http://informahealthcare.com/dre ISSN 0963-8288 print/ISSN 1464-5165 online Disabil Rehabil, 2014; 36(18): 1549–1554 ! 2014 Informa UK Ltd. DOI: 10.3109/09638288.2013.854840
ASSESSMENT PROCEDURES
Disabil Rehabil Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
Comparison of three instruments to assess changes of motor impairment in acute hemispheric stroke: the Stroke Impairment Assessment Set (SIAS), the National Institute of Health Stroke Scale (NIHSS) and the Canadian Neurological Scale (CNS) Masaru Seki1, Kimitaka Hase2, Hidetoshi Takahashi3, and Meigen Liu4 1
Department of Liberal Arts and Human Development, Kanagawa University of Human Services, Kanagawa, Japan, 2Department of Rehabilitation Medicine, Kansai Medical University, Osaka, Japan, 3Department of Rehabilitation Medicine, International Medical Center, Saitama Medical University, Saitama, Japan, and 4Department of Rehabilitation Medicine, Keio University School of Medicine, Tokyo, Japan Abstract
Keywords
Purpose: The purpose was to compare the performance of three representative instruments to measure changes of motor recovery with acute hemispheric stroke. Method: In 41 consecutive patients with acute hemispheric infarction, motor recovery was assessed within 3 days of onset and then every 2 weeks until the 12th week with the Stroke Impairment Assessment Set (SIAS), the National Institutes of Health stroke scale (NIHSS) and the Canadian Neurological Scale (CNS). We analyzed the relationships among the corresponding subscales of the three instruments with the Spearman’s rank correlation method, and compared their responsiveness by plotting the temporal profiles of scores of each instrument and by testing the significance of changes over time with the Friedman test. Results: High correlations were observed among the three instruments at each examination point. However, the SIAS scores were more widely dispersed at the same NIHSS and CNS scores. Friedman test revealed that the scores changed significantly during the observation period with the SIAS and the CNS but not with the NIHSS. The changes were detected later with the CNS than with the SIAS. Conclusions: Although the assessment results with the three instruments were highly inter-correlated, the SIAS performed better with respect to responsiveness to changes.
Hemiplegia, measurement, motor recovery, stroke History Received 18 March 2013 Revised 29 September 2013 Accepted 9 October 2013 Published online 18 November 2013
ä Implication for Rehabilitation
When providing rehabilitation services to patients with stroke, it is important to document objectively the level of their impairment from the acute stage and during the entire course of rehabilitation with standardized instrument, and to predict their functional outcomes as early and accurately as possible. This study therefore compared three representative instruments with established psychometric properties, the SIAS, the NIHSS and the CNS, with respect to their ability to document motor impairment and their responsiveness to recovery in patients with acute hemispheric stroke.
Introduction Stroke poses us important health and economic problems in many countries [1], and rehabilitation plays an important role in its management. When providing rehabilitation services to patients with stroke, it is important to document objectively the level of their impairment from the acute stage and during the entire course of rehabilitation with standardized instrument, and to predict their functional outcomes as early and as
Address for correspondence: Masaru Seki, MD, PhD, Department of Liberal Arts and Human Development, Kanagawa University of Human Services, 1-10-1 Heiseicho, Yokosuka, Kanagawa 238-8522, Japan. Tel: +81 46 828 2772. Fax: +81 46 828 2773. E-mail: [email protected]
accurately as possible. This is necessary to improve the quality of stroke rehabilitation programs while making the best use of the limited resources. For this purpose, a number of instruments have been proposed with varying degrees of descriptions of their psychometric properties, such as the Brunnstrom test [2] and the Motricity index for motor assessment [3], the Fugl– Meyer test [4], the National Institutes of Health Stroke Scale (NIHSS) [5], the Canadian Neurological Scale (CNS) [6], the Scandinavian Stroke Scale [7] and the Stroke Impairment Assessment Set (SIAS) [8] for comprehensive assessment of stroke impairment. However, it is not clear which instrument we are advised to use in daily clinical practice to evaluate patients with stroke. To answer this question, it is imperative to compare the
1550
M. Seki et al.
Disabil Rehabil, 2014; 36(18): 1549–1554
performance of each instrument to describe the degree of impairment and track changes over time. Although several studies are available that compared currently used stroke scales crosssectionally [9–14], no study has yet been reported that compares them longitudinally from the acute stage of onset in detail. The purpose of the present study was therefore to compare three representative instruments with established psychometric properties, the SIAS, the NIHSS and the CNS, with respect to their ability to document motor impairment and their responsiveness to recovery in patients with acute hemispheric stroke.
Disabil Rehabil Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
Methods The participants were 41 consecutive patients with CT- or MRIconfirmed cerebral hemispheric infarction (22 males and 19 females, 21 with right-sided and 20 with left-sided brain lesion, mean age: 62.7 11.1 years), who were admitted to one of the two acute care general hospitals, Odawara City Hospital (n ¼ 31) or Shizuoka City Hospital (n ¼ 10). To be included in the study, patients had to: (1) have measurable hemiparesis, (2) be hospitalized within 24 h of onset of neurological symptoms, (3) have no preexisting stroke deficits, (4) be able to understand the test tasks related to motor paralysis assessment and (5) be able to assume the sitting position at least with a support. Patients were excluded if they had: (1) hemorrhagic stroke to make them homogeneous with respect to stroke type, (2) disturbance of consciousness, (3) severe communication deficits, (4) motor apraxia, (5) unstable medical condition(s) and/or (6) other coexisting disabling condition(s) (e.g. arthritis, amputations). All the patients underwent a standard acute rehabilitation program together with acute medical care. Before enrollment, the purpose and procedures of the study were fully explained, and informed consent was obtained from the patients and their family members. This study followed principles in the Declaration of Helsinki. Within 3 days from stroke onset and every 2 weeks thereafter until the 12th week, the same board-certified physiatrist experienced in acute stroke rehabilitation assessed motor impairment of all patients with the motor items of the SIAS (SIAS-M), the NIHSS (NIH-M) and the CNS (CNS-M). Tables 1 and 2 illustrate the motor items and scaling of each instrument. The details of the assessment methods are described elsewhere [5,6,15]. Although the SIAS-M contains both proximal and distal components for motor assessment, only the proximal item were used to facilitate direct comparisons with corresponding items of the NIH-M and CNS-M that have only proximal items. We analyzed the relationships among the corresponding subscales of the three instruments at each examination point Table 1. Comparison of three representative motor assessment instruments for patients with stroke, the NIH-M, the CNS-M and the SIAS-M.
Instrument NIH-M CNS-M
SIAS-M
Severity grading Four grades (0–1–2–3) Four grades (0–1.5–1–2.0) Six grades (0–1–2–3–4–5) (eight grades: Finger function)
Evaluation concept MMT and ROM MMT and ROM
MMT and synergy
Examination items Motor arm Motor leg Arm-Proximal Arm-Distal Leg-Proximal Leg-Distal Knee–Mouth Finger-Function Hip-Flexion Knee-Extension Foot–Pat
with the Spearman’s rank correlation method [16], and compared their responsiveness by plotting the temporal profiles of the scores and by testing the significance of changes over time with the Friedman test [16]. The statistical analyses were performed with a Statview 4Ô software (Abacus Concepts Inc., San Diego, CA) developed for a MacintoshÔ computer (Apple computer, Cupertino, CA).
Results As demonstrated in Table 3, statistically significant correlations were found among the items of the three instruments at each examination point (r ¼ 0.80–0.93, p50.01). The Spearman rank correlation coefficient between SIAS-M and NIH-M was 0.93 for the proximal upper extremity and for the proximal lower extremity, between SIAS-M and CNS-M was 0.92 for the proximal upper extremity, and 0.90 for the proximal lower extremity (p50.01). Thus, high correlations were seen in each case. To analyze the relationships between the SIAS-M and the NIH-M or the CNS-M, we plotted each pair of data on scatter diagrams (Figure 1a for the upper extremity and Figure 1b for the lower extremity scores). As can be seen, the SIAS-M scores were widely dispersed at the same NIH-M or CNS-M scores. For example, a score of 2 with the NIH-M was dispersed in four stages with the SIAS-M, while a score of 3 with the NIH-M was dispersed in three stages with the SIAS-M. Similarly, a score of 1 with the CNS-M proximal lower extremity item was dispersed in four stages with the corresponding SIAS-M item and scores of 0.5 and 1.5 with the CNS-M were each dispersed in three stages with the SIAS-M. The changes in motor recovery from the early stage of onset were compared from the profiles over time of the course of changes for NIH-M, CNS-M and SIAS-M (Figures 2 and 3). When we plotted the temporal profiles of the scores of the three instruments from stroke onset to 12-week post-stroke, the SIASM looked more responsive to changes than the NIH-M or the CNS-M. Compared with the SIAS-M (Figures 2c and 3c), the NIH-M showed less temporal changes in the level of paralysis both for the upper (Figure 2a) and lower extremities (Figure 3a). The assessment with the CNS-M (Figures 2b and 3b) indicated more temporal changes than those with the NIH-M, but more time was needed to detect changes in motor recovery when compared with the SIAS-M (Figures 2c and 3c). We examined the changes over time of each instruments. Friedman test for the upper and lower extremity scoring revealed that the scores changed significantly from onset to the 12th week with the SIAS-M (p50.01) and the CNS-M (p50.05), but not with the NIH-M (p ¼ 0.22).
Discussion For measures to be clinically and scientifically meaningful, they must possess the following properties; reliability, validity, responsiveness and practicality [17]. Although these psychometric properties of the NIHSS, the CNS and the SIAS are well described in the literature as summarized in Table 4 [5,6,8,10,11,13,15,18–30], this is the first study that compared the performance of these standardized instruments with regard to their abilities to assess stroke motor impairment longitudinally for an extended period, i.e. every 2 weeks from the acute phase till the 12th week. Our results demonstrated that the test item scores of the three instruments were highly inter-correlated, and suggested that they could be used interchangeably to describe motor impairment resulting from acute hemispheric stroke. This would facilitate comparisons of stroke studies using one of the three instruments.
Comparison of three instruments: SIAS, NIHSS, CNS
DOI: 10.3109/09638288.2013.854840
1551
Table 2. Grading of the SIAS-M, NIH-M and the CNS-M (abbreviations as in Table 1).
NIH-M
0: no effort against gravity
SIAS-M
CNS-M 0 : absence of motion
0: no voluntary movement 1: minimal voluntary movement
1: can resist gravity
0.5: can’t overcome gravity
Disabil Rehabil Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
2: individual movement is not complete
2: drift
3: carry out the task with severe or moderate clumsiness
3: no drift
4 : mild clumsiness 5: as smoothly as on the unaffected side
1.0: mild weakness 1.5: no detectable gravity
Table 3. Relationships between the SIAS-M item scores and the corresponding NIH-M and CNS-M items scores at each evaluation point from stroke onset to the 12th week (abbreviations as in Table 1). NIH-M SIAS-M
CNS-M
Onset 2 weeks 4 weeks 6 weeks 8 weeks 10 weeks 12 weeks onset 2 weeks 4 weeks 6 weeks 8 weeks 10 weeks 12 weeks
Knee–Mouth 0.80*
0.85*
0.91*
Hip-Flexion
0.87*
0.89*
0.83*
Motor-Arm 0.90* 0.92* Motor-Leg 0.93* 0.90*
0.90*
0.93*
0.82*
0.88*
0.90*
0.92*
0.84*
0.88*
Arm-Proximal 0.92* 0.90* Arm-Distal 0.87* 0.88* 0.90* 0.90*
0.89*
0.89*
0.89*
0.88*
(N ¼ 41, Spearman’s coefficient of rank correlation). *p50.01.
In addition, the more widely dispersed scores of the SIAS-M as compared with the corresponding item scores of either the NIH-M or the CNS-M indicate that the SIAS-M is more responsive to changes of motor impairment than the other two instruments. This would be beneficial in the acute to subacute phase of stroke when a change in motor function is likely to occur. The better responsiveness of the SIAS-M as compared with either the NIH-M or the CNS-M could be explained by the differences in the scoring methods of the three instruments (Table 2). For example, when a patient can move his/her arm but cannot elevate it against gravity, a single score of 0 is given with the NIH-M and 0.5 with the CNS-M, while with the SIAS-M, a score of either 1 or 2 is possible according to the degree of voluntary control. Furthermore, when a patient can elevate his/her arm against gravity but demonstrates weakness, he/she is scored as 2 with the NIH-M and 1.0 with the CNS-M, but either as 3 or 4 with the SIAS-M. This indicates that the SIAS-M is more suited to track changes of motor function longitudinally from stroke onset. This contention is supported by our longitudinal observation that little change was recorded with the NIH-M and longer time was needed to detect changes with the CNS-M, while changes were more readily and clearly discernable with the SIAS-M. The NIH-M and the CNS-M have been extensively used to document neurological impairment (severity of stroke) in acute stroke [10,18–24], but as a longitudinal evaluation tool, the SIASM seemed superior to describe changes over time. Thus in rehabilitation settings and in clinical trials of interventions to
improve motor function, in which accurate documentation of function and its longitudinal change becomes important, the SIAS-M is recommended as an evaluation instrument. Furthermore, the SIAS-M has the advantage that it can evaluate distal upper and lower extremity functions in addition to proximal ones. There are several limitations of the present study. First, the participants were limited to those who could assume sitting position at least with a support, because the SIAS-M was originally developed to assess patients in this position. It is necessary to validate its assessment in lying position to make our findings more generalizable. Second, we only included patients with hemispheric cerebral infarction for this study, and although we believe that our results could be extrapolated to hemorrhagic stroke as well, we need to confirm it. Thirdly, the focus of this paper was on comparisons of the three instruments with regard to their ability to assess motor impairment, but it is important to compare the performance of stroke instruments to assess other aspects of impairments as well, such as sensory, visuospatial and speech impairments. In this sense, it would be necessary to compare the SIAS, developed as a comprehensive assessment instrument of stroke impairment, with the Fugl–Meyer test, which is another widely used comprehensive stroke impairment scale. Despite these limitations, we believe that we successfully clarified the characteristics of commonly used stroke impairment instrument as applied to patients with acute hemispheric stroke.
1552
M. Seki et al.
Disabil Rehabil, 2014; 36(18): 1549–1554
(a)
Disabil Rehabil Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
Figure 1. Relationships between the SIAS-M item scores and the corresponding NIH-M and the CNS-M item scores. (a) Proximal upper extremity, (b) proximal lower extremity). The SIAS-M scores were widely dispersed at the same NIH-M or CNS-M scores. For example, a score of 2 with the NIH-M was dispersed in four stages with the SIASM, while a score of 3 with the NIH-M was dispersed in three stages with the SIAS-M. Similarly, a score of 1 with the CNS-M proximal lower extremity item was dispersed in four stages with the corresponding SIASM item and scores of 0.5 and 1.5 with the CNS-M were each dispersed in three stages with the SIAS-M. Abbreviations: SIAS-M, motor items of the Stroke Impairment Assessment Set; NIH-M, motor items of the National Institute of Stroke Scale; CNS-M, motor items of the Canadian Neurological Scale.
SI A S- M
SI A S- M
score
score
5
5
4
4
3
3
2
2
1
1 0
0 0
1
2
3
0
score
0.5
1
SI A S- M
SI A S- M
sco r e
sco r e
5
5
4
4
3
3
2
2
1
1 0
0 0
1
2
3
0
score
0.5
1
(b)
NIH Stroke Scale （Motor Arm） N＝41
score
(c)
CNS Arm-Proximal N＝41
score
1.5
score
CNS-M
NIH-M
(a)
score
CNS-M
NIH-M (b)
1.5
SIAS Knee-Mouth Test N＝41
score
5 1.5
3
4 2
1.0
1
0.5
3 2 1
0 0
2 4
6
8 10 12
weeks from onset
0 0
0 2 4
6
8 10 12
weeks from onset
0 2 4
6
8 10 12
weeks from onset
Figure 2. Comparison of the temporal profiles of motor recovery of proximal arm function as assessed with the NIH-M (a), the CNS-M (b) and the SIAS-M (c). When the temporal profiles of the scores of the three instruments were plotted from stroke onset to 12-week post-stroke, the SIAS-M looked more responsive to changes than the NIH-M or the CNS-M (abbreviations as in Figure 1). (Broken line; the level is enough for the task to be performed).
Comparison of three instruments: SIAS, NIHSS, CNS
DOI: 10.3109/09638288.2013.854840 NIH Stroke Scale Motor Leg
(a)
(b) N＝41
score
(c)
CNS Leg-Proximal
N＝41
score
1553
SIAS Hip-Flexion Test
N＝41
score
5 1.5
3
4 1.0
2
3 2
00.5 5 1
1 0
0 0
2 4
6
8 10 12
0 2 4
Disabil Rehabil Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
weeks from onset
6
8 10 12
weeks from onset
0 0 2 4
6
8 10 12
weeks from onset
Figure 3. Comparison of the temporal profiles of motor recovery of proximal lower extremity function as assessed with the NIH-M (a), the CNS-M (b) and the SIAS-M (c). When the temporal profiles of the scores of the three instruments were plotted from stroke onset to 12-week post-stroke, the SIAS-M looked more responsive to changes than the NIH-M or the CNS-M (abbreviations as in Figure 1). (Broken line; the level is enough for the task to be performed).
Table 4. Psychometric properties of the Stroke Impairment Assessment Set (SIAS), National Institute of Health Stroke Scale (NIHSS) and Canadian Neurological Scale (CNS). SIAS Scale quality
Unidimensionality with Rasch analysis Six subscales with factor analysis
Reliability Interrater Test–retest Validity Concurrent
Weighted – correlations (1) Between SIAS motor items and the Motricity Index (2) Between SIAS motor items and the Brunnstrom stage (3) Between SIAS lower extremity scores and the Functional Independence Measure (FIMSM) locomotion scores (4) Visuospatial scores and the line bisection and copying task scores (5) Speech scores and the FIMSM communication scores addition of the SIAS as one of the predictors enhanced the predictive power of the discharge FIMSM scores
Predictive
Responsiveness
(1) More responsive to changes than the Motricity Index or the Brunnstrom stage (2) More responsive to changes than the NIHSS or the CNS
Declaration of interest The authors report no conflicts of interest.
NIHSS
CNS
Unidimensionality with Rating scale or Rasch analysis –
Internal consistency with Cronbach’s alpha –
Weighted or intraclass correlation coefficient statistics
Weighted or intraclass correlation coefficient –
(1) Between the NIHSS and the CNS (2) Between the NIHSS and the Middle Cerebral Artery Neurological Scale (MCANS) (3) Between the NIHSS and Guy’s prognostic score (4) Correlating the NIHSS with computed tomography measurement of infarction and clinical outcome at 3 months
(1) Between the CNS and the standard neurological examination (2) Between the CNS and the NIHSS (3) Between the CNS and the MCANS (4) Between the CNS and Guy’s prognostic score
(1) Utility of the NIHSS as a predictor of hospital disposition (2) The NIHSS provides a better prediction of 3-month outcome than the CNS or the MCANS
(1) Retrospective scoring of initial stroke severity based on the CNS is valid in hospital discharge summaries (2) The total initial CNS score to be significant predictor of death, morbidity and recovery of the Katz activities of daily living index
Neurologic recovery in the first hours assessed by the NIHSS
References 1. Liu M, Chino N, Takahashi H. Current status of rehabilitation, especially in patients with stroke. Scand J Rehabil Med 2000;32: 148–58.
Disabil Rehabil Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
1554
M. Seki et al.
2. Brunnstrom S. Movement therapy in hemiplegia. New York: Harper & Row; 1970. 3. Demeurisse G, Demol O, Robaye E. Motor evaluation in vascular hemiplegia. Eur Neurol 1980;19:382–9. 4. Fugl-Meyer AR, Jaasko L, Leyman I, et al. The post-stroke hemiplegic patient. A method for evaluation of physical performance. Scand J Rehabil Med 1975;7:13–31. 5. Goldstein LB, Bertels C, Davis JN. Interrater reliability of the NIH stroke scale. Arch Neurol 1989;46:660–2. 6. Cote R, Battista RN, Wolfson C, et al. The Canadian neurological scale: validation and reliability assessment. Neurology 1989;39: 638–43. 7. Lindenstrom E, Boysen G, Christiansen LW, et al. Reliability of Scandinavian neurological stroke scale. Cerebrovasc Dis 1991;1: 103–7. 8. Chino N, Sonoda S, Domen K, et al. Stroke impairment assessment set (SIAS) – a new evaluation instrument for stroke patients. Jpn J Rehabil Med 1994;31:119–25. 9. Malouin F, Pichard L, Bonneau C, et al. Evaluating motor recovery after stroke: comparison of the Fugl–Meyer Assessment and the Motor Assessment Scale. Arch Phys Med Rehabil 1994;75:1206–12. 10. Muir KW, Grosset DG, Lees KR. Interconversion of stroke scales: implications for therapeutic trials. Stroke 1994;25:1366–70. 11. Muir KW, Weir CJ, Murray GD, et al. Comparison of neurological scales and scoring systems for acute stroke. Stroke 1996;27: 1817–20. 12. Broeks J, Lankhorst G, Rumping K, Prevo A. The long-term outcome of arm function after stroke: results of a follow-up study. Disabil Rehabil 1999;21:357–64. 13. Bushnell CD, Johnston DC, Goldstein LB. Retrospective assessment of initial stroke severity: comparison of the NIH Stroke Scale and the Canadian Neurological Scale. Stroke 2001;32:656–60. 14. Nilanont Y, Komoltri C, Saposnik G, et al. The Canadian Neurological Scale and the NIHSS: development and validation of a simple conversion model. Cerebrovasc Dis 2010;30:120–6. 15. Liu M, Chino N, Tsuji T, et al. Psychometric properties of the Stroke Impairment Assessment Set (SIAS). Neurorehabil Neural Repair 2002;16:339–51. 16. Armitage P, Berry G. Statistical methods in medical research. 3rd ed. Oxford: Blackwell Scientific Publications; 1994. 17. Johnston MV, Keith RA. Measurement standards for medical rehabilitation and clinical applications. Phys Med Clin North Am 1993;4:417–49.
Disabil Rehabil, 2014; 36(18): 1549–1554
18. Brott T, Adams HP, Olinger CP, et al. Measurements of acute cerebral infarction: a clinical examination scale. Stroke 1989;39: 638–43. 19. Cote R, Hachinski V, Scurvell BI, et al. The Canadian neurological scale: a preliminary study in acute stroke. Stroke 1986;17:731–7. 20. Kasner SE, Chalela JA, Luciano JM, et al. Reliability and validity of estimating the NIH Stroke Scale score from medical records. Stroke 1999;30:1534–7. 21. Goldstein LB, Chilukuri V. Retrospective assessment of initial stroke severity with the Canadian Neurological Scale. Stroke 1997; 28:1181–4. 22. Biller J, Love BB, Marsh EE, et al. Spontaneous improvement after acute ischemic. A pilot study. Stroke 1990;21:1008–12. 23. Schlegel D, Kolb SJ, Luciano JM, et al. Utility of the NIH Stroke Scale as a predictor of hospital disposition. Stroke 2003;34: 134–7. 24. Heinemann AW, Harvey RL, McGuire JR, et al. Measurement properties of the NIH Stroke Scale during acute rehabilitation. Stroke 1997;28:1174–80. 25. Chino N, Sonoda S, Domen K, et al. Stroke impairment assessment set (SIAS). In: Chino N, Melvin JL, eds. Functional evaluation of stroke patients. Tokyo, Japan: Springer-Verlag; 1996:19–31. 26. Domen K, Sonoda S, Chino N, et al. Evaluation of motor function in stroke patients using the Stroke Impairment Assessment Set (SIAS). In: Chino N, Melvin JL, eds. Functional evaluation of stroke patients. Tokyo, Japan: Springer-Verlag; 1996:33–44. 27. Tsuji T, Liu M, Sonoda S, et al. The stroke impairment assessment set: its internal consistency and predictive validity. Arch Phys Med Rehabil 2000;81:863–8. 28. Liu M, Domen K, Chino N. Comorbidity measures stroke outcome research: a preliminary study. Arch Phys Med Rehabil 1997;78: 166–72. 29. Sonoda S, Chino N, Domen K, Saitoh E. Changes in impairment and disability from the 3rd to the 6th month after stroke and its relationship evaluated by an artificial neural network. Am J Phys Med Rehabil 1997;76:395–400. 30. Sonoda S, Saitoh E, Domen K, Chino N. Prognostication of stroke patients using the Stroke Impairment Assessment Set and the Functional Independence Measure. In: Chino N, Melvin JL, eds. Functional evaluation of stroke patients. Tokyo, Japan: SpringerVerlag; 1996:103–14.
ASSESSMENT PROCEDURES
Disabil Rehabil Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
Comparison of three instruments to assess changes of motor impairment in acute hemispheric stroke: the Stroke Impairment Assessment Set (SIAS), the National Institute of Health Stroke Scale (NIHSS) and the Canadian Neurological Scale (CNS) Masaru Seki1, Kimitaka Hase2, Hidetoshi Takahashi3, and Meigen Liu4 1
Department of Liberal Arts and Human Development, Kanagawa University of Human Services, Kanagawa, Japan, 2Department of Rehabilitation Medicine, Kansai Medical University, Osaka, Japan, 3Department of Rehabilitation Medicine, International Medical Center, Saitama Medical University, Saitama, Japan, and 4Department of Rehabilitation Medicine, Keio University School of Medicine, Tokyo, Japan Abstract
Keywords
Purpose: The purpose was to compare the performance of three representative instruments to measure changes of motor recovery with acute hemispheric stroke. Method: In 41 consecutive patients with acute hemispheric infarction, motor recovery was assessed within 3 days of onset and then every 2 weeks until the 12th week with the Stroke Impairment Assessment Set (SIAS), the National Institutes of Health stroke scale (NIHSS) and the Canadian Neurological Scale (CNS). We analyzed the relationships among the corresponding subscales of the three instruments with the Spearman’s rank correlation method, and compared their responsiveness by plotting the temporal profiles of scores of each instrument and by testing the significance of changes over time with the Friedman test. Results: High correlations were observed among the three instruments at each examination point. However, the SIAS scores were more widely dispersed at the same NIHSS and CNS scores. Friedman test revealed that the scores changed significantly during the observation period with the SIAS and the CNS but not with the NIHSS. The changes were detected later with the CNS than with the SIAS. Conclusions: Although the assessment results with the three instruments were highly inter-correlated, the SIAS performed better with respect to responsiveness to changes.
Hemiplegia, measurement, motor recovery, stroke History Received 18 March 2013 Revised 29 September 2013 Accepted 9 October 2013 Published online 18 November 2013
ä Implication for Rehabilitation
When providing rehabilitation services to patients with stroke, it is important to document objectively the level of their impairment from the acute stage and during the entire course of rehabilitation with standardized instrument, and to predict their functional outcomes as early and accurately as possible. This study therefore compared three representative instruments with established psychometric properties, the SIAS, the NIHSS and the CNS, with respect to their ability to document motor impairment and their responsiveness to recovery in patients with acute hemispheric stroke.
Introduction Stroke poses us important health and economic problems in many countries [1], and rehabilitation plays an important role in its management. When providing rehabilitation services to patients with stroke, it is important to document objectively the level of their impairment from the acute stage and during the entire course of rehabilitation with standardized instrument, and to predict their functional outcomes as early and as
Address for correspondence: Masaru Seki, MD, PhD, Department of Liberal Arts and Human Development, Kanagawa University of Human Services, 1-10-1 Heiseicho, Yokosuka, Kanagawa 238-8522, Japan. Tel: +81 46 828 2772. Fax: +81 46 828 2773. E-mail: [email protected]
accurately as possible. This is necessary to improve the quality of stroke rehabilitation programs while making the best use of the limited resources. For this purpose, a number of instruments have been proposed with varying degrees of descriptions of their psychometric properties, such as the Brunnstrom test [2] and the Motricity index for motor assessment [3], the Fugl– Meyer test [4], the National Institutes of Health Stroke Scale (NIHSS) [5], the Canadian Neurological Scale (CNS) [6], the Scandinavian Stroke Scale [7] and the Stroke Impairment Assessment Set (SIAS) [8] for comprehensive assessment of stroke impairment. However, it is not clear which instrument we are advised to use in daily clinical practice to evaluate patients with stroke. To answer this question, it is imperative to compare the
1550
M. Seki et al.
Disabil Rehabil, 2014; 36(18): 1549–1554
performance of each instrument to describe the degree of impairment and track changes over time. Although several studies are available that compared currently used stroke scales crosssectionally [9–14], no study has yet been reported that compares them longitudinally from the acute stage of onset in detail. The purpose of the present study was therefore to compare three representative instruments with established psychometric properties, the SIAS, the NIHSS and the CNS, with respect to their ability to document motor impairment and their responsiveness to recovery in patients with acute hemispheric stroke.
Disabil Rehabil Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
Methods The participants were 41 consecutive patients with CT- or MRIconfirmed cerebral hemispheric infarction (22 males and 19 females, 21 with right-sided and 20 with left-sided brain lesion, mean age: 62.7 11.1 years), who were admitted to one of the two acute care general hospitals, Odawara City Hospital (n ¼ 31) or Shizuoka City Hospital (n ¼ 10). To be included in the study, patients had to: (1) have measurable hemiparesis, (2) be hospitalized within 24 h of onset of neurological symptoms, (3) have no preexisting stroke deficits, (4) be able to understand the test tasks related to motor paralysis assessment and (5) be able to assume the sitting position at least with a support. Patients were excluded if they had: (1) hemorrhagic stroke to make them homogeneous with respect to stroke type, (2) disturbance of consciousness, (3) severe communication deficits, (4) motor apraxia, (5) unstable medical condition(s) and/or (6) other coexisting disabling condition(s) (e.g. arthritis, amputations). All the patients underwent a standard acute rehabilitation program together with acute medical care. Before enrollment, the purpose and procedures of the study were fully explained, and informed consent was obtained from the patients and their family members. This study followed principles in the Declaration of Helsinki. Within 3 days from stroke onset and every 2 weeks thereafter until the 12th week, the same board-certified physiatrist experienced in acute stroke rehabilitation assessed motor impairment of all patients with the motor items of the SIAS (SIAS-M), the NIHSS (NIH-M) and the CNS (CNS-M). Tables 1 and 2 illustrate the motor items and scaling of each instrument. The details of the assessment methods are described elsewhere [5,6,15]. Although the SIAS-M contains both proximal and distal components for motor assessment, only the proximal item were used to facilitate direct comparisons with corresponding items of the NIH-M and CNS-M that have only proximal items. We analyzed the relationships among the corresponding subscales of the three instruments at each examination point Table 1. Comparison of three representative motor assessment instruments for patients with stroke, the NIH-M, the CNS-M and the SIAS-M.
Instrument NIH-M CNS-M
SIAS-M
Severity grading Four grades (0–1–2–3) Four grades (0–1.5–1–2.0) Six grades (0–1–2–3–4–5) (eight grades: Finger function)
Evaluation concept MMT and ROM MMT and ROM
MMT and synergy
Examination items Motor arm Motor leg Arm-Proximal Arm-Distal Leg-Proximal Leg-Distal Knee–Mouth Finger-Function Hip-Flexion Knee-Extension Foot–Pat
with the Spearman’s rank correlation method [16], and compared their responsiveness by plotting the temporal profiles of the scores and by testing the significance of changes over time with the Friedman test [16]. The statistical analyses were performed with a Statview 4Ô software (Abacus Concepts Inc., San Diego, CA) developed for a MacintoshÔ computer (Apple computer, Cupertino, CA).
Results As demonstrated in Table 3, statistically significant correlations were found among the items of the three instruments at each examination point (r ¼ 0.80–0.93, p50.01). The Spearman rank correlation coefficient between SIAS-M and NIH-M was 0.93 for the proximal upper extremity and for the proximal lower extremity, between SIAS-M and CNS-M was 0.92 for the proximal upper extremity, and 0.90 for the proximal lower extremity (p50.01). Thus, high correlations were seen in each case. To analyze the relationships between the SIAS-M and the NIH-M or the CNS-M, we plotted each pair of data on scatter diagrams (Figure 1a for the upper extremity and Figure 1b for the lower extremity scores). As can be seen, the SIAS-M scores were widely dispersed at the same NIH-M or CNS-M scores. For example, a score of 2 with the NIH-M was dispersed in four stages with the SIAS-M, while a score of 3 with the NIH-M was dispersed in three stages with the SIAS-M. Similarly, a score of 1 with the CNS-M proximal lower extremity item was dispersed in four stages with the corresponding SIAS-M item and scores of 0.5 and 1.5 with the CNS-M were each dispersed in three stages with the SIAS-M. The changes in motor recovery from the early stage of onset were compared from the profiles over time of the course of changes for NIH-M, CNS-M and SIAS-M (Figures 2 and 3). When we plotted the temporal profiles of the scores of the three instruments from stroke onset to 12-week post-stroke, the SIASM looked more responsive to changes than the NIH-M or the CNS-M. Compared with the SIAS-M (Figures 2c and 3c), the NIH-M showed less temporal changes in the level of paralysis both for the upper (Figure 2a) and lower extremities (Figure 3a). The assessment with the CNS-M (Figures 2b and 3b) indicated more temporal changes than those with the NIH-M, but more time was needed to detect changes in motor recovery when compared with the SIAS-M (Figures 2c and 3c). We examined the changes over time of each instruments. Friedman test for the upper and lower extremity scoring revealed that the scores changed significantly from onset to the 12th week with the SIAS-M (p50.01) and the CNS-M (p50.05), but not with the NIH-M (p ¼ 0.22).
Discussion For measures to be clinically and scientifically meaningful, they must possess the following properties; reliability, validity, responsiveness and practicality [17]. Although these psychometric properties of the NIHSS, the CNS and the SIAS are well described in the literature as summarized in Table 4 [5,6,8,10,11,13,15,18–30], this is the first study that compared the performance of these standardized instruments with regard to their abilities to assess stroke motor impairment longitudinally for an extended period, i.e. every 2 weeks from the acute phase till the 12th week. Our results demonstrated that the test item scores of the three instruments were highly inter-correlated, and suggested that they could be used interchangeably to describe motor impairment resulting from acute hemispheric stroke. This would facilitate comparisons of stroke studies using one of the three instruments.
Comparison of three instruments: SIAS, NIHSS, CNS
DOI: 10.3109/09638288.2013.854840
1551
Table 2. Grading of the SIAS-M, NIH-M and the CNS-M (abbreviations as in Table 1).
NIH-M
0: no effort against gravity
SIAS-M
CNS-M 0 : absence of motion
0: no voluntary movement 1: minimal voluntary movement
1: can resist gravity
0.5: can’t overcome gravity
Disabil Rehabil Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
2: individual movement is not complete
2: drift
3: carry out the task with severe or moderate clumsiness
3: no drift
4 : mild clumsiness 5: as smoothly as on the unaffected side
1.0: mild weakness 1.5: no detectable gravity
Table 3. Relationships between the SIAS-M item scores and the corresponding NIH-M and CNS-M items scores at each evaluation point from stroke onset to the 12th week (abbreviations as in Table 1). NIH-M SIAS-M
CNS-M
Onset 2 weeks 4 weeks 6 weeks 8 weeks 10 weeks 12 weeks onset 2 weeks 4 weeks 6 weeks 8 weeks 10 weeks 12 weeks
Knee–Mouth 0.80*
0.85*
0.91*
Hip-Flexion
0.87*
0.89*
0.83*
Motor-Arm 0.90* 0.92* Motor-Leg 0.93* 0.90*
0.90*
0.93*
0.82*
0.88*
0.90*
0.92*
0.84*
0.88*
Arm-Proximal 0.92* 0.90* Arm-Distal 0.87* 0.88* 0.90* 0.90*
0.89*
0.89*
0.89*
0.88*
(N ¼ 41, Spearman’s coefficient of rank correlation). *p50.01.
In addition, the more widely dispersed scores of the SIAS-M as compared with the corresponding item scores of either the NIH-M or the CNS-M indicate that the SIAS-M is more responsive to changes of motor impairment than the other two instruments. This would be beneficial in the acute to subacute phase of stroke when a change in motor function is likely to occur. The better responsiveness of the SIAS-M as compared with either the NIH-M or the CNS-M could be explained by the differences in the scoring methods of the three instruments (Table 2). For example, when a patient can move his/her arm but cannot elevate it against gravity, a single score of 0 is given with the NIH-M and 0.5 with the CNS-M, while with the SIAS-M, a score of either 1 or 2 is possible according to the degree of voluntary control. Furthermore, when a patient can elevate his/her arm against gravity but demonstrates weakness, he/she is scored as 2 with the NIH-M and 1.0 with the CNS-M, but either as 3 or 4 with the SIAS-M. This indicates that the SIAS-M is more suited to track changes of motor function longitudinally from stroke onset. This contention is supported by our longitudinal observation that little change was recorded with the NIH-M and longer time was needed to detect changes with the CNS-M, while changes were more readily and clearly discernable with the SIAS-M. The NIH-M and the CNS-M have been extensively used to document neurological impairment (severity of stroke) in acute stroke [10,18–24], but as a longitudinal evaluation tool, the SIASM seemed superior to describe changes over time. Thus in rehabilitation settings and in clinical trials of interventions to
improve motor function, in which accurate documentation of function and its longitudinal change becomes important, the SIAS-M is recommended as an evaluation instrument. Furthermore, the SIAS-M has the advantage that it can evaluate distal upper and lower extremity functions in addition to proximal ones. There are several limitations of the present study. First, the participants were limited to those who could assume sitting position at least with a support, because the SIAS-M was originally developed to assess patients in this position. It is necessary to validate its assessment in lying position to make our findings more generalizable. Second, we only included patients with hemispheric cerebral infarction for this study, and although we believe that our results could be extrapolated to hemorrhagic stroke as well, we need to confirm it. Thirdly, the focus of this paper was on comparisons of the three instruments with regard to their ability to assess motor impairment, but it is important to compare the performance of stroke instruments to assess other aspects of impairments as well, such as sensory, visuospatial and speech impairments. In this sense, it would be necessary to compare the SIAS, developed as a comprehensive assessment instrument of stroke impairment, with the Fugl–Meyer test, which is another widely used comprehensive stroke impairment scale. Despite these limitations, we believe that we successfully clarified the characteristics of commonly used stroke impairment instrument as applied to patients with acute hemispheric stroke.
1552
M. Seki et al.
Disabil Rehabil, 2014; 36(18): 1549–1554
(a)
Disabil Rehabil Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
Figure 1. Relationships between the SIAS-M item scores and the corresponding NIH-M and the CNS-M item scores. (a) Proximal upper extremity, (b) proximal lower extremity). The SIAS-M scores were widely dispersed at the same NIH-M or CNS-M scores. For example, a score of 2 with the NIH-M was dispersed in four stages with the SIASM, while a score of 3 with the NIH-M was dispersed in three stages with the SIAS-M. Similarly, a score of 1 with the CNS-M proximal lower extremity item was dispersed in four stages with the corresponding SIASM item and scores of 0.5 and 1.5 with the CNS-M were each dispersed in three stages with the SIAS-M. Abbreviations: SIAS-M, motor items of the Stroke Impairment Assessment Set; NIH-M, motor items of the National Institute of Stroke Scale; CNS-M, motor items of the Canadian Neurological Scale.
SI A S- M
SI A S- M
score
score
5
5
4
4
3
3
2
2
1
1 0
0 0
1
2
3
0
score
0.5
1
SI A S- M
SI A S- M
sco r e
sco r e
5
5
4
4
3
3
2
2
1
1 0
0 0
1
2
3
0
score
0.5
1
(b)
NIH Stroke Scale （Motor Arm） N＝41
score
(c)
CNS Arm-Proximal N＝41
score
1.5
score
CNS-M
NIH-M
(a)
score
CNS-M
NIH-M (b)
1.5
SIAS Knee-Mouth Test N＝41
score
5 1.5
3
4 2
1.0
1
0.5
3 2 1
0 0
2 4
6
8 10 12
weeks from onset
0 0
0 2 4
6
8 10 12
weeks from onset
0 2 4
6
8 10 12
weeks from onset
Figure 2. Comparison of the temporal profiles of motor recovery of proximal arm function as assessed with the NIH-M (a), the CNS-M (b) and the SIAS-M (c). When the temporal profiles of the scores of the three instruments were plotted from stroke onset to 12-week post-stroke, the SIAS-M looked more responsive to changes than the NIH-M or the CNS-M (abbreviations as in Figure 1). (Broken line; the level is enough for the task to be performed).
Comparison of three instruments: SIAS, NIHSS, CNS
DOI: 10.3109/09638288.2013.854840 NIH Stroke Scale Motor Leg
(a)
(b) N＝41
score
(c)
CNS Leg-Proximal
N＝41
score
1553
SIAS Hip-Flexion Test
N＝41
score
5 1.5
3
4 1.0
2
3 2
00.5 5 1
1 0
0 0
2 4
6
8 10 12
0 2 4
Disabil Rehabil Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
weeks from onset
6
8 10 12
weeks from onset
0 0 2 4
6
8 10 12
weeks from onset
Figure 3. Comparison of the temporal profiles of motor recovery of proximal lower extremity function as assessed with the NIH-M (a), the CNS-M (b) and the SIAS-M (c). When the temporal profiles of the scores of the three instruments were plotted from stroke onset to 12-week post-stroke, the SIAS-M looked more responsive to changes than the NIH-M or the CNS-M (abbreviations as in Figure 1). (Broken line; the level is enough for the task to be performed).
Table 4. Psychometric properties of the Stroke Impairment Assessment Set (SIAS), National Institute of Health Stroke Scale (NIHSS) and Canadian Neurological Scale (CNS). SIAS Scale quality
Unidimensionality with Rasch analysis Six subscales with factor analysis
Reliability Interrater Test–retest Validity Concurrent
Weighted – correlations (1) Between SIAS motor items and the Motricity Index (2) Between SIAS motor items and the Brunnstrom stage (3) Between SIAS lower extremity scores and the Functional Independence Measure (FIMSM) locomotion scores (4) Visuospatial scores and the line bisection and copying task scores (5) Speech scores and the FIMSM communication scores addition of the SIAS as one of the predictors enhanced the predictive power of the discharge FIMSM scores
Predictive
Responsiveness
(1) More responsive to changes than the Motricity Index or the Brunnstrom stage (2) More responsive to changes than the NIHSS or the CNS
Declaration of interest The authors report no conflicts of interest.
NIHSS
CNS
Unidimensionality with Rating scale or Rasch analysis –
Internal consistency with Cronbach’s alpha –
Weighted or intraclass correlation coefficient statistics
Weighted or intraclass correlation coefficient –
(1) Between the NIHSS and the CNS (2) Between the NIHSS and the Middle Cerebral Artery Neurological Scale (MCANS) (3) Between the NIHSS and Guy’s prognostic score (4) Correlating the NIHSS with computed tomography measurement of infarction and clinical outcome at 3 months
(1) Between the CNS and the standard neurological examination (2) Between the CNS and the NIHSS (3) Between the CNS and the MCANS (4) Between the CNS and Guy’s prognostic score
(1) Utility of the NIHSS as a predictor of hospital disposition (2) The NIHSS provides a better prediction of 3-month outcome than the CNS or the MCANS
(1) Retrospective scoring of initial stroke severity based on the CNS is valid in hospital discharge summaries (2) The total initial CNS score to be significant predictor of death, morbidity and recovery of the Katz activities of daily living index
Neurologic recovery in the first hours assessed by the NIHSS
References 1. Liu M, Chino N, Takahashi H. Current status of rehabilitation, especially in patients with stroke. Scand J Rehabil Med 2000;32: 148–58.
Disabil Rehabil Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
1554
M. Seki et al.
2. Brunnstrom S. Movement therapy in hemiplegia. New York: Harper & Row; 1970. 3. Demeurisse G, Demol O, Robaye E. Motor evaluation in vascular hemiplegia. Eur Neurol 1980;19:382–9. 4. Fugl-Meyer AR, Jaasko L, Leyman I, et al. The post-stroke hemiplegic patient. A method for evaluation of physical performance. Scand J Rehabil Med 1975;7:13–31. 5. Goldstein LB, Bertels C, Davis JN. Interrater reliability of the NIH stroke scale. Arch Neurol 1989;46:660–2. 6. Cote R, Battista RN, Wolfson C, et al. The Canadian neurological scale: validation and reliability assessment. Neurology 1989;39: 638–43. 7. Lindenstrom E, Boysen G, Christiansen LW, et al. Reliability of Scandinavian neurological stroke scale. Cerebrovasc Dis 1991;1: 103–7. 8. Chino N, Sonoda S, Domen K, et al. Stroke impairment assessment set (SIAS) – a new evaluation instrument for stroke patients. Jpn J Rehabil Med 1994;31:119–25. 9. Malouin F, Pichard L, Bonneau C, et al. Evaluating motor recovery after stroke: comparison of the Fugl–Meyer Assessment and the Motor Assessment Scale. Arch Phys Med Rehabil 1994;75:1206–12. 10. Muir KW, Grosset DG, Lees KR. Interconversion of stroke scales: implications for therapeutic trials. Stroke 1994;25:1366–70. 11. Muir KW, Weir CJ, Murray GD, et al. Comparison of neurological scales and scoring systems for acute stroke. Stroke 1996;27: 1817–20. 12. Broeks J, Lankhorst G, Rumping K, Prevo A. The long-term outcome of arm function after stroke: results of a follow-up study. Disabil Rehabil 1999;21:357–64. 13. Bushnell CD, Johnston DC, Goldstein LB. Retrospective assessment of initial stroke severity: comparison of the NIH Stroke Scale and the Canadian Neurological Scale. Stroke 2001;32:656–60. 14. Nilanont Y, Komoltri C, Saposnik G, et al. The Canadian Neurological Scale and the NIHSS: development and validation of a simple conversion model. Cerebrovasc Dis 2010;30:120–6. 15. Liu M, Chino N, Tsuji T, et al. Psychometric properties of the Stroke Impairment Assessment Set (SIAS). Neurorehabil Neural Repair 2002;16:339–51. 16. Armitage P, Berry G. Statistical methods in medical research. 3rd ed. Oxford: Blackwell Scientific Publications; 1994. 17. Johnston MV, Keith RA. Measurement standards for medical rehabilitation and clinical applications. Phys Med Clin North Am 1993;4:417–49.
Disabil Rehabil, 2014; 36(18): 1549–1554
18. Brott T, Adams HP, Olinger CP, et al. Measurements of acute cerebral infarction: a clinical examination scale. Stroke 1989;39: 638–43. 19. Cote R, Hachinski V, Scurvell BI, et al. The Canadian neurological scale: a preliminary study in acute stroke. Stroke 1986;17:731–7. 20. Kasner SE, Chalela JA, Luciano JM, et al. Reliability and validity of estimating the NIH Stroke Scale score from medical records. Stroke 1999;30:1534–7. 21. Goldstein LB, Chilukuri V. Retrospective assessment of initial stroke severity with the Canadian Neurological Scale. Stroke 1997; 28:1181–4. 22. Biller J, Love BB, Marsh EE, et al. Spontaneous improvement after acute ischemic. A pilot study. Stroke 1990;21:1008–12. 23. Schlegel D, Kolb SJ, Luciano JM, et al. Utility of the NIH Stroke Scale as a predictor of hospital disposition. Stroke 2003;34: 134–7. 24. Heinemann AW, Harvey RL, McGuire JR, et al. Measurement properties of the NIH Stroke Scale during acute rehabilitation. Stroke 1997;28:1174–80. 25. Chino N, Sonoda S, Domen K, et al. Stroke impairment assessment set (SIAS). In: Chino N, Melvin JL, eds. Functional evaluation of stroke patients. Tokyo, Japan: Springer-Verlag; 1996:19–31. 26. Domen K, Sonoda S, Chino N, et al. Evaluation of motor function in stroke patients using the Stroke Impairment Assessment Set (SIAS). In: Chino N, Melvin JL, eds. Functional evaluation of stroke patients. Tokyo, Japan: Springer-Verlag; 1996:33–44. 27. Tsuji T, Liu M, Sonoda S, et al. The stroke impairment assessment set: its internal consistency and predictive validity. Arch Phys Med Rehabil 2000;81:863–8. 28. Liu M, Domen K, Chino N. Comorbidity measures stroke outcome research: a preliminary study. Arch Phys Med Rehabil 1997;78: 166–72. 29. Sonoda S, Chino N, Domen K, Saitoh E. Changes in impairment and disability from the 3rd to the 6th month after stroke and its relationship evaluated by an artificial neural network. Am J Phys Med Rehabil 1997;76:395–400. 30. Sonoda S, Saitoh E, Domen K, Chino N. Prognostication of stroke patients using the Stroke Impairment Assessment Set and the Functional Independence Measure. In: Chino N, Melvin JL, eds. Functional evaluation of stroke patients. Tokyo, Japan: SpringerVerlag; 1996:103–14.
