, targeted research about health care–related outcomes is a relatively new field and optimal analytic approaches are still being developed. In this issue of JAMA, Fonarow and colleagues5 evaluate the influence of including or excluding stroke severity in prognostic stroke outcome models in a large Medicare insurance database. The authors used data from almost 128 000 patients with ischemic stroke from 782 Get With The Guidelines–Stroke participating hospitals. For all patients included in the analysis, information on the severity of the stroke was available in the form of the National Institutes of Health Stroke Scale (NIHSS) score. The primary outcome was 30-day mortality, and prognostic models were evaluated for their overall model discrimination as well as for differences in rankings of the hospital performance when including or excluding information on stroke severity. The main prognostic model included information on age, sex, prior stroke or transient ischemic attack, and a large number of comorbid conditions and was compared with a model that additionally included information on stroke severity. All statistical measures of model performance showed that the model including stroke severity was superior, indicating that stroke severity substantially improved the prediction of 30-day mortality above and beyond other clinical predictors. Considered at the level of the individual patient, this result does not seem surprising: a patient who has a relatively more severe stroke is at increased risk of death. Moreover, when the prediction models were used to classify and rank hospitals based on 30-day mortality, the model including stroke severity demonstrated substantially more accurate classification and substantially changed hospital rankings. Of the 782 participating hospitals, the absolute change of the median hospital rank position was 79 places. More than half (58%) of hospitals first classified as having higher than expected mortality were reclassified to having the expected mortality rate after incorporating severity into the model. Further, among hospitals in the top fifth percentile, 41% (16 of 39) were newly identified when using the model including stroke severity compared with a model without. Overall, when considering hospitals ranked in the top 20% and bottom 20%, close to one-third of 294 hospitals would have been reclassified. These results suggest the importance of carefully assessing case mix at different hospitals before classifying them as high or low performing. Several aspects should be borne in mind when interpreting the results of the study by Fonarow et al. From a methodological point of view, it is important to consider the setup and interplay of variables included in prognostic models. In their analysis, Fonarow et al included information on stroke severity as a single covariate in the model. Whether interactions between stroke severity and other comorbid conditions would further improve the prognostic model performance remains unclear. In models evaluating the association between acute treatment and mortality after ischemic stroke, the results can substantially differ if the underlying interrelationship between treatment and stroke severity is ignored,6 which may be part of the reason thrombolysis in patients with ischemic stroke has been associated with increased risk of mortality in observational studies.7 In some cases, particularly when there are significant interactions with other patient characteristics, it may be better to stratify patients by stroke severity at baseline, rather than adjust for it in a prediction model. For studies evaluating mortality after stroke, it also may be of importance to consider the number of patients with acute ischemic stroke who are treated at an individual hospital, as this factor influences prediction of mortality after stroke.8 The effects of a stroke depend strongly on its location: a small area of necrosis in a critical brain location can cause a large deficit. The NIHSS score is weighted toward deficits in language and cognition; thus, patients with left hemispheric stroke will consistently score worse on the NIHSS than patients with right hemispheric stroke, even when stroke volume is identical. The clinical outcomes, however, are no worse or better among patients with left hemispheric stroke.9 Stroke severity also may change substantially within a few hours or days after symptom onset, and the performance of the NIHSS depends on the time from stroke.10 Such considerations, however, would likely have biased the results of this study toward finding less influence of stroke severity, further emphasizing the importance of including at least some measure of stroke severity. Further studies may better refine the use and timing of measures of stroke severity. Another consideration is that data on stroke severity for this analysis were collected from hospitals involved with the Get With The Guidelines–Stroke program. These hospitals all have an interest in stroke, and hospital personnel have been trained to collect the NIHSS score. At other hospitals, stroke severity may not be routinely collected or may be collected less accurately. It is also plausible that some centers might systematically overestimate the severity of their stroke patients if severity became an important measure for ranking the hospitals. In addition, mortality should be regarded as only one important measure of quality of care after stroke. The effect of ischemic stroke often manifests more as disability than mortality, and many patients are left disabled but alive. At 6 months, 30% of stroke survivors are unable to walk without assistance, 26% are dependent on others for activities of daily living, 35% have symptoms of depression, and 26% are institutionalized.1 Even though mortality is relatively easy to measure, it may not fully capture the ability of the hospital to improve the functional outcomes of patients with stroke. Incorporation of patient-centered outcomes may be particularly important for a disease like stroke. Despite these considerations, the results of the study by Fonarow and colleagues5 clearly highlight the importance of incorporating information on stroke severity when conducting health outcomes research in stroke. Excluding this information will lead to incorrect ranking of hospital performance by failing to consider that hospitals care for different patient populations. The influence of stroke severity on these outcome measures, moreover, seems different from that of measures of severity in other conditions. For other cardiovascular diseases, risk adjustment using demographic characteristics and claims-derived comorbid conditions may sufficiently account for the underlying case mix.11 Ischemic stroke is a much more heterogeneous condition than ischemic heart disease and is characterized by multiple subtypes, etiologies, and diverse outcomes. The assumption that what is true of myocardial infarction is also true of stroke, therefore, is flawed, as the present data underscore. The particular characteristics of stroke have to be taken into consideration by clinicians, insurance companies, and policy makers when comparing disease-specific health outcomes. Back to top Article Information Corresponding Author: Tobias Kurth, MD, ScD, INSERM U708—Neuroepidemiology, University Bordeaux Segalen, 146 rue Leo Saignat, Case 11, 33076 Bordeaux CEDEX, France (tobias.kurth@univ-bordeaux.fr). Conflict of Interest Disclosures: Both authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Kurth reported having received investigator-initiated research funding from the French National Research Agency, the US National Institutes of Health (NIH), the Migraine Research Foundation, and the Parkinson's Disease Foundation and having received honoraria from the BMJ for editorial services; from Allergan, the American Academy of Neurology, and Merck for educational lectures; and from MAP Pharmaceutical for contributing to a scientific advisory panel. Dr Elkind reported receiving research support from diaDexus, Bristol-Myers Squibb/Sanofi Pharmaceuticals Partnership, and the NIH; serving on an event adjudication committee for Jarvik Heart; having received compensation for participation in litigation on behalf of Novartis, Organon, and GlaxoSmithKline; and receiving compensation from the American Academy of Neurology for service as Resident and Fellow Section Editor for the journal Neurology. Editorials represent the opinions of the authors and JAMA and not those of the American Medical Association. References 1. Roger VL, Go AS, Lloyd-Jones DM, et al; American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Heart disease and stroke statistics: 2012 update: a report from the American Heart Association. Circulation. 2012;125(1):e2-e22022179539PubMedGoogle ScholarCrossref 2. Alberts MJ, Latchaw RE, Jagoda A, et al; Brain Attack Coalition. Revised and updated recommendations for the establishment of primary stroke centers: a summary statement from the Brain Attack Coalition. Stroke. 2011;42(9):2651-266521868727PubMedGoogle ScholarCrossref 3. Alberts MJ, Latchaw RE, Selman WR, et al; Brain Attack Coalition. Recommendations for comprehensive stroke centers: a consensus statement from the Brain Attack Coalition. Stroke. 2005;36(7):1597-161615961715PubMedGoogle ScholarCrossref 4. Govan L, Weir CJ, Langhorne P.for the Stroke Unit Trialists' Collaboration. Organized inpatient (stroke unit) care for stroke. Stroke. 2008;39:2402-240318556580PubMedGoogle ScholarCrossref 5. Fonarow GC, Pan W, Saver JL, et al. Comparison of 30-day mortality models for profiling hospital performance in acute ischemic stroke with vs without adjustment for stroke severity. JAMA. 2012;308(3):joc120053257-264Google ScholarCrossref 6. Kurth T, Walker AM, Glynn RJ, et al. Results of multivariable logistic regression, propensity matching, propensity adjustment, and propensity-based weighting under conditions of nonuniform effect. Am J Epidemiol. 2006;163(3):262-27016371515PubMedGoogle ScholarCrossref 7. Dubinsky R, Lai SM. Mortality of stroke patients treated with thrombolysis: analysis of nationwide inpatient sample. Neurology. 2006;66(11):1742-174416769953PubMedGoogle ScholarCrossref 8. Heuschmann PU, Kolominsky-Rabas PL, Roether J, et al; German Stroke Registers Study Group. Predictors of in-hospital mortality in patients with acute ischemic stroke treated with thrombolytic therapy. JAMA. 2004;292(15):1831-183815494580PubMedGoogle ScholarCrossref 9. Elkind MS, Prabhakaran S, Pittman J, Koroshetz W, Jacoby M, Johnston KC.GAIN Americas Investigators. Sex as a predictor of outcomes in patients treated with thrombolysis for acute stroke. Neurology. 2007;68(11):842-84817353472PubMedGoogle ScholarCrossref 10. Olavarría VV, Delgado I, Hoppe A, et al. Validity of the NIHSS in predicting arterial occlusion in cerebral infarction is time-dependent. Neurology. 2011;76(1):62-6821205696PubMedGoogle ScholarCrossref 11. Krumholz HM, Normand SL. Public reporting of 30-day mortality for patients hospitalized with acute myocardial infarction and heart failure. Circulation. 2008;118(13):1394-139718725492PubMedGoogle ScholarCrossref

Journal

JAMA – American Medical Association

Published: Jul 18, 2012

Keywords: cerebrovascular accident,ischemic stroke

There are no references for this article.