Evaluation of Ataxia in Mild Ischemic Stroke Patients Using the Scale for the Assessment and Rating of Ataxia (SARA)

Article information

Ann Rehabil Med. 2018;42(3):375-383

Publication date (electronic) : 2018 June 27

doi : https://doi.org/10.5535/arm.2018.42.3.375

Sung Won Choi, MD, Nami Han, MD, PhD, Sang Hoon Jung, MD, Hyun Dong Kim, MD, PhD, Mi Ja Eom, MD, Hyun Woo Bae, MD

Department of Physical Medicine and Rehabilitation, Inje University Busan Paik Hospital, Inje University College of Medicine, Busan, Korea

Corresponding author: Nami Han Department of Physical Medicine and Rehabilitation, Inje University Busan Paik Hospital, 75 Bokji-ro, Busanjin-gu, Busan 47392, Korea. Tel: +82-51-890-6295, Fax: +82-51-891-1430, E-mail: rehabit@inje.ac.kr

Received 2017 July 19; Accepted 2017 September 11.

Abstract

INTRODUCTION

Ataxia is a neurological dysfunction of motor coordination, which affects the fundamental activities such as gaze, speech, gait, and balance [1]. It induces disorders of limb movement such as dyssynergia, dysmetria, kinetic or postural tremor and dysdiadochokinesia. Ataxia also affects trunk control resulting in balance or gait dysfunction [2]. These manifestations of ataxia interrupt physical performances and often contribute to poor functional outcome despite relatively well preserved muscle strength [3].

Although ataxia is caused mostly by cerebellar lesions, lesion in the basis pontis, corona radiata, thalamus, posterior limb of internal capsule also cause ataxia with ipsilateral pyramidal signs, and most of these regions are associated with posterior circulation [4,5]. This clinical syndrome is known as ataxic hemiparesis (AH), which was first defined by Fisher and Cole [4] and later modified by Fisher [5]. AH was initially described as a lacunar syndrome correlating with lacunar infarctions, which resulted from the occlusion of small penetrating arteries. However, recent studies demonstrated that cardioembolic and large-artery atherosclerosis as well as lacunar infarctions cause AH [6]. Therefore, AH is not a rare clinical manifestation among ischemic stroke patients.

The National Institutes of Health Stroke Scale (NIHSS) provides a quantitative measure of key symptoms in stroke patients and consists of 15 subscales: level of consciousness, questions, commands, visual fields, facial palsy, left/right arm/leg strength, ataxia, sensation, language, dysarthria and extinction/inattention. It is most widely used as a standard neurologic examination for the assessment of acute stroke patients, and its reliability, validity and predictive value is established [7,8]. However, NIHSS has limited application for features of posterior circulation stroke such as cranial nerve signs or ataxia because it mostly focuses on features of anterior circulation stroke [9,10]. NIHSS contains only one item assessing limb ataxia, and none related to trunk function.

Impaired trunk control is common in stroke patients even if limb ataxia is not manifested explicitly. Ataxia affects trunk control resulting in balance or gait dysfunction, increased risk of fall, and decline in independence of activities of daily living (ADL). Studies reported that the balance status is related to hospitalization period [11]. Trunk control in early stages of stroke is a predictor of comprehensive ADL function at 6 months [12]. Standing balance is correlated with change in functional state over the course of acute rehabilitation [13]. Sitting balance also shows strong correlation with ADL function [14]. Therefore, appropriate trunk control is fundamental to motor function and ADL. Apparently, detailed assessment of not only muscle strength but also trunk and limb control is essential for comprehensive evaluation of stroke patients.

Among the scales for ataxia, the International Cooperative Ataxia Rating Scale (ICARS) [15] is widely used as a tool to evaluate the severity and treatment efficacy. Consisting of 19 sub-items divided into 4 subscales, various ataxic symptoms can be assessed comprehensively with ICARS [1]. However, its clinical utility is questionable because of the prolonged duration to complete the test.

Schmitz-Hubsch et al. [16] devised a new scale known as the Scale for the Assessment and Rating of Ataxia (SARA) in 2006 to assess cerebellar ataxia. SARA is semiquantitative and is a 40-point scale consisting of 8 items: gait, stance, sitting, speech disturbance, finger chase, nose-finger test, fast alternating hand movements, and heel-shin slide. SARA has fewer scoring items than ICARS, and is therefore, faster than ICARS. SARA is more efficient than ICARS for the assessment of ataxia of trunk because it has 3 items related to trunk control (gait, stance, and sitting). SARA is characterized by high interrater reliability, test-retest reliability and internal consistency in the spinocerebellar atrophy patient group [16]. Further, Kim et al. [17,18] presented that SARA was useful as a functional measure in stroke patients due to high intra-rater and inter-rater reliability. However, subjects in these studies included patients with both ischemic and hemorrhagic stroke characterized by lesions of either anterior or posterior circulation and at varying severities of stroke. Their functional severity was not considered except for muscle strength measured using the Medical Research Council (MRC) scale [17,18].

Therefore, we focused on subjects diagnosed with mild stroke including NIHSS scores ranging from 0 to 5, and ischemic lesions of cerebellum and AH region. The specific definition of ‘mild’ has not been agreed upon universally, and there is variability in the interpretation and implementation [19]. In this study, we used the definition of ‘mild’ as NIHSS 0–5, referring to the previous studies of Spokoyny et al. [19] and Logallo et al. [20]. We investigated the need and utility of SARA for assessment of posterior circulation symptoms related to motor control function, which can be overlooked during assessment only based on NIHSS.

MATERIALS AND METHODS

RESULTS

Demographic characteristics of the subjects are summarized in Table 1. The mean age was 63.96±11.57 years, and the participants included 25 males and 20 females. The mean±standard deviation of subjects’ duration from onset to admission was 1.64±2.72 days, and that of hospitalization period was 10.36±4.46 days. Among the subjects, 48.9% (22/45) scored SARA above 5.5, and even 11.1% (5/45) scored above 14.25 which is the cut-off value of ‘severe dependency’ in ADL (Fig. 1).

Table 1.

Patient demographics (n=45)

Fig. 1.

Scale for the Assessment and Rating of Ataxia (SARA) score graded by the cut-off value of severity in dependency of activities of daily living (ADL). SARA was assessed immediately after the patients’ release from absolute bed rest (within 5 days from admission). Values are presented as mean±standard deviation, ratio (%), and number of patients.

The stepwise multiple linear regression analysis showed that SARA and TCT scores were significant predictors of mRS at discharge in patients with mild ischemic stroke (Table 2). Other variables (BBS and TUG) were originally included in analysis but not presented because of insignificant relationships with mRS at discharge after controlling for the variables. The largest variance inflation factor was 1.194, indicating lack of multi-collinearity among variables. Residual error showed normal distribution. In predicting mRS at discharge, the regression coefficient of SARA was 0.223, which was statistically significant (p<0.001). The SARA score was the strongest predictive variable among the four functional outcome measures (standardized β coefficient : SARA=0.873, TCT=0.345). The coefficient of determination was 61.5%.

Table 2.

Stepwise multiple linear regression analysis of variables predicting functional outcomes related to ‘mRS at discharge’ in patients diagnosed with mild ischemic stroke with acute ataxia

The SARA score showed strong negative correlation with BBS in mild ischemic stroke patients (r=-0.946, p<0.001) (Fig. 2A). It also indicated that SARA showed strong negative correlation with TCT (r=-0.799, p<0.001) (Fig. 2C). However, the ceiling effect should be considered for accurate interpretation because 85.7% of subjects obtained complete scores on TCT. The SARA score showed positive correlation with TUG (r=-0.584, p<0.001) (Fig. 2B). The subjects were assigned to groups A and B based on the mRS at admission (group A, mRS 0–2; group B, mRS 3–5). Group A included 26 subjects and group B included 19 subjects. The group B showed weaker results in all assessments compared with group A. Significant differences were found in SARA as well as BBS, TUG and TCT scores between groups A and B (Table 3). No significant difference in cognitive function (Mann-Whitney test of MMSE, p=0.677) existed between the two groups. Other complications that affected the functional status were not detected.

Fig. 2.

Pearson correlation analysis between the SARA total scores and BBS (A), TUG (B), and TCT (C). Measuring tools were assessed immediately after the patients’ release from absolute bed rest (within 5 days from admission). SARA, Scale for the Assessment and Rating of Ataxia; BBS, Berg Balance Scale; TUG, Timed Up-andGo; TCT, Trunk Control Test. ***p<0.001.

Table 3.

Comparison of SARA, BBS, and TCT between groups A and B

DISCUSSION

The term, ‘mild’ ischemic stroke suggests possible favorable outcomes. However, several previous studies demonstrated otherwise. Schlegel et al. [23] reported that not all (81%) of mild stroke patients were discharged home, which suggests that 19% of mild stroke patients required long-term medical support such as rehabilitation or nursing facilities after acute period. Another study showed that a third of all mild stroke patients were not functionally independent after 90 days from the onset [19]. This mismatch was attributed to other symptoms not focused by NIHSS. Even patients scoring zero on NIHSS often presented with other stroke symptoms, especially trunk ataxia (45%) [24]. In this study, SARA played an adjunct role in differentiating the presence of ataxic symptoms and grades. Subjects in this study had homogeneous scores of NIHSS (0–5) and all were graded as mild stroke. However, their SARA scores were distributed across a wide range (Fig. 1). Mild ischemic stroke patients should evaluated with SARA as well as NIHSS and an individualized treatment plan should be developed.

Especially in patients carrying a posterior circulation lesion, the initial NIHSS score is often lower. In a recent cohort study, 76% of patients with posterior circulation stroke presented with baseline NIHSS scores ≤5 [25]. Again, lower NIHSS scores are often interpreted as assurance of a favorable outcome. However, in one retrospective study, death or disability occurred in 15% of patients with posterior circulation stroke with NIHSS scores ≤4 at 3-month follow-up as a direct consequence of their stroke [26]. In subjects enrolled in this study, who had mild ataxic hemiparesis, 48.9% of subjects (22/45) scored above the cut-off value suggesting ‘moderate dependent’ and 11.1% (5/45) showed that they needed maximal assistance in ADL. It is clear that NIHSS cannot be used as an independent and comprehensive evaluation tool for posterior circulation stroke. Other tools such as SARA are required for concurrent evaluation and reliable prediction with NIHSS.

BBS, TUG, and TCT are typical tools for assessment of overall motor control function, and they include items for evaluation of trunk control as well. Since these tools do not entail additional cost or equipment, and are rapid and easy to use, they are widely used in the clinical setting. BBS is a simple clinical tool for assessing static and dynamic balance in sitting or standing positions. Validity, reliability and sensitivity of BBS are well established in stroke patients. BBS consists of 14 subscales with a total score of 56, and a higher score suggests better functional state [27,28]. If the patient can walk, TUG test is reliable and applicable for the measurement of functional mobility based on the time taken by a patient to rise from a chair, walk 3 m, turn around and back to the chair and sit down [29-31]. TCT is not only reliable and valid, but also sensitive for the assessment of the progression of stroke. It consist of 4 items of functional movement (rolling to weak side, rolling to strong side, balance in sitting position, and sitting up from lying down) [32-34]. Based on the certainty and universality of these tools in assessing the overall limb and trunk motor function, we analyzed their correlation with SARA to determine the value of SARA in motor functional assessment of patients with mild ataxic hemiparetic stroke.

In regression analysis, SARA showed statistically significant results for prediction of mRS at discharge. Compared with BBS, TUG and TCT, SARA showed the highest standardized β coefficient (=0.873), which suggests that SARA has the strongest predictive value for functional prognosis. There are many factors for predicting the stroke patient’s functional outcome, and the strongest and most consistent predictor is the functional ability at admission. [35]. In a patient with mild stroke manifesting impaired limb or trunk control associated with ataxia, SARA is a sound tool for the prediction of functional outcome at discharge.

SARA showed a strong correlation with BBS and significant correlation with TUG and TCT (p<0.001) in mild stroke patients. However, most of the subjects scored complete TCT scores, and the result showed a ceiling effect. TCT comprises items of movement with relatively low level of functional activity compared with BBS or TUG, which contributed to the emergence of ceiling effect in mild (NIHSS 0–5) stroke patients. In the comparison between group A with group B, SARA as well as BBS, TUG and TCT showed statistical significance in functional severity.

Until now, few studies investigated the utility of SARA in assessing posterior circulation in patients with mild (low NIHSS) stroke. Kim et al. [18] reported the benefit of SARA in that it corresponded with the gait status and ADL dependency. However, no criteria were available to determine the severity excluding MRC or criteria for stroke lesions. Accordingly, this study represents an advance over previous studies.

The study limitation relates to the small sample size of subjects and non-standardized hospital period for follow-up of mRS. Several patients were excluded during screening due to failure to undergo SARA, BBS, TUT, TCT, and mRS testing. As a retrospective study, the hospitalized period and test duration were not controlled. As the study of Jorgensen et al. [36] which showed that 95% of 239,886 patients diagnosed with stroke attained optimal neurological recovery in 3 months, scores of long-term follow-up are needed. Furthermore, most of the subjects were not transferred to rehabilitation units so that their medical records did not sufficiently cover the data related to functional ability such as Modified Barthel Index (MBI) or Functional Independence Measure (FIM). To investigate the predictive value of functional outcome using SARA, only mRS was included in this study. Further, the correlation between SARA and TCT showed a ceiling effect suggesting that the result was of limited value. A prospective study design with sufficient sample size and controlled follow-up period is needed. Functional outcome assessment with more detailed and universal tools used in rehabilitation, such as MBI or FIM, is also needed.

Considering limited resources and patients’ clinical frustration, efficient selection of measurement tools is essential. Carefully selected tools of assessment facilitate the most appropriate and critical evaluation rapidly and cost-effectively, eliminating redundancy. SARA is a more efficient and complementary evaluation tool compared with other conventional tools such as BBS, TUG, and TCT for the assessment of mild stroke patients manifesting symptoms of posterior circulation.

In conclusion, SARA represents a useful tool for evaluation of severity of deficit and prediction of functional outcome in patients with mild ischemic stroke characterized by symptoms of posterior circulation. SARA is recommended as an adjunct to NIHSS in the initial evaluation of patients with mild ischemic stroke.

Article information Continued