The International Journal of Neuroscience

Transcription

1 The Underlying Factor Structure of National Institutes of Health Stroke Scale: An Exploratory Factor Analysis Journal: The International Journal of Neuroscience Manuscript ID: GNES-0-0 Manuscript Type: Original Article Date Submitted by the Author: -Aug-0 Complete List of Authors: Zandieh, Ali; Tehran University of Medical Sciences, Iranian Center of Neurological Research Zeinali Kohoki, Zahra; Tehran University of Medical Sciences, Department of Neurology, Imam Khomeini hospital complex Sadeghian, Homa; Tehran University of Medical Sciences, Iranian Center of Neurological Research Pourashraf, Maryam; Tehran University of Medical Sciences, Iranian Center of Neurological Research Parviz, Sara; Tehran University of Medical Sciences, Iranian Center of Neurological Research Ghaffarpour, Majid; Tehran University of Medical Sciences, Iranian Center of Neurological Research; Tehran University of Medical Sciences, Department of Neurology, Imam Khomeini hospital complex Ghabaee, Mojdeh; Tehran University of Medical Sciences, Iranian Center of Neurological Research; Tehran University of Medical Sciences, Department of Neurology, Imam Khomeini hospital complex Keywords: National institutes of health stroke scale, ischemic stroke, factor analysis

2 Page of The International Journal of Neuroscience

3 Page of The Underlying Factor Structure of National Institutes of Health Stroke Scale: An Exploratory Factor Analysis Ali Zandieh, M.D., M.P.H., Zahra Zeinali Kohoki, M.D., Homa Sadeghian, M.D., Maryam Pourashraf, M.D., Sara Parviz, M.D., Majid Ghaffarpour, M.D.,,and Mojdeh Ghabaee, M.D., * Iranian Center of Neurological Research, Tehran University of Medical Sciences, Tehran, Iran. Department of Neurology, Imam Khomeini hospital complex, Tehran University of Medical Sciences, Tehran, Iran. Running title: Underlying structure of NIHSS Keywords: National institutes of health stroke scale, ischemic stroke, factor analysis addresses: Ali_zandieh@yahoo.com Drzzeynalik@yahoo.com Homa.sadeghian@gmail.com Pourashraf_m@yahoo.com Sarapzu@yahoo.com Ghaffarpourm@tums.ac.ir Ghabaeem@tums.ac.ir * Correspondence to: Mojdeh Ghabaee, M.D., Professor of Neurology Department of Neurology, Iranian Center of Neurological Research, Tehran University of Medical Sciences, Tehran, Iran. P.O.Box:, Tehran, Iran Fax: +()

4 Page of The International Journal of Neuroscience Tel: +()- ghabaeem@tums.ac.ir, mojdeh.ghabaee@gmail.com The authors declare no conflicts of interests Word count: abstract:, main manuscript: Number of Tables:, Figures: 0, References:

5 Page of Abstract The underlying structure of national institutes of health stroke scale (NIHSS) as the most widely used scale in clinical trials has been the focus of little attention. The aim of the current study was to elucidate the clustering pattern of NIHSS items in ischemic stroke patients. A series of consecutive patients with first ever ischemic strokes admitted to a university affiliated hospital were enrolled. NIHSS score was estimated on admission and correlation coefficients between its items were calculated. Further, exploratory factor analysis was used to study the clustering pattern of NIHSS items. Extinction neglect, visual field and facial palsy were weakly associated with other NIHSS items. Factor analysis led to a four factor structure. Factors and were determined by left brain function as items of right arm and leg motor, language and dysarthria loaded on both of them. By contrast factor reflected right brain involvement. Since visual field and ataxia loaded on factor, this factor was primarily associated with posterior strokes. Our study shows that the four factor structure model is plausible for NIHSS. Further, for the first time a single distinct factor is identified for posterior strokes.

6 Page of The International Journal of Neuroscience Introduction Stroke is the leading cause of adult long-term disability and the second most common cause of death in Eastren-Mediterranean countries [, ]. Considering epidemiological transitions in populations and subsequent further rising in burden of stroke [], it is of principal importance to develop a valid and reliable stroke outcome scale. In this respect, one of the most widely used scales in clinical trials is national institutes of health stroke scale (NIHSS) [-]. It is comprised of scores related to consciousness, cognition, inattention, speech, sensory and motor impairments, visual field abnormalities and ataxia []. NIHSS is simple, easy to administer and can be assessed by diverse groups of health care professionals [, ]. Further, review of previous studies shows that NIHSS has high ability in predicting stroke outcomes at different times following onset of symptoms [-0]. In spite of aforementioned positive points, NIHSS has also some problems. For instance, it is argued that some items of this scale because of poor reliability indices are redundant [, ]. In this respect, Tirschwell et al. have shown that five selected items of NIHSS scale reflect much of the predictive performance of the original scale []. Studying interrelationships of NIHSS items in terms of a small set of underlying factors would be helpful for diagnosis of internal structure of this scale. Factor analysis is potentially a way of approaching this body of research. In this statistical method, each extracted factor refers to a pathophysiologic mechanism such as left or right cerebral function. In this study, we evaluated the clustering pattern of NIHSS items, using factor analysis. Methods A series patients with stroke who consecutively admitted within hours after onset of symptoms to our main University affiliated hospital-based stroke center from February 00 to

7 Page of March 00 were enrolled in the current study. Those subjects whose symptoms were not consistent with the findings in computed tomography (CT) or magnetic resonance imaging (MRI) were not included. After excluding individuals with prior stroke events (n=) and those with hemorrhagic stroke (n=), analyses were conducted on the remaining subjects. All patients were followed up for at least days Data regarding the demographic and clinical characteristics of all subjects were obtained. Neurological examinations, -lead electrocardiography and echocardiography were performed on all participants. National institutes of health stroke scale (NIHSS) was assessed by a neurologist to estimate stroke severity on admission. The NIHSS was categorized as <, - and >. Analyses of data were conducted by SPSS software (version.0; SPSS Inc., Chicago, USA). Kendall s tau correlation coefficients were estimated for the associations between NIHSS items. Using principal component analysis, exploratory factor analysis was performed to identify the clustering pattern of NIHSS items. This statistical method reduces the number of original variables into fewer latent factors, in accordance with the linear correlation that exist among the variables. Factors with an eigenvalue (the amount of variance attributable to the factor) of greater than one were retained. Further, quartimax rotation method was conducted to facilitate the interpretation. This method helps to identify the variables that correlate with each extracted factor by maximizing factor loadings to one or minimizing it to zero. The factor loading of a variable on a factor is similar to the correlation coefficient between that variable and the factor score. In accordance with several previous reports, Factor loadings of 0. were considered meaningful for interpretation; because this ensures that the variable shares at least % of the variance with the corresponding factor [, ]. P <0.0 was considered statistically significant.

8 Page of The International Journal of Neuroscience Results Table summarizes the principal characteristics of the study population. The mean age ± standard deviation (SD) of subjects was. ±.0 years. The cumulative rate of mortality was equal to.% at -day follow-up period. Among NIHSS items, motor impairments (.%-.% for different limbs) and facial palsy (.%) were the most frequent symptoms observed in stroke patients. On the contrary, only.%,.% and.% of subjects suffered from impairments attributed to the items of extinction neglect, limb ataxia and gaze, respectively. Correlation coefficients of NIHSS items are shown in Table. Items of right arm and leg motor impairment and language had the highest association with other items. On the contrary, extinction neglect, visual field and facial palsy were tenuously linked to other NIHSS items. Exploratory factor analysis revealed four underlying factors for NIHSS items (Table ). Items related to consciousness and right-sided motor impairment as well as language were highly contributed to first extracted factor. Items of left arm and leg motor loaded exclusively on factor. Factor was associated with facial palsy, language and right arm and leg motor impairment. Moreover, factor loading of dysarthria on factor was close to the value of 0.. Finally, factor was primarily comprised of visual field and limb ataxia. Factors to explained.%,.%, 0.0% and.% of the total variance, respectively. Discussion We found that extinction neglect, visual field and facial palsy are least associated to other items. Performing factor analysis on NIHSS items led to a four factor structure. Factors and represent left brain lesions. Factor reflects right brain motor function. Finally, since ataxia and

9 Page of impairment in visual field are associated with vertebra-basilar circulation, factor represents lesions in that territory. To the best of our knowledge, two previous studies have conducted factor analysis on NIHSS items [, ]. Generally, they have found similar results. For instance, both have identified four underlying factors, two for each hemisphere. Further, in those studies factors and were associated with left cortical function and left brain motor function, respectively, and in a similar way factors and represented other side of the brain. In accordance with aforementioned findings, we identified a four factor structure. Similarly, in our study factors and were determined by left brain function and items related to right arm and leg motor involvement as well as language and dysarthria loaded on both of them, simultaneously. Our results showed that much of the total variance attributed to NIHSS items was explained by factor. Nine out of fifteen NIHSS items loaded positively on this factor. Facial palsy, limb ataxia and extinction neglect, which have been reported to have low reliability indices [,, ], are among the remaining items. Moreover, two of the items that did not contributed to the factor were related to right brain motor function. In our study, since factor is associated with limb ataxia and visual field, it represents posterior circulation strokes. None of the previous studies have identified such a factor. Because of low factor loadings, ataxia was not considered as an item of NIHSS in previous studies reporting the underlying factors of NIHSS scale [, ]. On the other hand, in those studies visual field loaded on the second extracted factor, which was suggested to be related to right brain function [, ]. It was postulated that confounding effects of aphasia and extinction neglect might be responsible for that phenomenon []. Therefore, this study for the first time extracts a factor representing posterior circulation strokes.

10 Page of The International Journal of Neuroscience Factors analysis has been widely used to investigate the suitability and internal structure of various scales in psychology and internal medicine [,, 0-]. Nevertheless, factor analysis of the NIHSS items has been the focus of little attention. Thus, data regarding the underlying factor structure of NIHSS are limited. This study provides more insight into the pathophysiologic mechanisms which weaves together much of the NIHSS items. However, the potential limitation of the current study is that we used a hospital-based study population, which may confine our ability to extend our findings. In conclusion, we identified a four factor structure for NIHSS items. Each factor was determined by either left brain, right brain or posterior stroke events. In this study, for the first time visual field and limb ataxia load on a single factor, reflecting posterior circulation strokes. However, further studies may be needed to determine whether this factor model is suitable for different clinical settings and how it improves the ability of NIHSS score to predict stroke outcomes. Declaration of interest The authors declare no conflicts of interests.

11 Page 0 of References [] WHO Global Burden of Disease Estimated Death Number and Mortality Rate, WHO EASTERN MEDITERRANEAN REGION, [] WorldHealthOrganization. The World Health Report: Shaping the Future. Geneva, Switzerland: WHO; 00. [] Feigin VL, Lawes CM, Bennett DA, et al. Stroke epidemiology: a review of population-based studies of incidence, prevalence, and case-fatality in the late 0th century. Lancet Neurol. 00;:-. [] Kasner SE, Chalela JA, Luciano JM, et al. Reliability and validity of estimating the NIH stroke scale score from medical records. Stroke. ;:-. [] Pezzella FR, Picconi O, De Luca A, et al. Development of the Italian version of the National Institutes of Health Stroke Scale: It-NIHSS. Stroke. 00;:-. [] Meyer BC, Lyden PD. The modified National Institutes of Health Stroke Scale: its time has come. Int J Stroke. 00;:-. [] Brott T, Adams HP, Jr., Olinger CP, et al. Measurements of acute cerebral infarction: a clinical examination scale. Stroke. ;0:-0. [] Adams HP, Jr., Bendixen BH, Leira E, et al. Antithrombotic treatment of ischemic stroke among patients with occlusion or severe stenosis of the internal carotid artery: A report of the Trial of Org 0 in Acute Stroke Treatment (TOAST). Neurology. ;:-. [] Nedeltchev K, Renz N, Karameshev A, et al. Predictors of early mortality after acute ischaemic stroke. Swiss Med Wkly. :-. [0] Nogueira RG, Liebeskind DS, Sung G, et al. Predictors of good clinical outcomes, mortality, and successful revascularization in patients with acute ischemic stroke undergoing thrombectomy: pooled analysis of the Mechanical Embolus Removal in Cerebral Ischemia (MERCI) and Multi MERCI Trials. Stroke. 00;:-. [] Lyden PD, Lu M, Levine SR, et al. A modified National Institutes of Health Stroke Scale for use in stroke clinical trials: preliminary reliability and validity. Stroke. 00;:0-. [] Meyer BC, Hemmen TM, Jackson CM, et al. Modified National Institutes of Health Stroke Scale for use in stroke clinical trials: prospective reliability and validity. Stroke. 00;:-. [] Tirschwell DL, Longstreth WT, Jr., Becker KJ, et al. Shortening the NIH Stroke scale for use in the prehospital setting. Stroke. 00;:0-0. [] Esteghamati A, Zandieh A, Khalilzadeh O, et al. Clustering of metabolic syndrome components in a Middle Eastern diabetic and non-diabetic population. Diabetol Metab Syndr. 00;:. [] Esteghamati A, Zandieh A, Khalilzadeh O, et al. Clustering of leptin and physical activity with components of metabolic syndrome in Iranian population: an exploratory factor analysis. Endocrine. 00;:0-. [] Lyden P, Claesson L, Havstad S, et al. Factor analysis of the National Institutes of Health Stroke Scale in patients with large strokes. Arch Neurol. 00;:-0. [] Lyden P, Lu M, Jackson C, et al. Underlying structure of the National Institutes of Health Stroke Scale: results of a factor analysis. NINDS tpa Stroke Trial Investigators. Stroke. ;:-. [] Lyden P, Brott T, Tilley B, et al. Improved reliability of the NIH Stroke Scale using video training. NINDS TPA Stroke Study Group. Stroke. ;:0-. [] Lyden P, Raman R, Liu L, et al. NIHSS training and certification using a new digital video disk is reliable. Stroke. 00;:-. [0] Kendler KS, Aggen SH, Patrick CJ. A Multivariate Twin Study of the DSM-IV Criteria for Antisocial Personality Disorder. Biol Psychiatry. 0;:-.

12 Page of The International Journal of Neuroscience [] Otto C, Bischof G, Rumpf HJ, et al. Multiple dimensions of health locus of control in a representative population sample: ordinal factor analysis and cross-validation of an existing three and a new four factor model. BMC Med Res Methodol. 0;:. [] Liu XQ, Georgiades S, Duku E, et al. Identification of genetic loci underlying the phenotypic constructs of autism spectrum disorders. J Am Acad Child Adolesc Psychiatry. 0;0:- e.

13 Page of Tables: Table. Principal characteristics of the study population. Characteristic Value Age, mean (SD). ±.0 Male gender, %. Mortality, %. NIHSS, mean (SD). (.0) NIHSS, median (IQR).00 (.0) NIHSS, % <. -. >.

14 Page of The International Journal of Neuroscience Table. Correlation coefficients between NIHSS items. LOC LOCquestioncommands field palsy motor motor motor motor ataxia e LOC- Gaze Visual Facial Left arm Right arm Left leg Right leg Limb Sensory Languag LOC-questions 0.*** LOC-commands 0 0.*** 0.*** Gaze 0.0** * Visual field * -0.0*** -0.0** Facial palsy * Left arm motor * 0.*** Right arm motor 0.*** 0.*** 0.*** 0.*** -0.0* 0.*** 0.* Left leg motor *** -0.0 Right leg motor 0.*** 0.*** 0.*** 0.* -0.** 0.*** 0.** 0.*** 0.0 Limb ataxia ** ** -0.** -0.** 0.0*** 0.0*** 0.*** Sensory 0 0.*** 0.*** 0.*** 0.0* *** *** -0.0 Language 0.*** 0.*** 0.0*** 0.0** *** *** - 0.*** -0.0* 0.*** 0.*** Dysarthria 0.*** 0.*** 0.*** 0.0** *** *** *** *** 0.*** Extinction neglect ** -0.0* * * * * P<0.0, **P<0.0, ***P<0.00 Dysarthri a

15 Page of Table. Factor loadings of NIHSS items on the selected factors. Factors LOC 0.* LOC-questions 0.* LOC-commands 0.* Gaze 0.* Visual field * Facial palsy * -0.0 Left arm motor * Right arm motor 0.* * -0. Left leg motor 0. 0.* Right leg motor 0.* * -0.0 Limb ataxia * Sensory 0.* Language 0.0* * -0. Dysarthria 0.* Extinction neglect * * factor loadings 0..