(Stroke. 1997;28:1181-1184.)
© 1997 American Heart Association, Inc.
Articles |
Retrospective Assessment of Initial Stroke Severity With the Canadian Neurological Scale
From the Division of Neurology, Durham Department of Veterans Affairs Medical Center, and the Division of Neurology, Department of Medicine, and Center for Health Policy Research and Education (L.B.G.), Duke University, Durham, NC.
Correspondence to Larry B. Goldstein, MD, Box 3651, Duke University Medical Center, Durham, NC 27710. E-mail golds004{at}mc.duke.edu
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionAppendix 1References |
|---|
Background and Purpose The severity of the initial neurological deficit is a critical determinant of outcome after acute stroke. Retrospective outcome studies are generally limited by a lack of quantitative data relating to this initial stroke severity. We evaluated the validity and reliability of measuring initial stroke severity retrospectively with the Canadian Neurological Scale (CNS).
Methods The CNS was used to prospectively score the initial neurological deficit in a series of patients with acute ischemic stroke (n=24). An algorithm was devised for applying the CNS retrospectively on the basis of information in the patient's hospital discharge summary. Those dictating the discharge summaries were not aware of the study, and the retrospective scoring was performed without reference to other scores. The level of agreement between the prospective and retrospective scores (validity) and both intraobserver and interobserver reliability for the retrospective scores were determined.
Results Agreement was high between retrospective and
prospective scores (r=.84, R2=.71,
P<.0001), between two sets of retrospective scores obtained
by one rater (r=.95, R2=.91,
P<.0001), and between retrospective scores obtained by
different raters (r=.91, R2=.82,
P<.0001). Weighted kappa statistics (
w) for
prospectively versus retrospectively scored items varied from almost
perfect (w >0.81 for level of consciousness and
orientation) to substantial (w=0.68 for speech) and
moderate (w=0.41 to 0.60 for facial weakness, proximal
arm, distal arm, proximal leg, and distal leg strength). Using the
retrospective algorithm, there was almost perfect intraobserver and
interobserver reliability for each of the individual CNS items
(w=0.81 to 1.00).
Conclusions These data show that retrospective scoring of initial stroke severity using an algorithm based on the CNS is valid and can be reliably performed using information available in hospital discharge summaries.
Key Words: outcome • prognosis • retrospective studies • stroke assessment
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionAppendix 1References |
|---|
Although a variety of factors may influence the rate and ultimate degree of recovery after stroke, the single most important outcome variable is the severity of the initial neurological deficit.1 2 3 4 5 However, outcome studies frequently rely on retrospective data. The inferences and conclusions that can be drawn from these types of studies are limited because retrospective quantitative data relating to the severity of the initial neurological deficit are generally unavailable. For example, in several recent studies, the authors indicated that a lack of data relating to the initial severity of the neurological deficit was a limitation of the study design, or they were forced to infer that initial stroke severity was similar between groups with differing outcomes.6 7 8 9 A valid and reliable instrument that can be used to measure initial stroke severity retrospectively is currently unavailable.
The Canadian Neurological Scale (CNS) is a highly reliable and validated stroke scoring system.10 11 Long-term outcome can be predicted soon after acute stroke with a simple mathematical model based on the patient's age and initial CNS score.12 Whereas other stroke impairment scales that are in common use include assessments of deficits such as neglect, coordination, sensation, and gait that may be variably recorded in the hospital discharge summary,13 14 15 16 the CNS focuses on level of consciousness, speech, and strength. Because impairments of these modalities are basic to the evaluation of any stroke patient, the data required for the retrospective application of the CNS are more likely to be recorded in the discharge note. We developed and assessed the validity and reliability of an algorithm designed to permit the CNS to be applied retrospectively on the basis of information routinely available in patients' hospital discharge summaries.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionAppendix 1References |
|---|
On the basis of a review of a series of hospital discharge summaries, an algorithm was empirically devised for the retrospective scoring of initial stroke severity using the CNS (see "Appendix"). To test the validity of this retrospective algorithm, a consecutive series of 24 patients with acute ischemic stroke was prospectively rated at the time of hospital admission with the CNS by neurology resident physicians who were unaware of the study. On the basis of information available in the patients' routine hospital discharge summaries as dictated by the resident physician caring for the patient, the admission CNS was later assessed independently by a rater (L.B.G.) who was not involved in the patients' care and without reference to the prospectively obtained scores. Those dictating the hospital discharge summaries were unaware of the study. The validity of the retrospective assessment was assessed by comparing prospectively and retrospectively obtained scores for each patient with linear regression analysis. Weighted kappa statistics (
w) were also calculated to determine the
level of agreement between prospectively and retrospectively obtained
scores for individual items constituting the CNS.17 The
weighted kappa score measures the agreement among observers adjusted
for the amount of agreement expected by chance and the magnitude of
disagreements.18 To determine the level of interrater reliability of the retrospectively obtained scores, a second observer (V.C.) independently scored each of the 24 patients. To determine the level of intrarater reliability, each case was rated a second time by one observer without reference to the initial set of scores. In each case, reliability was assessed with linear regression analysis. Weighted kappa statistics were calculated to determine the level of agreement for retrospectively obtained individual CNS items.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionAppendix 1References |
|---|
Patients were a mean of 66±2 years old and all were men; 67% were white and 33% were black. According to clinical and radiographic (brain CT and/or MRI scan) data, 29% of the 24 patients had strokes in the distribution of a small penetrating vessel; 25% of strokes were in a middle cerebral, 17% vertebrobasilar, 8% posterior cerebral, and 12% in an uncertain arterial distribution. Eight percent of the strokes were associated with carotid artery occlusions. Based on the prospective evaluations, 21% of the patients had a depressed level of consciousness, 17% were disoriented, 42% had a speech disturbance, 42% had facial weakness, 58% had proximal arm weakness, 46% had distal arm weakness, 48% had proximal leg weakness, and 24% had distal leg weakness. Although no patients were comatose on admission, the scores otherwise encompass the full range of possible scores for each item of the CNS. Fig 1
gives the
distribution of total CNS scores for each of the patients based on
prospective and retrospective assessments.
|
Except for two patients, all of the data necessary to apply the CNS retrospectively were available in each of the dictated discharge summaries. The evaluations of speech and facial strength were not specifically recorded in two of the notes. Three of the patients had limb amputations.
Fig 1
gives the regression analysis for the prospectively and
retrospectively obtained CNS scores. More than 70% of the variance in
the retrospective scores is explained by the prospective scores. The
Table gives the levels of agreement between
prospectively and retrospectively obtained scores for individual items
constituting the CNS. Agreement ranged from almost perfect
(w >0.81) for level of consciousness and orientation to
substantial (w=0.61 to 0.80) for speech and moderate
(w=0.41 to 0.60) for facial weakness, proximal arm,
distal arm, proximal leg, and distal leg
strength.17
|
Fig 2 (top) gives the regression analysis for
the scores obtained by two independent raters (interrater reliability).
The scores obtained by one rater explain more than 80% of the variance
in the scores obtained by the second rater. The largest discrepancy
between the two observers was 3 points. In this case, one rater scored
the proximal and distal limb strengths of a patient with a leg
amputation as "no weakness" (score, 1.5 proximal strength and 1.5
distal strength). The second rater scored the proximal and distal leg
strengths of this patient as "total paralysis" (score, 0 proximal
strength and 0 distal strength). Based on the algorithm, amputated
limbs are not scored (score 0).
|
Fig 2 (bottom) also gives the regression analysis for the
two independent sets of ratings by the same observer (intrarater
reliability). The first set of scores explains more than 90% of the
variance in the second set of scores. The largest discrepancy was 3
points due to rating a patient with a leg amputation as having "no
weakness" in the limb on one occasion and as having "total
paralysis" of the limb on the second occasion. With the
retrospective algorithm, there was almost perfect interrater and
intrarater reliability for each of the individual CNS items
(w=0.81 to 1.00, data not shown).
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionAppendix 1References |
|---|
These data demonstrate a high level of overall agreement between retrospectively and prospectively obtained CNS scores, supporting the validity of the retrospective scoring algorithm. Although the level of agreement between retrospective and prospective scores for speech, orientation, and level of consciousness was substantial or almost perfect, the level of agreement for limb-strength items was moderate (Table
). This likely occurred because the results of
strength testing for the specified muscle groups were not always
specifically indicated in the hospital discharge summaries (ie,
strength in an entire extremity may have been rated as "4/5").
The reliability of individual motor strength items might be further
improved by reference to the hospital admitting note, where the results
of more detailed strength testing may be recorded. These data also show that the CNS rating of initial stroke severity can be reconstructed reliably from retrospective data. The major source of disagreement was related to errors in coding strength in patients with amputations. This potential source of error could be eliminated if a separate category for coding limb amputation was included for each of the strength testing items.
One limitation of the present analysis is that all of the included patients were admitted to a neurology service or assessed by neurology residents in a single academic center. This may have biased the type of information included in the hospital discharge summaries. However, the types of data required to apply the CNS retrospectively are basic to the clinical evaluation of any stroke patient and would be expected to be reported in the hospital admission note even if not indicated in the discharge summary. Another potential limitation is that the study is based on the assessments of a relatively small, albeit consecutive, series of patients. One third of the patients were black, and because the study was conducted at a Veterans Administration hospital, all of the patients were men. Although these demographic biases would not be expected to affect the assessment of stroke severity, it would be desirable to extend these data to other centers and larger numbers of patients.
Many retrospective studies have attempted to adjust for general severity of illness through the application of a comorbidity index.19 20 However, a retrospective comorbidity index does not measure the severity of the initial neurological impairment resulting from a stroke. These data are critical for the interpretation of any study measuring utilization of resources, costs of care, and outcome. The availability of a valid and reliable instrument to retrospectively measure initial stroke severity will enhance the scope and potential applicability of these types of studies.
|
Appendix 1 |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionAppendix 1References |
|---|
|
Received January 27, 1997; revision received March 20, 1997; accepted April 8, 1997.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionAppendix 1References |
|---|
1. Goldstein LB, Matchar DB, Morgenlander JC, Davis JN. The influence of drugs on the recovery of sensorimotor function after stroke. J Neurol Rehabil. 1990;4:137-144.
2.
Heinemann AW, Roth EJ, Cichowski K, Betts HB.
Multivariate analysis of improvement and
outcome following stroke rehabilitation. Arch
Neurol. 1989;44:1167-1172.
3.
Lincoln NB, Blackburn M, Ellis S, Jackson J, Edmans
JA, Nouri FM, Walrer MF, Haworth H. An investigation of factors
affecting progress of patients on a stroke unit. J
Neurol Neurosurg Psychiatry. 1989;52:493-496.
4. Oxbury JM, Greenhall RCD, Grainger KMR. Predicting the outcome of stroke: acute stage after cerebral infarction. Br Med J. 1975;3:125-127.
5. Goldstein LB. Sygen in Acute Stroke Study Investigators: common drugs may influence motor recovery after stroke. Neurology. 1995;45:865-871.[Abstract]
6.
Bowen J, Yaste C. Effect of a stroke protocol
on hospital costs of stroke patients. Neurology. 1994;44:1961-1964.
7. Holloway RG, Witter DM Jr, Lawton KB, Lipscomb J, Samsa G. Inpatient costs of specific cerebrovascular events at five academic medical centers. Neurology. 1996;46:854-860.[Medline] [Order article via Infotrieve]
8.
Webb DJ, Fayad PB, Wilbur C, Thomas A, Brass
LM. Effects of a specialized team on stroke care: the first two
years of the Yale Stroke Program. Stroke. 1995;26:1353-1357.
9.
Mitchell JB, Ballard DJ, Whisnant JP, Ammering CJ,
Samsa GP, Matchar DB. What role do neurologists play in
determining the costs and outcomes of stroke patients?
Stroke. 1996;27:1937-1943.
10.
Cote R, Battista RN, Wolfson C, Boucher J, Adams J,
Hachinski VC. The Canadian Neurological Scale: validation and
reliability assessment. Neurology. 1989;39:638-643.
11.
Cote R, Hachinski VC, Shurvell BL, Norris JW, Wolfson
C. The Canadian Neurological Scale: a preliminary study in acute
stroke. Stroke. 1986;17:731-737.
12.
Fiorelli M, Alperovitch A, Argentino C, Sacchetti ML,
Toni D, Sette G, Cavalletti C, Gori MC, Fieschi C. Prediction of
long-term outcome in the early hours following acute ischemic
stroke. Arch Neurol. 1995;52:250-255.
13. Hantson L, De Weerdt W, De Keyser J, Diener HC, Franke C, Palm R, Van Orshoven M, Schoonderwalt H, De Klippel N, Herroelen L, Feys H. The European Stroke Scale. Stroke. 1994;25:2215-2219.[Abstract]
14.
Brott T, Adams HP Jr, Olinger CP, Marler JR, Barsan WG,
Biller J, Spilker J, Holleran R, Eberle R, Hertzberg V, Rorick M,
Moomaw CJ, Walker M. Measurements of acute cerebral infarction:
a clinical examination scale. Stroke. 1989;20:864-870.
15. Mathew NT, Meyer JS, Rivera VM, Charney JZ, Hartmann A. Double-blind evaluation of glycerol therapy in acute cerebral infarction. Lancet. 1972;2:1327-1329.[Medline] [Order article via Infotrieve]
16. Lindenstrom E, Boysen G, Christiansen LW, Hansen BR, Nielsen PW. Reliability of Scandinavian Neurological Stroke Scale. Cerebrovasc Dis. 1991;1:103-107.
17. Kramer MS, Feinstein AR. Clinical biostatistics, LIV: the biostatistics of concordance. Clin Pharmacol Ther. 1983;29:111-123.
18. Cohen J. Weighted kappa: nominal scale agreement with a provision for scaled disagreement or partial credit. Psychol Bull. 1968;70:213-220.
19. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373-383.[Medline] [Order article via Infotrieve]
20. Romano PS, Roos LL, Jollis JG. Adapting a clinical comorbidity index for use with ICD-9-CM administrative data: differing perspectives. J Clin Epidemiol. 1993;46:1075-1079.[Medline] [Order article via Infotrieve]
This article has been cited by other articles:
 |
|
 |
|
  C. A. Smith Hammond, L. B. Goldstein, R. D. Horner, J. Ying, L. Gray, L. Gonzalez-Rothi, and D. C. Bolser Predicting Aspiration in Patients With Ischemic Stroke: Comparison of Clinical Signs and Aerodynamic Measures of Voluntary Cough Chest, March 1, 2009; 135(3): 769 - 777. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. M. Nelson, M. A. Smith, B. C. Martinson, A. Kind, and R. V. Luepker Declining Patient Functioning and Caregiver Burden/Health: The Minnesota Stroke Survey-Quality of Life After Stroke Study Gerontologist, October 1, 2008; 48(5): 573 - 583. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. D. Bushnell, J. Griffin, L. K. Newby, L. B. Goldstein, K. W. Mahaffey, C. A. Graffagnino, R. A. Harrington, H. D. White, R. J. Simes, R. M. Califf, et al. Statin Use and Sex-Specific Stroke Outcomes in Patients With Vascular Disease Stroke, June 1, 2006; 37(6): 1427 - 1431. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A. Smith Hammond and L. B. Goldstein Cough and Aspiration of Food and Liquids Due to Oral-Pharyngeal Dysphagia: ACCP Evidence-Based Clinical Practice Guidelines Chest, January 1, 2006; 129(1_suppl): 154S - 168S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I-P. Hsueh, W.-C. Wang, C.-F. Sheu, and C.-L. Hsieh Rasch Analysis of Combining Two Indices to Assess Comprehensive ADL Function in Stroke Patients Stroke, March 1, 2004; 35(3): 721 - 726. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D.J. Anderson, L.B. Goldstein, W.E. Wilkinson, G.R. Corey, C.H. Cabell, L.L. Sanders, and D.J. Sexton Stroke location, characterization, severity, and outcome in mitral vs aortic valve endocarditis Neurology, November 25, 2003; 61(10): 1341 - 1346. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I-P. Hsueh, C.-H. Wang, C.-F. Sheu, and C.-L. Hsieh Comparison of Psychometric Properties of Three Mobility Measures for Patients With Stroke Stroke, July 1, 2003; 34(7): 1741 - 1745. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N U Weir, C E Counsell, M McDowall, A Gunkel, and M S Dennis Reliability of the variables in a new set of models that predict outcome after stroke J. Neurol. Neurosurg. Psychiatry, April 1, 2003; 74(4): 447 - 451. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. C. Hobart, L. S. Williams, K. Moran, and A. J. Thompson Quality of Life Measurement After Stroke: Uses and Abuses of the SF-36 Stroke, May 1, 2002; 33(5): 1348 - 1356. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Stavem and O.M. Ronning Survival of unselected stroke patients in a stroke unit compared with conventional care QJM, March 1, 2002; 95(3): 143 - 152. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Weih, N. Amberger, S. Wegener, U. Dirnagl, T. Reuter, and K. Einhaupl Sulfonylurea Drugs Do Not Influence Initial Stroke Severity and In-Hospital Outcome in Stroke Patients With Diabetes Stroke, September 1, 2001; 32(9): 2029 - 2032. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. D. Bushnell, G. P. Samsa, and L. B. Goldstein Hormone replacement therapy and ischemic stroke severity in women: A case-control study Neurology, May 22, 2001; 56(10): 1304 - 1307. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. D. Bushnell, D. C.C. Johnston, and L. B. Goldstein Retrospective Assessment of Initial Stroke Severity : Comparison of the NIH Stroke Scale and the Canadian Neurological Scale Stroke, March 1, 2001; 32(3): 656 - 660. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C.A. S. Hammond, L.B. Goldstein, D.J. Zajac, L. Gray, P.W. Davenport, and D.C. Bolser Assessment of aspiration risk in stroke patients with quantification of voluntary cough Neurology, February 27, 2001; 56(4): 502 - 506. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Lopez-Yunez, A. Bruno, L. S. Williams, E. Yilmaz, C. Zurru, and J. Biller Protocol Violations in Community-Based rTPA Stroke Treatment Are Associated With Symptomatic Intracerebral Hemorrhage Stroke, January 1, 2001; 32(1): 12 - 16. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. B. Bosworth, R. D. Horner, L. J. Edwards, and D. B. Matchar Depression and Other Determinants of Values Placed on Current Health State by Stroke Patients : Evidence From the VA Acute Stroke (VASt) Study Stroke, November 1, 2000; 31(11): 2603 - 2609. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. S. Williams, E. Y. Yilmaz, and A. M. Lopez-Yunez Retrospective Assessment of Initial Stroke Severity With the NIH Stroke Scale Stroke, April 1, 2000; 31(4): 858 - 862. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. R. Jones, R. D. Horner, L. J. Edwards, J. Hoff, S. B. Armstrong, C. A. Smith-Hammond, D. B. Matchar, and E. Z. Oddone Racial Variation in Initial Stroke Severity Stroke, March 1, 2000; 31(3): 563 - 567. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. S. Williams, M. Weinberger, L. E. Harris, and J. Biller Measuring quality of life in a way that is meaningful to stroke patients Neurology, November 1, 1999; 53(8): 1839 - 1839. [Abstract] [Full Text]
|
|
|
|
|
|
M. Weih, K. Kallenberg, A. Bergk, U. Dirnagl, L. Harms, K. D. Wernecke, and K. M. Einhaupl Attenuated Stroke Severity After Prodromal TIA : A Role for Ischemic Tolerance in the Brain? Stroke, September 1, 1999; 30(9): 1851 - 1854. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. S. Williams, M. Weinberger, L. E. Harris, D. O. Clark, and J. Biller Development of a Stroke-Specific Quality of Life Scale Stroke, July 1, 1999; 30(7): 1362 - 1369. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||