European Journal of Physical and Rehabilitation

COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically © 2018 EDIZIONI MINERVA MEDICA European Journal of Physical and Rehabilitation Medicine 2019 February;55(1):1-7 or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access Online version at http://www.minervamedica.it DOI: 10.23736/S1973-9087.18.05161-4 to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. ORIGINAL ARTICLE Assessing functional recovery in the first six months after acute ischemic stroke: a prospective, observational study João P. BRANCO 1, 2, 3 *, Sandra OLIVEIRA 3, João SARGENTO-FREITAS 1, 4, Jorge LAÍNS 3, João PINHEIRO 2, 4 1Faculty of Medicine, University of Coimbra, Coimbra, Portugal; 2Department of Physical and Rehabilitation Medicine, University Hospital Center of Coimbra, Coimbra, Portugal; 3Department of Physical and Rehabilitation Medicine, Centro de Medicina de Reabilitação da Região Centro, Rovisco Pais, Tocha, Portugal; 4Department of Neurology, University Hospital Center of Coimbra, Coimbra, Portugal *Corresponding author: João P. Branco, Faculty of Medicine of the University of Coimbra, Pólo I - Edífico Central, Rua Larga, 3004-504 Coimbra, Portugal. E-mail: [email protected] ABSTRACT BACKGROUND: Stroke is a major cause of short-term functional impairment, but the recovery timeframes are not well-established. AIM: This study aims to evaluate the progression of functional recovery during the first 24 weeks after acute ischemic stroke, to determine the timeframes for motor, cognitive, and overall recovery. DESIGN: Prospective observational study. SETTING: Tertiary care center; 12-week inpatient period, followed by 12-week outpatient period. POPULATION: A group of 131 patients with acute stroke in the territory of the middle cerebral artery with age 18-85 years-old. METHODS: Patients received treatment according to routine clinical practice and underwent a closely-controlled rehabilitation program for 24 weeks. Functionality assessments were conducted at 48 hours, 3 weeks, 12 weeks, and 24 weeks after acute stroke and included the modified Rankin Scale (mRS), the Functional Independence Measure (FIM), and the Stroke Upper Limb Capacity Scale (SULCS). RESULTS: Over the study period, patient functionality improved significantly (P<0.001) as measured by all scales. Assessment scores improved significantly from 48 hours to 3 weeks for all scales; the same occurred from 3 weeks to 12 weeks, except for C-FIM. From 12 weeks to 24 weeks, there were no statistically significant functional improvements for any scale. In comparative terms, cognitive impairment was less severe than motor disability in the acute phase. C-FIM scores at 48 hours were significantly higher than M-FIM and mRS scores. Upper limb function- ing measured by SULCS, showed a intermediate degree of disability compared to the C-FIM, M-FIM, and mRS. CONCLUSIONS: Functional recovery occurs at least until 24 weeks after acute stroke, but most of the functional gains tend to be achieved dur- ing the first 12 weeks. Cognitive function tends to improve earlier than motor function, with the most substantial gains occurring within the first three weeks. From 12 to 24 weeks there are observable numerical gains in patient functionality, highlighting the need to maintain an adequate rehabilitation program. CLINICAL REHABILITATION IMPACT: This study provides insight into the recovery timeframe for stroke patients, which can support the development of more effective rehabilitation programs. (Cite this article as: Branco JP, Oliveira S, Sargento-Freitas J, Laíns J, Pinheiro J. Assessing functional recovery in the first six months after acute ischemic stroke: a prospective, observational study. Eur J Phys Rehabil Med 2019;55:1-7. DOI: 10.23736/S1973-9087.18.05161-4) Key words: Stroke - Rehabilitation - Upper extremity. Stroke is a major cause of short-term functional impair- Cognitive and motor impairments are major causes of ment in developed countries, which can lead to sig- disability after stroke.8, 9 Cognitive impairment can, by nificant long-term functional impairment in the absence itself, result in long-term patient dependency, but most of adequate treatment and rehabilitation.1-4 Up to 70% of frequently, patients regain cognitive function shortly after stroke patients show hemiparesis in the acute phase and stroke and continue to reveal motor impairment for longer 40% of those do not regain independency in the execution periods.9-11 Lack of control of the extremities is usually of activities of daily living (ADL) within the first 6 months the main driver of motor impairment in this context. While after stroke.5-8 65% patients show full recovery of lower limb function, Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 1

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, BRANCO FUNCTIONAL RECOVERY AFTER STROKE cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. recovering upper limb function is generally more difficult. After enrolment, patients underwent a standard treat- Some studies indicate that only 15% of patients actually ment and rehabilitation program for 24 weeks, accord- recover hand function, which significantly impacts the ing to routine clinical practice. Functionality assessments execution of ADLs and, consequently, overall quality of were conducted at 48 h, 3 weeks, 12 weeks, and 24 weeks life.5, 6, 8 after acute stroke. Despite the impact of stroke on functional capacity, the The study was evaluated and approved by the Scientific timeframes for functional recovery after stroke are not Council of the Ethics Committee of the Faculty of Medi- fully established. Most authors agree the vast majority of cine of the University of Coimbra (reference letter 104-CE- functional gains tend to occur in the first weeks after the 2014). Patients or their legal representatives provided their acute episode, but there is no consensus on the timeframe written informed consent prior to enrolment in the study. from which substantial gains should no longer be expect- ed.4-6, 8, 12, 13 Some studies suggest that recovery tends to Treatment and rehabilitation plateau at 4 weeks after stroke, but others note meaningful recovery at least until 12 weeks.5, 6, 8, 14 In any case, few During the period of the study, patients received treat- studies actually evaluated long-term functional recovery ment according to routine clinical practice and underwent after an acute episode of stroke, which is a significant limi- a closely-controlled rehabilitation program. In the first tation when designing stroke rehabilitation programs. 12 weeks participants received inpatient care at the study center, which included 60 minutes of physical therapy, 30 Therefore, there is a need to accurately assess the level minutes of occupational therapy, and 30 minutes of speech of functional recovery over longer periods of time. Better therapy (if aphasia was present) daily. Between week 12 establishing the timeframes in which patients are expected and 14 patients were discharged, following clinical assess- to show greater functional gains would help allocate re- ment, and continued as outpatients until week 24. After sources more effectively, providing the basis for better fol- discharge, patients were instructed to continue the daily low-up and treatment. This type of evaluation should use rehabilitation program at home, including 60 minutes of standardized scales that could be easily applied in clinical physical therapy, 30 minutes of occupational therapy, and practice to assist in the development of highly effective 30 minutes of speech therapy (if aphasia was present). The treatment and rehabilitation programs, with the ultimate study team made all reasonable efforts to ensure patients goal of ensuring that all patients reach their full recovery complied with the rehabilitation program. potential. Neurological rehabilitation exercises included balance- This study aims to evaluate the progression of function- coordination training, hand rehabilitation, stretching and al recovery during the first 24 weeks after acute ischemic relaxation exercises, walking exercises, and posture ex- stroke, in order to determine the timeframes for motor, ercises.15 Exercises focused mainly on using an affected cognitive and overall recovery. limb, symmetric weight bearing and transfer, mat activity, and gait training.5 Materials and methods Functional assessment Study design and sample Clinical and demographic characteristics were assessed We conducted a prospective observational study of pa- after admission, during the initial clinical evaluation, by a tients hospitalized for acute ischemic stroke in a tertiary neurologist. Then, patient functionality was assessed over care center in central Portugal (Centro Hospitalar e Uni- time using several scales at 4 different time points: (1) 48 versitário de Coimbra, Coimbra, Portugal). hours, (2) 3 weeks, (3) 12 weeks, and (4) 24 weeks after stroke. These scales included the modified Rankin Scale All patients with age 18-85 years and clinical and ra- (mRS), the Functional Independence Measure (FIM), and diological evaluation indicative of ischemic stroke in the the Stroke Upper Limb Capacity Scale (SULCS). territory of the middle cerebral artery (confirmed through head CT) were eligible to participate. Patients were ex- mRS is a widely used assessment scale that evaluates the cluded from the study if symptom onset occurred more degree of disability and dependency in activities of daily than 3 hours prior to hospital admission, if there were living.6 The scale is divided into 7 degrees of ascending dis- signs of hemorrhagic transformation, if the they showed ability, in which 0 corresponds to patients with no visible premorbid disability (mRS>2), and if there were other symptoms of disability and 6 corresponds to death. In this neurological, inflammatory, or neoplastic conditions. study, the Portuguese validated version of mRS was used. 2 European Journal of Physical and Rehabilitation Medicine February 2019

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, FUNCTIONAL RECOVERY AFTER STROKE BRANCO cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. FIM allows the measurement of disability in both physi- Patients admitted with acute cal and cognitive dimensions.16 The scale consists of 18 ischemic stroke of the MCA confirmed items scored 1 to 7, in which higher scores indicate high- er degree of functioning and independence in executing with neuroimaging (N.=556) ADLs. The items fall into 2 domains: (1) motor domain (M- FIM), which evaluates self-care, sphincter control, trans- Total excluded (N.=390) fers, and locomotion, and (2) cognitive domain (C-FIM), - No CT at admission (N.=25) which evaluates communication and social cognition. The - Stroke >3 hours prior to admission (N.=55) Portuguese translation of FIM currently recommended by - Premorbid mRS>2 (N.=45) the National Health General Directorate was used. - Active infection (N.=35) - Renal insufficiency (N.=25) SULCS is a unidimensional, hierarchical scale specifi- - Active malignancy (N.=36) cally designed to evaluate upper limb functioning in stroke - No blood sample within 48 hours of admission patients.17, 18 It is based on the execution of 10 tasks of increasing difficulty. The score corresponds to the number (N.=45) of tasks the patient is able to execute. Scores 0-2 indicate - P remorbid neurocognitive pathology (N.=50) no hand function, scores 3-7 indicate basic hand function, - Symptom onset >24 hours (N.=28) and scores 8-10 indicate good hand function. The recently - Recent surgery (N.=24) published Portuguese version of SULCS was used in this - Coma (N.=22) study.17 Patients enrolled in the study (N.=166) Follow-up over 6 months Total excluded (N.=35) - Hemorragic trasformation (N.=10) - Loss to follow-up (N.=14) - Death (N.=11) Statistical analysis Final analysis population (N.=131) Quantitative variables are presented as mean±SD if nor- Figure 1.—Study flowchart. mality of distribution is confirmed through the Shapiro- MCA: middle cerebral artery. Wilk Test or otherwise presented as median plus 25th, 75th percentiles. Categorical variables are presented as absolute rolled in the study (Figure 1 details the reasons for exclu- and relative frequencies. sion of the remaining patients). Later, 35 patients were excluded from the final analysis population (N.=131), due Changes in functionality throughout the study period to hemorrhagic transformation (N.=10), loss to follow-up were assessed with 5 variables: mRS, SULCS, Motor- (N.=14), and death (N.=11). FIM, Cognitive-FIM, and Total-FIM. To test for changes in these variables over time, one-way repeated measures Table I summarizes the demographic and clinical char- analysis of variance (ANOVA) or the Friedman test were acteristics of participants. Mean±standard deviation age applied as appropriate. in the study population was 68.82±11.88 years. Genders were overall evenly represented, but women tended to be In a later phase, measured scores were transformed in older than men, being overrepresented in the 66-79 and percentages of the maximum score for each variable, to as- ≥80 age strata (P=0.002). The most frequent comorbidi- sess the relative change in functionality observed in differ- ties were hypertension, dyslipidemia, and atrial fibrilla- ent scales. Two-way repeated measures ANOVA was ap- tion. Older patients tended to show higher rates of atrial plied, considering the interaction of time with the several fibrillation, hypertension, and diabetes, while younger pa- functionality variables. tients tended to show higher rates of dyslipidemia, hyper- uricemia, obesity, and heart failure. However, these trends IBM SPSS Statistics version 23 (IBM Corp, Armonk, were mostly non-significant, except for atrial fibrillation NY, USA) was used for statistical analysis, adopting a 5% (P=0.031). significance level. Functional recovery over 24 weeks post-stroke Results Table II presents median scores observed in all functional Study population assessments applied in the study. Over the study period patient functionality improved significantly (P<0.001) as A total of 556 patients were admitted to the study center measured by all scales. From 48 hours to 3 weeks there during the enrolment period with acute stroke. Of those, 166 patients fulfilled the selection criteria and were en- Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 3

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, BRANCO FUNCTIONAL RECOVERY AFTER STROKE cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. Table I.—Clinical and demographic characteristics of participants stratified by age. Gender, N. (%) 18-65 years (N.=45) 66-79 years (N.=63) ≥80 years (N.=23) P value Female (N.=66) Male (N.=65) 15 (33.3) 33 (52.4) 18 (78.3) 0.002 30 (66.7) 30 (47.6) 5 (21.7) Comorbidities/lifestyle factors, N. (%) 0.031 Atrial fibrillation (N.=59) 19 (42.2) 24 (38.1) 16 (69.6) 0.529 Dyslipidemia (N.=59) 22 (48.9) 29 (46.0) 8 (34.8) 0.076 Hypertension (N.=91) 26 (57.8) 46 (73.0) 19 (82.6) 0.187 Diabetes (N.=38) 10 (22.2) 18 (28.6) 10 (43.5) 0.933 Hyperuricemia (N.=16) 3 (13.0) 0.124 Obesity (N.=25) 6 (13.3) 7 (11.1) 1 (4.3) 0.897 Heart failure (N.=27) 11 (24.4) 13 (20.6) 4 (17.4) 0.847 Smoking (N.=16) 10 (22.2) 13 (20.6) 2 (8.7) 6 (13.3) 0.882 Laterality, N. (%) 8 (12.7) 12 (52.2) Right (N.= 65) 21 (46.7) 11 (47.8) 0.312 Left (N.= 66) 24 (53.3) 32 (50.8) 31 (49.2) 14 (60.9) Aphasia, N. (%) 23 (51.1) 9 (39.1) Yes (N.= 64) 22 (48.9) 27 (42.9) No (N.= 67) 36 (57.1) Table II.—M edian scores in patient functionality assessments over the first 24 weeks after stroke. Assessment scales Effect of time 48 hours 3 weeks P value 12 weeks P value 24 weeks P value P value (48 h vs. (3 vs. (12 vs. 3 weeks) 12 weeks) 24 weeks) SULCS (0-10) <0.001 1 (0; 7) 3 (1; 9) <0.001 5 (1; 10) 0.019 6 (3; 10) NS mRS (0-5) <0.001 4 (4; 5) 4 (2; 4) <0.001 3 (1; 4) 0.008 2 (1; 4) NS M-FIM (13-91) <0.001 20 (13; 33) 44 (20; 87) <0.001 66 (31; 91) 0.002 72 (35; 91) NS C-FIM (5-35) <0.001 16 (5; 30) 23 (13; 31) 25 (17; 33) NS 27 (17; 35) NS T-FIM (18-126) <0.001 36 (18; 61) 68 (35; 115) 0.002 93 (53; 120) 0.001 96 (56; 123) NS <0.001 C-FIM: Cognitive Functional Independence Measure; M-FIM: Motor Functional Independence Measure, mRS: modified Rankin Scale; SULCS: Stroke Upper Limb Capacity Scale; NS: non-significant; T-FIM: Total Functional Independence Measure. Table III.—R esults of two-way repeated measures analysis of vari- time points. These results indicate significantly different ance. rates of recovery for the different facets of patient func- tionality assessed by these scales. Variables Degrees of F P freedom Table IV provides a comparison of functional gains measured through the various scales, expressed as per- Assessment scales 2.332 56.061 <0.001 centage of the maximum score, and Figure 2 provides a Time 2.196 132.940 <0.001 visual representation of the relative gains over 24 weeks Assessment scales x time 5.548 <0.001 post-stroke. Cognitive impairment was less severe than 22.161 motor disability in the acute phase. C-FIM scores at 48 hours were significantly higher than M-FIM and mRS were significant functional improvements for all assess- scores, and remained higher for the different time points, ment scales. From 3 weeks to 12 weeks there were also although the differences tended to be reduced overtime. significant functionality improvements for all scales, ex- Laterality did not significantly affect cognitive function at cept C-FIM. Finally, during the outpatient period, from 12 the different time points (P>0.05). Patients with aphasia weeks to 24 weeks, there were no statistically significant (N.=64) showed significantly worse cognitive function functional improvements for any scale, although median compared to those without aphasia at 48 hours (P<0.001) scores were numerically higher for all scales at 24 weeks and 3 weeks (P<0.001); at 12 and 24 weeks differences compared to 12 weeks. were non-significant. Table III presents the results of repeated measures two- At 48 hours, scores for both M-FIM and mRS were way ANOVA, which revealed a significant interaction (P<0.001) between the different scales used and follow-up 4 European Journal of Physical and Rehabilitation Medicine February 2019

COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA FUNCTIONAL RECOVERY AFTER STROKE BRANCO This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow accessTable IV.—Relative change in patient functionality according to different assessment scales. to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher.Assessments 48 h 3 weeks 12 weeks 24 weeks scales Percentage of maximum score Estimate (SE) 95% CI Estimate (SE) 95% CI Estimate (SE) 95% CI Estimate (SE) 95% CI SULCS 32.75 (3.19)* ° 26.45-39.05 47.56 (3.39) 40.86-54.26 53.66 (3.36) 47.03-60.30 59.01 (3.26) 52.56-65.45 mRS 16.49 (1.77)§ # 12.99-19.99 39.85 (3.15)# 33.63-46.07 48.55 (3.08)# ^ 42.46-54.64 53.74 (3.07)# 47.68-59.81 M-FIM 17.53 (2.17)§ # 13.25-21.81 49.34 (3.42) 42.58-56.11 60.03 (3.20) 53.69-66.37 64.24 (3.31) 57.70-70.78 C-FIM 43.05 (3.48)* ° ^ 36.16-49.95 55.52 (3.06)* 49.47-61.58 63.46 (2.79)* 57.93-68.99 66.77 (2.71)* 61.41-72.12 T-FIM 24.63 (2.37)# 19.94-29.31 50.80 (3.26) 44.35-57.24 61.30 (2.99)* 55.38-67.22 64.90 (3.01) 58.95-70.85 Data presented as percentage of the maximum score for each scale, summarized here as mean (standard error) and 95% Confidence interval. Post-hoc multiple comparison analysis: *significant versus mRS; °significant versus M-FIM; §significant versus SULCS; #significant versus C-FIM; ^significant versus T-FIM. C-FIM: Cognitive Functional Independence Measure; CI: confidence interval; M-FIM: Motor Functional Independence Measure; mRS: modified Rankin Scale; SE: standard error; SULCS: Stroke Upper Limb Capacity Scale; T-FIM: Total Functional Independence Measure. very low in comparative terms — not even reaching 20%, 70 compared to 43% for C-FIM — but improved vastly at 3 weeks and continued to show functional gains at 12 60 weeks and 24 weeks, though in the later phase the rate of improvement was substantially reduced. Upper limb 50 functioning measured by SULCS, showed a somewhat intermediate degree of disability compared to the previ- 40 ously mentioned cognitive and motors facets of patient functionality. Importantly, SULCS percentage scores 30 SULCS were significantly higher than mRS scores at 48 hours. mRS The rates of improvement overtime with SULCS score were comparable to those of C-FIM, with greater im- 20 M-FIM provements in the first 3 weeks followed by more residual C-FIM gains at 12 and, especially, at 24 weeks. 10 T-FIM Discussion 0 3 weeks 12 weeks 24 weeks We assessed the long-term functional recovery after acute 48 hours stroke of the middle cerebral artery and found significant functional improvements up to 3 months (12 weeks), fol- Time after stroke lowed by only numerical improvements until 6 months (24 weeks) after stroke. Cognitive function tends to be less im- Figure 2.—Comparison of the percentage of functional gains according paired by stroke than motor function and shows early re- to different assessment scales. covery, which occurs mostly within the first 3 weeks. Mo- See Table IV for post-hoc analysis results. tor function is severely impacted by stroke and although it C-FIM: cognitive functional independence measure; M-FIM: mo- shows stark recovery within the first 3 weeks, significant tor functional independence measure; mRS: modified Rankin Scale; motor function improvements continue to be observed up SULCS: Stroke Upper Limb Capacity Scale; T-FIM: total functional to 12 weeks, when the rate of recovery is substantially re- independence measure. duced. found a very high rate of motor function recovery in the The rate of functional recovery observed in this study first 4 weeks, which was then somewhat reduced until did not follow a linear trend. Overall, patients showed sta- week 12, at which point 91% of the maximum gains had tistically significant functional improvements in the first already been achieved. Verheyden et al.8 highlighted the 3 to 12 weeks, followed by non-significant, lower magni- first 12 weeks as the most important period for upper and tude improvements from 12 to 24 weeks. Previous stud- lower limb functional recover, with residual gains until ies assessing functional recovery in the post-stroke period 24 weeks. These results are comparable to our findings, also identified non-linear rates of recovery.6 Lee et al.5 although comparisons with such studies are limited and should be taken with caution, due to the use of different assessment scales, different samples (with both ischemic and hemorrhagic stroke), different clinical characteristics, and different artery territory involved. In the first 3 weeks, the rate of functional recovery was substantially higher than in the remaining periods and after 12 weeks there were no statistically significant improvements, which can be explained by increased neuroplasticity in the first 12 weeks after stroke.19 Nonetheless, it is important to con- Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 5

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, BRANCO FUNCTIONAL RECOVERY AFTER STROKE cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. sider that there were numerical gains throughout the en- Limitations of the study tire study period. Although these did not reach statistical significance in the aggregate analysis, such numerical dif- This study has some limitations to consider. The study ferences might still have clinically relevant implications population is relatively heterogeneous — especially in in some patients, which could, ultimately, translate into terms of age — potentially leading to different patterns of patients’ dependency or independency in the execution of functional recovery, although we enrolled a substantially ADLs. larger sample size than comparable previous studies. The type of rehabilitation received during the study period (12 The FIM Scale was adopted in this study because it weeks inpatient daily rehabilitation followed by 12 weeks is frequently used in clinical practice and is designed to outpatient daily rehabilitation) can also introduce some measure disability in both the cognitive and motor do- limitations when interpreting the functional gains (or lack mains.15, 16 In terms of the cognitive domain, it is esti- thereof in the outpatient setting) in the different periods. mated that 30-50% of patients show cognitive impairment The observational, routine clinical practice nature of this after stroke.9 Mellon et al.9 characterized cognitive dis- study prevented the use of the same type of rehabilitation ability after stroke using Montreal Cognitive Assessment for the entire duration of the study. Still, this approach fol- (MoCA) and found that cognitive impairments tend to re- lows current best practices for rehabilitation of stroke pa- solve rapidly after acute stroke. In this study, there were tients, providing insight into the expected recovery in rou- no significant improvements in cognitive function after tine clinical practice. In the future, real-world studies with 3 weeks and approximately 50% of patients continued to larger populations and longer follow-up times will be of show some degree of cognitive impairment 24 weeks after great interest to establish the patterns of functional recov- stroke. These results are comparable to our findings. Aydin ery and identify the most effective rehabilitation strategies. et al.15 found that contrary to the cognitive domain, the To- tal FIM scores (which encapsulate the results of the motor Conclusions domain) improve significantly until 12 weeks after stroke, which is also comparable to our findings. Functional recovery occurs at least until 24 weeks after acute stroke, but most the of the functional gains tend to The mRS was used because it is one of the most well- be achieved during the first 12 weeks, reflecting the well- established scales in the assessment of patient functionality established period of increased neuroplasticity. Cognitive after stroke.20 We also adopted SULCS to specifically as- function tends to improve earlier than motor function, sess upper limb functionality, since it is one of the most im- with the most substantial gains occurring within the first 3 pactful facets of overall functionality for the execution of weeks. From 12 to 24 weeks there are observable numeri- ADLs.21 Interestingly, the relative comparison of percent- cal gains in patient functionality, highlighting the need to age scores for SULCS and mRS indicates that these scales maintain an adequate rehabilitation program, even if in an are not very well correlated. mRS shows significantly lower outpatient setting. score overall (in percentage terms), which could be attrib- uted to the inherent characteristics of the scales. However, The patterns of functional recovery after stroke ob- the rates of improvements in SULCS and mRS appear to served here can provide valuable insight to support the be substantially different. At 48 hours SULCS percentage development and optimization of rehabilitation programs, scores were significantly higher, leading to lower rates of allowing better resource allocation that could improve pa- improvement compared to mRS until 12 weeks. Such find- tient outcomes. Future studies should evaluate functional ings indicate that mRS might not accurately encapsulate up- recovery over longer periods of time in real-word condi- per limb functioning in the initial weeks after acute stroke. tions, in order to further establish the threshold over which The types of assessments used in mRS are more geared to- patients stop experiencing functional gains and the extent wards evaluating walking ability and, therefore, upper limb to which those gains are maintained overtime. functioning is not as accurately measured as with a specific scale. These findings highlight the importance of using a References specific upper limb function assessment scale. Importantly, similarly to the results seen with T-FIM, both SULCS and 1.  Branco JP, Costa JS, Sargento-Freitas J, Oliveira S, Mendes B, Laíns J, mRS showed significant gains until 12 weeks after stroke et al. [Neuroimaging and Blood Biomarkers in Functional Prognosis after and non-significant numerical gains until 24 weeks, which Stroke]. Acta Med Port 2016;29:749–54. Portuguese. is in agreement with previous studies.5, 8 2.  Brito R, Lins L, Almeida C, Ramos-Neto E, Araújo D, Franco C. In- 6 European Journal of Physical and Rehabilitation Medicine February 2019

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, FUNCTIONAL RECOVERY AFTER STROKE BRANCO cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. strumentos de Avaliação Funcional Específicos Para o Acidente Vascular ery after stroke: a systematic review of the literature. Arch Phys Med Re- Cerebral. Rev. Neurociencias 2014;21:593–9. habil 2002;83:1629–37. 3.  Ward A, Payne KA, Caro JJ, Heuschmann PU, Kolominsky-Rabas PL. 14.  Kwakkel G, Kollen B, Lindeman E. Understanding the pattern of Care needs and economic consequences after acute ischemic stroke: the functional recovery after stroke: facts and theories. Restor Neurol Neu- Erlangen Stroke Project. Eur J Neurol 2005;12:264–7. rosci 2004;22:281–99. 4.  Duncan PW, Goldstein LB, Matchar D, Divine GW, Feussner J. Mea- 15.  Aydin T, Taspinar O, Kepekci M, Keskin Y, Erten B, Gunel M, et surement of motor recovery after stroke. Outcome assessment and sample al. Functional independence measure scores of patients with hemiple- size requirements. Stroke 1992;23:1084–9. gia followed up at home and in university hospitals. J Phys Ther Sci 5.  Lee KB, Lim SH, Kim KH, Kim KJ, Kim YR, Chang WN, et al. Six- 2016;28:553–7. month functional recovery of stroke patients: a multi-time-point study. Int 16.  Sangha H, Lipson D, Foley N, Salter K, Bhogal S, Pohani G, et al. A J Rehabil Res 2015;38:173–80. comparison of the Barthel Index and the Functional Independence Mea- 6.  Au-Yeung SS, Hui-Chan CW. Predicting recovery of dextrous hand sure as outcome measures in stroke rehabilitation: patterns of disability function in acute stroke. Disabil Rehabil 2009;31:394–401. scale usage in clinical trials. Int J Rehabil Res 2005;28:135–9. 7.  Higgins J, Mayo NE, Desrosiers J, Salbach NM, Ahmed S. Upper-limb 17.  Branco JP, Oliveira S, Páscoa Pinheiro J, L Ferreira P. Assessing up- function and recovery in the acute phase poststroke. J Rehabil Res Dev per limb function: transcultural adaptation and validation of the Portu- 2005;42:65–76. guese version of the Stroke Upper Limb Capacity Scale. BMC Sports Sci 8.  Verheyden G, Nieuwboer A, De Wit L, Thijs V, Dobbelaere J, Devos Med Rehabil 2017;9:15. H, et al. Time course of trunk, arm, leg, and functional recovery after 18.  Roorda LD, Houwink A, Smits W, Molenaar IW, Geurts AC. Mea- ischemic stroke. Neurorehabil Neural Repair 2008;22:173–9. suring upper limb capacity in poststroke patients: development, fit of 9.  Mellon L, Brewer L, Hall P, Horgan F, Williams D, Hickey A; AS- the monotone homogeneity model, unidimensionality, fit of the double PIRE-S study group. Cognitive impairment six months after ischaemic monotonicity model, differential item functioning, internal consistency, stroke: a profile from the ASPIRE-S study. BMC Neurol 2015;15:31. and feasibility of the stroke upper limb capacity scale, SULCS. Arch Phys 10.  Salvadori E, Pasi M, Poggesi A, Chiti G, Inzitari D, Pantoni L. Pre- Med Rehabil 2011;92:214–27. dictive value of MoCA in the acute phase of stroke on the diagnosis of 19.  Shen J, Leishear K. Novel Techniques for Upper Extremity Train- mid-term cognitive impairment. J Neurol 2013;260:2220–7. ing for Hemiparesis after Stroke; 2011 [Internet]. Available from: https:// 11.  Douiri A, Rudd AG, Wolfe CD. Prevalence of poststroke cogni- www.rehabmedicine.pitt.edu/docs/RGR_PDFs/RGR_10_11.pdf [cited tive impairment: South London Stroke Register 1995-2010. Stroke 2018, Dec 11]. 2013;44:138–45. 20.  Saver JL, Filip B, Hamilton S, Yanes A, Craig S, Cho M, et al.; FAST- 12.  Papp Z, Édes I, Fruhwald S, De Hert SG, Salmenperä M, Leppikan- MAG Investigators and Coordinators. Improving the reliability of stroke gas H, et al. Levosimendan: molecular mechanisms and clinical implica- disability grading in clinical trials and clinical practice: the Rankin Fo- tions: consensus of experts on the mechanisms of action of levosimendan. cused Assessment (RFA). Stroke 2010;41:992–5. Int J Cardiol 2012;159:82–7. 21.  Houwink A, Nijland RH, Geurts AC, Kwakkel G. Functional recov- 13.  Hendricks HT, van Limbeek J, Geurts AC, Zwarts MJ. Motor recov- ery of the paretic upper limb after stroke: who regains hand capacity? Arch Phys Med Rehabil 2013;94:839–44. Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material discussed in the manuscript. Acknowledgements.—The authors thank Graça Fernandes, BSN (Stroke Unit, Centro Hospitalar Universitário de Coimbra) for providing nursing care throughout the study; Marisa Loureiro, MSc (Laboratory of Biostatistics and Medical Informatics, Faculty of Medicine of the University of Coimbra) for providing assistance during statistical analysis; and Tiago Campos, MSc (ARC Publishing) for providing editorial assistance. Article first published online: May 14, 2018. - Manuscript accepted: May 14, 2018. - Manuscript revised: April 16, 2018. - Manuscript received: January 9, 2018. Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 7

COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically © 2018 EDIZIONI MINERVA MEDICA European Journal of Physical and Rehabilitation Medicine 2019 February;55(1):8-18 or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access Online version at http://www.minervamedica.it DOI: 10.23736/S1973-9087.18.05168-7 to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. ORIGINAL ARTICLE How effective is physical therapy for gait muscle activity in hemiparetic patients who receive botulinum toxin injections? Kazuki FUJITA 1, 2, Hiroichi MIAKI 3 *, Hideaki HORI 1, Yasutaka KOBAYASHI 4, Takao NAKAGAWA 3 1Department of Rehabilitation Physical Therapy, Faculty of Health Science, Fukui Health Science University, Fukui-city Fukui, Japan; 2Division of Health Sciences, Graduate School of Medical Sciences, Kanazawa University, Kanazawa-city, Ishikawa, Japan; 3Faculty of Health Sciences, Institute of Medical, Pharmaceutical and Health Sciences, Kanazawa University, Kanazawa-city, Ishikawa, Japan; 4Department of Rehabilitation Medicine, Fukui General Hospital, Fukui-city Fukui, Japan *Corresponding author: Hiroichi Miaki, Faculty of Health Sciences, Institute of Medical, Pharmaceutical and Health Sciences, Kanazawa University, 5-11-80 Kodatsuno, 920-0942, Kanazawa-city Ishikawa, Japan. E-mail: [email protected] ABSTRACT BACKGROUND: Administration of botulinum neurotoxin A (BoNT-A) to the ankle plantar flexors in patients with hemiplegia reduces the strength of knee extension, which may decrease their walking ability. Studies have reported improvements in walking ability with physical therapy following BoNT-A administration. However, no previous studies have evaluated from an exercise physiology perspective the efficacy of physical therapy after BoNT-A administration for adult patients with hemiplegia. AIM: To investigate the effects of physical therapy following BoNT-A administration on gait electromyography for patients with hemiparesis secondary to stroke. DESIGN: Non-randomized controlled trial. SETTING: Single center. POPULATION: Thirty-five patients with chronic stroke with spasticity were assigned to BoNT-A monotherapy (N.=18) or BoNT-A plus physi- cal therapy (PT) (N.=17). METHODS: On the paralyzed side of the body, 300 single doses of BoNT-A were administered intramuscularly to the ankle plantar flexors. Physical therapy was performed for 2 weeks, starting from the day after administration. Gait electromyography was performed and gait param- eters were measured immediately before and 2 weeks after BoNT-A administration. Relative muscle activity, coactivation indices, and walking time/distance were calculated for each phase. RESULTS: For patients who received BoNT-A monotherapy, soleus activity during the loading response decreased 2 weeks after the intervention (P<0.01). For those who received BoNT-A+PT, biceps femoris activity and knee coactivation index during the loading response and tibialis ante- rior activity during the pre-swing phases increased, whereas soleus and rectus femoris activities during the swing phase decreased 2 weeks after the intervention (P<0.05). These rates of change were significantly greater than those for patients who received BoNT-A monotherapy (P<0.05). CONCLUSIONS: Following BoNT-A monotherapy, soleus activity during the stance phase decreased and walking ability either remained un- changed or deteriorated. Following BoNT-A+PT, muscle activity and knee joint stability increased during the stance phase, and abnormal muscle activity during the swing phase was suppressed. CLINICAL REHABILITATION IMPACT: If botulinum treatment of the ankle plantar flexors in stroke patients is targeted to those with low knee extension strength, or if it aims to improve leg swing on the paralyzed side of the body, then physical therapy following BoNT-A adminis- tration could be an essential part of the treatment strategy. (Cite this article as: Fujita K, Miaki H, Hori H, Kobayashi Y, Nakagawa T. How effective is physical therapy for gait muscle activity in hemiparetic patients who receive botulinum toxin injections? Eur J Phys Rehabil Med 2019;55:8-18. DOI: 10.23736/S1973-9087.18.05168-7) Key words: Botulinum toxins - Electromyography - Stroke - Gait. Spasticity is a significant sequela in patients with post- botulinum therapy for the treatment of spasticity, and a stroke hemiplegia, and it occurs in approximately 38% dramatic increase in the use of botulinum neurotoxin type of patients within 12 months of stroke.1 Numerous random- A (BoNT-A) was subsequently noted.2 According to the ized, comparative studies have been conducted regarding guidelines of the American Academy of Neurology and the 8 European Journal of Physical and Rehabilitation Medicine February 2019

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, GAIT ELECTROMYOGRAPHY AFTER BOTULINUM TOXIN ADMINISTRATION FUJITA cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. Japan Stroke Society, botulinum therapy for leg spasticity main uncertain. No previous study has investigated from not only alleviates spasticity but also promotes improve- the point of view of exercise physiology the efficacy of ments in walking and other active functions.3, 4 Further- physical therapy following BoNT-A administration for more, a meta-analysis of studies that investigated the ef- adult patients with hemiplegia. fects of botulinum neurotoxin type A (BoNT-A) treatment for equivarus deformity of the foot due to leg spasticity This study aimed to investigate the effects of BoNT-A reported improvements in walking speed.5 However, the administration to the ankle plantar flexors combined with meta-analysis included numerous studies in which physi- physical therapy on gait electromyography for adult stroke cal therapy was performed following BoNT-A administra- patients. We hypothesized that this combination therapy tion. Conversely, numerous studies reported that walking would increase the activities of muscles other than the speed did not increase following botulinum administration ankle plantar flexors. to the ankle plantar flexors in patients with chronic phase stroke.6-8 Furthermore, a previous study in which the only Materials and methods intervention was botulinum administration to the ankle plantar flexors reported no changes in the spatiotemporal Subjects parameters of walking.9 Fifty-eight patients with chronic stroke who received A meta-analysis that assessed the safety of BoNT-A botulinum toxin A (BOTOX®; Allergan Pharmaceuticals, therapy indicated that the most frequent adverse event was Dublin, Ireland) for lower limb spasticity at Fukui General localized muscle weakness.10 Studies have also reported Hospital between September 2014 and October 2016 were increases in the frequency of falls associated with muscle recruited for this study. The inclusion criteria were as fol- weakness following BoNT-A administration to the ankle lows: 1) unilateral cerebral lesions; 2) at least 6 months plantar flexors.11 In a previous study, some patients report- since the onset of stroke; 3) walking ability of monitor- ed deterioration in their walking ability following BoNT-A ing level or higher (can walk either without help or with administration to the ankle plantar flexors.9 Ankle plan- a T-shaped cane, without leg braces); 4) walking speed of tar flexors contribute to stability of knee extension during 0.1-1 m/s; and 5) spasticity of at least 1+ on the modified the stance phase;12 therefore, decreased walking ability Ashworth scale (MAS) in the ankle plantar flexors. following BoNT-A administration is possible. Deteriora- tions in gait may also be secondary to the physiological The exclusion criteria were as follows: 1) passive ankle mechanism of action of BoNT-A rather than muscle weak- dorsiflexion range of motion (ROM) ≤0; 2) higher cog- ness. When BoNT-A is administered to skeletal muscles, it nitive dysfunction that can hinder the intervention and/or enters the motor neurons in the spinal cord by retrograde evaluation; 3) cardiovascular diseases that restrict exer- axonal transport, thereby affecting Renshaw cells, which cise; and 4) BoNT-A administration within the previous control recurrent inhibition.13 For patients with hemiple- 4 months. gia, the motor neurons of the quadriceps femoris undergo excitatory synaptic stimulation via the soleus recurrent in- We presented two treatment strategies to 35 patients hibition pathway, resulting in involuntary coactivation.14 who were finally included, BoNT-A administration alone Therefore, following BoNT-A administration to the soleus, (BoNT-A monotherapy) and intensive physical therapy the activity of Renshaw cells is suppressed and the H-re- following BoNT-A administration (BoNT-A+PT). Pa- flex of vastus lateralis is reduced.15 As a result of these tients were asked to choose their treatment; subsequently, mechanisms, BoNT-A administration to the ankle plantar they were administered BoNT-A monotherapy (N.=18) or flexors in patients with hemiplegia reduces knee extension BoNT-A+PT (N.=17) (Figure 1). strength, which in turn increases the risk of falls. All measurements and interventions were performed at Studies of physical therapy including leg resistance Fukui General Hospital. This study was approved by the exercises, walking training, functional electrical stimu- Ethical Review Committee of the hospital (approval no.: lation of the ankle dorsiflexors, and biofeedback follow- Nittazuka Ethics 26-4). All subjects provided written in- ing BoNT-A administration have reported improvements formed consent. in walking ability.16-19 However, the improvements were based on simple parameters such as walking speed; there- BoNT-A administration fore, the mechanisms underlying these improvements re- The target muscles for BoNT-A administration were the medial head of the gastrocnemius, lateral head of the gas- trocnemius, soleus, tibialis posterior, flexor digitorum lon- Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 9

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically FUJITA COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, GAIT ELECTROMYOGRAPHY AFTER BOTULINUM TOXIN ADMINISTRATION cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. Assessed for eligibility (N.=58) Spare walkaway Measurement walkaway Spare walkaway 3m 10 m 3m Excluded (N.=23) - Did not meet inclusion criteria Calibration line Stride length (Right) (1 m) Step length (N.=23) (Right) - Declined to participate (N.=0) Step length - O ther reasons (N.=0) (Left) Stride length (Left) 5m Included (N.=35) Lateral view videocamera Botulinum neurotoxin A (BoNT-A) Botulinum neurotoxin A (BoNT-A) Figure 2.—Layout of the walkway used for measuring electromyogra- monotherapy monotherapy + physiotherapy phy and spatiotemporal parameters during gait. Allocated to intervention (N.=18) Allocated to intervention (N.=17) Evaluation - Received allocated intervention - Received allocated intervention The evaluations were performed immediately before and 2 (N.=18) (N.=17) weeks after BoNT-A administration by a single examiner. - Did not receive allocated interven- - Did not receive allocated interven- The primary outcome assessment included electromyog- raphy and determination of spatiotemporal parameters tion (N.=0) tion (N.=0) during gait; the secondary outcome assessment included determination of the spasticity of ankle plantar flexors and Lost to follow-up (N.=0) Lost to follow-up (N.=0) patient satisfaction. Discontinued intervention Discontinued intervention Electromyography was performed and spatiotemporal (other disease) (N.=1) (N.=0) parameters were evaluated at a comfortable walking speed and mode. A straight walkway (16 m) was prepared for Analyzed (N.=17) Analyzed (N.=18) measuring the walking distance (including an extra 3 m - E xcluded from analysis (N.=0) - Excluded from analysis (N.=0) at each end). A video camera (HD Pro WebCamera; Logi- cool, Inc., Tokyo, Japan) with a sampling frequency of 30 Figure 1.—Flow diagram of the study design and the processes. Hz was set at 5 m lateral to the midpoint of the walkway, and a 1-m line was drawn at the midpoint of the walkway gus, and flexor pollicis longus. The administration site was (Figure 2). The time taken by each patient to walk 10 m selected according to each patient’s condition. A total of was measured using a stopwatch. During the measure- 300 units were administered per muscle, with a minimum ments, the subjects wore shoes and were allowed to use dose of 50 units. Ultrasound was used to monitor the posi- walking canes, but not leg braces. tions of the needles and muscles during administration to the deep muscles. TELEmyo DTS (Noraxon Inc.) was used for electro- myographic recording. The sampling frequency was 1500 Physical therapy intervention Hz and the bandpass filter was set at 10-500 Hz. Electro- myography was assessed in the following five muscles on Physical therapy was performed for 2 weeks (two 1-hour the paralyzed side of the body: tibialis anterior, soleus, sessions per day), starting from the day after BoNT-A ad- medial head of the gastrocnemius, rectus femoris, and bi- ministration. All patients were randomly treated by three ceps femoris. Muscle action potential was induced using therapists who were blinded to the purpose of this study. bipolar leads. Skin impedance was reduced to no more The same physical therapy program was followed for each than 10 kΩ using alcohol-soaked cotton swabs and Abra- patient. It included the following: 1) stretching of the ankle sive (Skin Pure; Nihon Kohden Co., Ltd., Tokyo, Japan). plantar flexors; 2) leg resistance exercises; 3) low-frequen- Ag-AgCl electrodes (EM-272; Noraxon Inc.) were posi- cy electrical stimulation (PAS System GD-601; OG Well- tioned 2 cm apart and were placed on positions recom- ness Technologies Co., Ltd., Okayama, Japan) of the ankle mended by the SENIAM (Surface ElectroMyoGraphy for dorsiflexors; 4) electromyographic feedback (MyoTrace; the Non-Invasive Assessment of Muscle) project.20 Foot Noraxon Inc., Scottsdale, AZ, USA) for ankle dorsiflexion switches were placed on the soles of both feet (4 points exercises; and 5) walking exercises, including walking on a level surface and treadmill with body weight support (Un- weighing System; Biodex Medical Systems Inc., Shirley, NY, USA). 10 European Journal of Physical and Rehabilitation Medicine February 2019

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, GAIT ELECTROMYOGRAPHY AFTER BOTULINUM TOXIN ADMINISTRATION FUJITA cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. for each foot). Time axes of all devices were matched us- walking time (measured using a stopwatch), and cadence ing synchronization and optical signals. and gait cycle duration, which were calculated from the foot switch data. The asymmetry index for the swing Spasticity of the ankle plantar flexors was evaluated phase duration was calculated as follows:25 based on the MAS and clonus scores in the supine posi- tion. Scores of 1+ on the MAS were assigned as 2, and Asymmetry index = swing phase duration on the scores of 2 and higher were revised upward by 1. The paralyzed side / (swing phase duration on the paralyzed clonus score was recorded in five stages according to the side + swing phase duration on the non-paralyzed side) duration of the ankle clonus (0, no clonus can be induced; 1, clonus lasting for 1-4 seconds; 2, 5-9 seconds; 3, 9-15 Spatial gait parameters were measured using the still seconds; 4, >15 seconds.21 Additionally, ROM of passive image extracted from the video data when the subject and active ankle dorsiflexion were measured using a goni- passed through the intermediate points on the walk- ometer with the patient in the supine position and the knee way.26, 27 The image processing software ImageJ (Na- joint flexed to 90°. tional Institutes of Health, Bethesda, MD, USA) was used. Stride length was measured as the linear distance Patient satisfaction was evaluated 2 weeks after BoNT- between successive points of heel contact of the same A administration using the global rating of change scale foot. Step length was measured as the linear distance be- (GRCS).22 GRCS uses a questionnaire with the following tween corresponding successive points of heel contact of question: “Has your walking changed in comparison with opposite feet. The asymmetry index for step length was before treatment?,” and the patient can choose an answer calculated as follows:28 from 15 possible scores (-7 to 7). Asymmetry index = step length on the paralyzed side / Data analysis (step length on the paralyzed side + step length on the MR3 (Noraxon Inc.) was used for analyzing the electro- non-paralyzed side) myographic waveforms. First, full-wave rectification of all raw waveforms was performed. The analysis interval was For all electromyographic data and gait parameters, set to three continuous gait cycles around the middle part the means obtained from three walking trials were calcu- of the walkway. Initial contact was defined as the time of lated. Therefore, for electromyographic data and tempo- electric potential input from the foot switch on the para- ral parameters, the means calculated were derived from lyzed side of the body and was determined using synchro- nine gait cycles (three gait cycles × three trials); for spa- nized video cameras, with monitoring for abnormal electric tial parameters, the means calculated were derived from potentials from the foot switch due to leg-dragging during three gait cycles (one gait cycle × three trials). the swing phase. The duration of each of the three gait cy- cles was normalized after considering one gait cycle to be Statistical analysis 100%. The arithmetic mean of the three gait cycles was obtained, and 1000-point amplitudes were calculated at in- For all the data evaluated, the differences in each treatment tervals of 0.1%, followed by normalization using the mean group before and after the intervention were compared us- amplitude of the entire gait cycle. ing the paired t-test or Wilcoxon signed-rank test depend- ing on the normality and scale of the data. In addition, the According to the report by Turns et al.,23 muscle ac- effect size (ES; ES=r value) and the 95% confidence inter- tivities in each of the gait phases were distinguished vals (95% CI) were calculated.29 Differences between the based on the data from the foot switch, i.e., the loading treatment groups with respect to preintervention data and response, single support, pre-swing, and swing phases proportional changes before and after intervention ([post were calculated from the mean amplitudes of the respec- - pre] / pre) were evaluated using the t-test, Welch Test, tive phases (Figure 3). Additionally, according to Chow and Mann-Whitney U Test. The software used for these et al.,24 coactivation indices were calculated for the tibi- analyses was SPSS version 20 (IBM Co., Ltd., Armonk, alis anterior and gastrocnemius; they were calculated for NY, USA), with a significance level of 5%. the rectus femoris and biceps femoris by dividing the ar- eas of overlaps between the flexor and extensor muscle Results amplitudes for each gait phase by the duration of the relevant gait phase (Figure 4). The temporal gait param- One patient who was receiving BoNT-A monotherapy dis- eters were walking speed, which was calculated from the continued therapy due to other medical reasons; therefore, a total of 17 patients received BoNT-A monotherapy (13 Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 11

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically FUJITA COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, GAIT ELECTROMYOGRAPHY AFTER BOTULINUM TOXIN ADMINISTRATION cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. males and 4 females; mean age, 57.2 years; mean months Table I.—Baseline characteristics of patients in each group. since stroke onset, 75.2 months) and 17 received BoNT- A+PT (12 males and 5 females; mean age, 58.6 years; mean BoNT-A BoNT-A+PT months since stroke onset, 39.8 months) (Figure 1, Table I). monotherapy (N.=17) The total volume of BoNT-A administered to each muscle in both groups was approximately the same (Table I). (N.=17) Baseline Age 57.2±10.7 58.6±10.5 Sex (female/male) 4/13 5/12 No significant difference between groups was observed Type of stroke (CI/ICH/SAH) 6/10/1 6/10/1 for age (P=0.523), but there was a significant differ- Months since onset ence between groups for mean months since stroke on- Paretic side (L/R) 75.2±51.2 39.8±37.7 set (P=0.020) (Table I). No significant differences in the Fugl-Meyer assessment LE 6/11 7/10 baseline of all electromyogram data for tibialis anterior Assistive device (none/t-cane/AFO) 21±3 19±6 (loading response: P=0.291; single support: P=0.667; Total amount of BoNT-A administered pre-swing: P=0.341; swing: P=0.713), soleus (P=0.300, 3/11/13 2/13/14 P=0.437, P=0.262, and P=0.113, respectively), medial MG gastrocnemius (P=0.143, P=0.223, P=0.069, and P=0.052, Sol 1025 U 1025 U FDL 1075 U 1050 U LG 500 U 450 U TP 1025 U 1025 U 1075 U 1050 U FHL 400 U 450 U BoNT-A: botulinum neurotoxin A; PT: physical therapy; CI: cerebral infarction; ICH: intracerebral hemorrhage; SAH: subarachnoid hemorrhage; LE: lower extremity; AFO: ankle foot orthosis; MG: medialis gastrocnemius; LG: lateralis gastrocnemius; Sol: soleus; TP: tibialis posterior; FDL: flexor digitorum longus; FHL: flexor pollicis longus. Figure 3.—Calculation of the mean am- Loading Single Pre-swing Swing plitude for each gait phase. response support The electromyographic amplitude in soleus, with normalized waveform, is Normalized EMG (%) shown. Broken lines indicate the cal- culated mean amplitudes in each gait phase. Figure 4.—Calculation of the coactiva- Loading Single Gait cycle (%) Swing tion indices for each gait phase. response support Pre-swing The electromyographic amplitudes with normalized waveforms for the extensor Normalized EMG (%) muscles (broken lines) and flexor mus- cles (unbroken line). The areas of the parts of the two waveforms that overlap in each gait phase (gray: S) were cal- culated, and the co-activation index for each phase was calculated by dividing the respective area by the duration of each gait phase (t). Gait cycle (%) CoI of loading response = S/t 12 European Journal of Physical and Rehabilitation Medicine February 2019

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, GAIT ELECTROMYOGRAPHY AFTER BOTULINUM TOXIN ADMINISTRATION FUJITA cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. Table II.—Comparisons of electromyogram data between BoNT-A monotherapy and BoNT-A+PT groups. Muscles Gait phase BoNT-A monotherapy (N.=17) BoNT-A+PT (N.=17) P value Pre Post Change rate Pre Post Change rate TA LR 95.9±24.6 104.3±28.8 9.8±20.5 107.8±38.5 99.3±33.3 -5.2±24.9 NS SS 67.7±17.8 64.0±26.9 -6.2±31.0 69.6±24.7 58.9±24.4* -10.6±32.3 NS PSw 119.9±29.9 119.6±36.4 -0.7±14.0 110.1±29.4 124.7±40.0* 14.4±22.9 <0.05 Sw 116.6±26.2 112.1±24.2 -1.4±19.8 112.6±35.9 117.1±23.5 13.7±41.1 NS Sol LR 154.8±35.8 140.7±30.3** -8.0±12.2 142.9±30.0 142.7±29.6 1.4±19.0 NS SS 138.5±28.3 134.9±20.9 -1.0±13.6 130.3±32.5 130.3±30.2 2.7±20.6 NS PSw 62.4±18.1 75.3±22.0** 23.3±33.8 70.6±23.7 78.8±25.1 16.1±36.1 NS Sw 44.2±17.5 49.1±14.1 19.0±32.2 56.2±24.5 48.1±18.5* -8.2±31.7 <0.05 MG LR 153.4±31.4 148.5±33.9 -3.0±11.9 133.9±23.8 136.1±22.1 3.7±21.2 NS SS 123.8±18.4 127.1±16.7 3.4±11.5 114.1±26.6 120.8±21.0 8.5±18.2 NS PSw 62.5±19.4 65.9±20.2 7.7±23.6 74.5±17.9 76.8±15.1 7.0±28.1 NS Sw 60.3±19.2 58.5±19.7 -0.4±22.9 77.5±24.9 66.3±21.2* -11.2±23.3 NS RF LR 151.2±27.9 157.9±31.8 6.3±21.5 135.1±43.5 149.6±41.4 7.4±42.3 NS SS 92.6±30.4 93.5±27.6 3.7±20.5 85.2±24.4 85.9±22.4 3.7±20.5 NS PSw 86.0±33.5 76.5±27.8 -5.4±37.0 96.2±36.8 90.7±36.3 -1.3±35.2 NS Sw 70.2±26.1 72.1±23.5 6.8±27.9 83.5±26.9 73.7±21.1* -9.8±15.5 <0.05 BF LR 174.0±32.8 169.5±42.9 -3.1±11.9 153.3±35.6 176.0±31.2* 25.6±61.0 <0.05 SS 125.9±25.2 126.0±30.1 -0.4±8.8 122.0±37.2 109.7±35.2 -9.0±20.0 NS PSw 39.1±21.6 42.6±26.6 13.3±53.0 55.0±38.0 43.3±18.0 -5.4±45.5 NS Sw 61.0±26.6 61.8±29.7 3.4±24.7 69.7±28.0 71.0±30.7 2.8±24.2 NS Values are normalized mean amplitudes (%) during each gait phase (mean±standard deviation). Change rate = (post data - pre data) / pre data × 100. P value: comparison of the change rate between BoNT-A monotherapy and BoNT-A+PT groups (t-test or Welch’s t-test or Mann-Whitney U Test). BoNT-A: botulinum neurotoxin A; PT: physical therapy; TA: tibialis anterior; Sol: soleus; MG: medial gastrocnemius; RF: rectus femoris; BF: biceps femoris; LR: loading response; SS: single support; PSw: pre-swing; Sw: swing; NS: no significant difference. *P<0.05, **P<0.01 (pre vs. post, paired t-test or Wilcoxon signed-rank test). Table III.—C omparisons of the coactivation indices between BoNT-A monotherapy and BoNT-A+PT groups. Muscles Gait phase BoNT-A monotherapy (N.=17) BoNT-A+PT (N.=17) P value Pre Post Change rate Pre Post Change rate TA-MG LR 74.1±19.0 75.2±20.7 2.2±18.2 76.0±27.6 74.5±26.3 1.2±23.9 NS RF-BF SS 56.6±13.1 56.0±20.0 -0.8±30.8 54.5±18.4 51.9±19.5 -0.8±33.2 NS PSw 46.3±15.3 48.7±13.9 8.3±20.0 53.7±14.7 54.8±15.0 8.4±51.9 NS Sw 48.2±12.1 48.4±10.7 3.5±20.5 55.4±14.2 52.1±13.7 -4.2±19.8 NS LR 117.5±26.0 120.5±31.5 3.7±19.5 99.9±27.4 122.0±34.1* 32.3±65.8 <0.05 SS 76.8±25.1 77.1±24.3 2.2±15.1 68.0±20.0 71.8±22.2 7.4±23.4 NS PSw 31.6±14.2 32.4±17.9 3.4±27.1 41.9±25.3 35.0±10.4 0.8±54.2 NS Sw 38.5±13.4 38.5±13.5 2.5±20.2 45.6±13.3 42.3±14.2 -7.0±16.1 NS Values are coactivation indices (%) during each gait phase (mean±standard deviation). Change rate = (post data - pre data) / pre data × 100. P value: comparison of the change rate between BoNT-A monotherapy and BoNT-A+PT groups (t-test or Welch’s t-test or Mann-Whitney U Test). BoNT-A: botulinum neurotoxin A; PT: physical therapy; TA: tibialis anterior; Sol: soleus; MG: medial gastrocnemius; RF: rectus femoris; BF: biceps femoris; LR: loading response; SS: single support; PSw: pre-swing; Sw: swing; NS: no significant difference. *P<0.01 (pre vs. post, paired t-test or Wilcoxon signed-rank test). respectively), rectus femoris (P=0.163, P=0.438, P=0.278, (Table II, III). No significant differences in the baseline of and P=0.152, respectively), biceps femoris (P=0.091, all spatiotemporal parameters of gait velocity (P=0.098), P=0.729, P=0.147, and P=0.365, respectively), the co- cadence (P=0.999), single gait cycle period (P=0.835), activation index of the tibialis anterior and gastrocnemius loading response period (P=0.221), single support period (P=0.813, P=0.705, P=0.162, and P=0.117, respectively), (P=0.161), pre-swing period (P=0.174), double-support and the coactivation index of the rectus femoris and bi- period (P=0.134), swing period (P=0.736), swing pe- ceps femoris (P=0.069, P=0.277, P=0.150, and P=0.135, riod asymmetry index (P=0.312), stride length (P=0.053) respectively) between the treatment groups were found and step length asymmetry index (P=0.061) between Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 13

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically FUJITA COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, GAIT ELECTROMYOGRAPHY AFTER BOTULINUM TOXIN ADMINISTRATION cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. Table IV.—C omparisons of gait parameters between BoNT-A monotherapy and BoNT-A+PT groups. Assessment BoNT-A monotherapy (N.=17) BoNT-A+PT (N.=17) Pre Post Change rate P value Pre Post Change rate Gait velocity (m/s) 0.57±0.20 0.59±0.18 5.5±17.0 0.46±0.20 0.54±0.16** 21.7±16.0 <0.01 Cadence (steps/min) 81.3±18.0 81.7±15.7 1.3±7.9 81.3±20.7 87.2±17.3** 9.1±10.8 <0.05 1 stride period 1.55±0.35 1.52±0.31 -0.8±7.1 1.58±0.43 1.43±0.29** -7.5±8.7 <0.05 LR period (s) 0.25±0.16 0.23±0.13 -0.9±16.5 0.28±0.13 0.23±0.09† <0.01 SS period (s) 0.41±0.09 0.41±0.06 2.0±10.2 0.37±0.09 0.36±0.06 -17.5±10.4 PSw period (s) 0.28±0.13 0.28±0.12 2.2±14.3 0.34±0.15 0.29±0.11* 0.2±12.8 NS DS period (s) 0.53±0.28 0.51±0.23 0.6±14.2 0.62±0.24 0.52±0.17† -9.9±15.9 <0.05 Sw period (s) 0.61±0.14 0.60±0.11 -1.0±8.3 0.59±0.17 0.55±0.11* <0.01 AI-Sw period (s) 0.60±0.04 0.59±0.04 -1.0±4.0 0.61±0.05 0.60±0.04 -13.8±11.2 Stride length 83.4±19.6 83.6±16.6 1.6±10.1 69.6±20.5 76.6±18.4† -5.2±10.1 NS AI-step length (cm) 0.51±0.07 0.52±0.06 0.6±6.3 0.57±0.10 0.53±0.06** -1.8±4.3 NS 12.2±12.1 <0.01 -5.7±6.4 <0.01 Values represent the mean±standard deviation. *P<0.05, **P<0.01, †P<0.001 (pre vs. post, paired t-test or Wilcoxon signed-rank Test). Change rate = (post data - pre data) / pre data × 100. P value = comparison of the change rate between BoNT-A monotherapy and BoNT-A+PT groups (t-test or Welch’s t-test or Mann-Whitney U Test). BoNT-A: botulinum neurotoxin A; PT: physical therapy; LR: loading response; SS: single support; PSw: pre-swing; Sw: swing; DS: double support; AI: asymmetry index; NS: no significant difference. Table V.—C omparisons of ankle joint data between BoNT-A monotherapy and BoNT-A+PT groups. Assessment BoNT-A monotherapy (N.=17) BoNT-A+PT (N.=17) P value Pre Post Change Pre Post Change NS Modified Ashworth Scale 3 (0) 2 (0)† -1 (0) 3 (0) 2 (0)** -1 (0) P<0.05 Clonus Score 1 (1) 0 (1)* 0 (-1) 1 (4) 0 (1)** -1 (3) P<0.05 ROM-passive df 13.6±6.1 16.7±6.2** 3.1±3.4 13.9±6.9 20.8±6.0† 6.8±5.5 ROM-active df -9.9±14.8 -4.5±14.1** 5.4±5.6 -6.4±16.2 -1.4±11.7** 5.0±7.0 NS Values appear as mean±standard deviation or as median (quartile deviation). ROM is expressed in grades. Modified Ashworth Scale: score of 1+ was assigned as 2, and scores of 2 and higher were revised upward by 1. *P<0.05; **P<0.01; †P<0.001 (pre vs. post, paired t-test or Wilcoxon signed-rank Test). Change = (post data - pre data). P value: comparison of the change between BoNT-A monotherapy and BoNT-A+PT groups (t-test or Welch’s t-test or Mann-Whitney U Test). BoNT-A: botulinum neurotoxin A; PT: physical therapy; ROM: range of motion; df: dorsiflexion; NS: no significant difference. the treatment groups were found (Table IV). No signifi- ES=0.497; 95% CI: -0.282 to -0.011), whereas those in cant differences in the baseline of all ankle joint data for the soleus (P=0.047; ES=0.473; 95% CI: 0.001-0.160) and MAS (P=0.843), clonus score (P=0.127), passive ROM rectus femoris (P=0.022; ES=0.536; 95% CI: 0.016-0.179) (P=0.875), and active ROM (P=0.511) between treatment decreased significantly during the swing phase, and that groups were found (Table V). in biceps femoris increased significantly during the load- ing response phase (P=0.033; ES=0.506; 95% CI: -0.434 Muscle activity during each gait phase to -0.021) after the intervention. These rates of change were significantly greater than those for patients who re- Of the patients who received BoNT-A monotherapy, the ceived BoNT-A monotherapy (tibialis anterior: P=0.027; electromyographic amplitude in the soleus after the inter- soleus: P=0.019; rectus femoris: P=0.040; biceps femoris: vention decreased significantly during the loading response P=0.013) (Table II). phase (P=0.005; ES=0.628; 95% CI: 0.049-0.235) and in- creased significantly during the pre-swing phase (P=0.002; Furthermore, of those who received BoNT-A+PT, the ES=0.679, 95% CI: -0.203 to -0.055) (Table II). No other coactivation index of the rectus femoris and biceps femo- significant differences in electromyographic amplitudes ris after the intervention significantly increased during the were found after the intervention for these patients. loading response phase (P=0.006; ES=0.686; Z=2.741), and the rate of change was significantly greater than that Of the patients who received BoNT-A+PT, the electro- for those who received BoNT-A monotherapy (P=0.048) myographic amplitude in the tibialis anterior increased (Table III). significantly during the pre-swing phase (P=0.035; 14 European Journal of Physical and Rehabilitation Medicine February 2019

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, GAIT ELECTROMYOGRAPHY AFTER BOTULINUM TOXIN ADMINISTRATION FUJITA cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. Spatiotemporal gait parameters greater for those who received BoNT-A+PT than for those who received BoNT-A monotherapy (P=0.024 and For patients who received BoNT-A monotherapy, there P=0.021, respectively) (Table V). were no significant differences after the intervention. For those who received BoNT-A+PT, gait velocity (P=0.002; Patient satisfaction: GRCS ES=0.764; Z=3.148) and cadence (P=0.003; ES=0.663; 95% CI: -9.333 to -2.341) significantly increased after the After the interventions, GRCS was higher for those who intervention, and the rates of change for both parameters received BoNT-A+PT than for those who received BoNT- were significantly greater than those for patients who re- A monotherapy (BoNT-A monotherapy: median, 2; maxi- ceived BoNT-A monotherapy (gait velocity: P=0.002; ca- mum, 4; minimum, -2; BoNT-A+PT: median, 3; maxi- dence: P=0.023) (Table IV). Seven patients who received mum, 6; minimum, 0). Two subjects who received BoNT- BoNT-A monotherapy and one who received BoNT- A monotherapy reported deterioration in walking, whereas A+PT demonstrated decreased walking speed after the none who received BoNT-A+PT did. intervention. Discussion For patients who received BoNT-A+PT, the durations of the single gait cycle (P=0.005; ES=0.635; 95% CI: 0.052- In this study, improvements in MAS score, clonus score, 0.241), loading response phase (P<0.001; ES=0.856; and ROM for both treatment groups were evaluated, and Z=3.527), pre-swing phase (P=0.019; ES=0.569; BoNT-A was found to be effective for inhibiting spasticity. Z=2.343), and double support phase (P<0.001; ES=0.810; There was a significant difference in the mean number of Z=3.337) significantly decreased after the intervention; months since stroke onset between baseline groups, but it these rates of change were significantly greater than those was considered that the results were not affected because of patients who received BoNT-A monotherapy (loading both groups were in the chronic phase. With respect to response: P=0.001; pre-swing: P=0.026; double support: gait evaluation, for patients who received BoNT-A mono- P=0.003) (Table IV). therapy, changes in muscle activity were found only in the soleus, and no changes in the spatiotemporal parameters For patients who received BoNT-A+PT, stride length were observed. Additionally, approximately 40% of the increased significantly (P<0.001; ES=0.724; 95% CI: patients demonstrated decreased walking speed and some -10.480 to -3.442) and step length asymmetry index de- reported gait deterioration. For those who received BoNT- creased significantly (P=0.004; ES=0.638; 95% CI = A+PT, changes in muscle activity occurred in numerous 0.013-0.061) after the intervention, with the rate of change muscles and changes in spatiotemporal gait parameters in both parameters being significantly greater than those were observed; almost all subjects showed improvements of patients who received BoNT-A monotherapy (stride in walking speed. Improvements in the spatiotemporal gait length: P=0.003; step length asymmetry index: P=0.007) parameters in our study were similar to those reported in (Table IV). previous studies in which subjects with approximately the same severity of spasticity as that in our study patients un- Muscle tone and ROM derwent walking training after BoNT-A administration;16, 30 however, the proportional change was markedly greater in For both treatment groups, MAS scores (BoNT-A mono- our study, and the increase in walking speed reached 0.06 therapy: P<0.001; ES=0.827; Z=3.408, BoNT-A+PT: m/s,31 which is the minimal difference for clinically sig- P=0.002; ES=0.772; Z=3.180) and clonus scores (BoNT- nificance. The duration of physical therapy in this study A monotherapy: P=0.043; ES=0.491; Z=2.023, BoNT- was shorter than that in previous studies;16-19 nevertheless, A+PT: P=0.003; ES=0.712; Z=2.934) after the inter- physical therapy was performed for a longer duration each vention significantly decreased, whereas passive ROM day. Furthermore, intensive physical therapy immediately (BoNT-A monotherapy: P=0.002; ES=0.683; 95% CI: after BoNT-A administration could have a greater effect on -4.888 to -1.347, BoNT-A+PT: P<0.001; ES=0.790; 95% improving gait parameters than long-term physical therapy CI: -9.635 to -4.012) and active ROM (BoNT-A monother- such as home exercises or that during hospital visits. apy: P=0.001; ES=0.708; 95% CI: -8.275 to -2.549, BoNT- A+PT: P=0.010; ES=0.592; 95% CI: -8.613 to -1.387) af- Electromyographic evaluation of gait showed that so- ter the intervention significantly increased compared with leus activity during the loading response phase decreased those before the intervention. Changes in clonus score and following BoNT-A+PT, which is consistent with the find- passive ROM due to the intervention were significantly Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 15

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically FUJITA COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, GAIT ELECTROMYOGRAPHY AFTER BOTULINUM TOXIN ADMINISTRATION cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. ings of Hesse et al.32 BoNT-A acts at the terminals of ing BoNT-A administration. In our study, rectus femoris γ-motor neurons and reduces the excitation of type Ia af- activity decreased during the swing phase for patients who ferent neurons from muscle spindles;33, 34 therefore, sup- received BoNT-A+PT, suggesting that the required in- pression of the stretch reflex during the initial contact time crease in knee flexion angle during the swing phase could with equine gait is highly probable. be achieved without increasing the risk of decreased knee stability during the stance phase. For patients who received BoNT-A+PT, despite re- ceiving the same dose of BoNT-A as those who received Because soleus activity during the stance phase was de- BoNT-A monotherapy, no decrease in soleus activity was creased in patients who received BoNT-A monotherapy, observed after the intervention. Additionally, increases it is possible that there were decreases in knee extension were noted in biceps femoris activity as well as the co- strength, which could result in gait deterioration. Physical activation index of the rectus femoris and biceps femo- therapy following BoNT-A administration could increase ris. These findings, which were considered to be due to the muscle activity associated with knee joint stability, and physical therapy, were similar to those reported by Mulroy it possibly has positive effects on muscle activities associ- et al.35 The vertical force component of the floor reaction ated with leg swinging. If botulinum treatment of the ankle force is highest during the loading response phase; there- plantar flexors in stroke patients is targeted to those with fore, the stability of the knee joint during this phase is es- low knee extension strength or aims to improve leg swing sential.36 The increase in knee coactivation compensates on the paralyzed side of the body, then physical therapy for the decreased knee joint stability.37 Therefore, an in- following BoNT-A administration could be an essential crease in knee coactivation without a decrease in soleus part of the treatment strategy. Additionally, intensive phys- activity, which contributes to knee stability,12 could reduce ical therapy performed immediately after BoNT-A admin- the risk of the knee giving way, which is associated with istration is considered effective. the adverse effects of BoNT-A. Limitations of the study For patients who received BoNT-A+PT, tibialis anterior activity increased during the pre-swing phase and soleus BoNT-A was administered to the deep muscles such as the activity decreased during the swing phase after the inter- tibialis posterior and flexor digitorum longus in some of vention. Soleus activity during the pre-swing phase in- the patients in this study, and the subsequent effects could creased after BoNT-A monotherapy, whereas no increase not be confirmed by surface electromyography. Addition- in soleus activity was observed following BoNT-A+PT. ally, the effects of the changes in muscle activity on pa- Tang et al.38 previously suggested that increased tibialis rameters of kinematics are unknown. Because physical anterior activity after BoNT-A administration is due to the therapy included multiple components such as resistance decreased excitation of type Ia centripetal neurons from exercises, electrical stimulation, and walking training, the the ankle plantar flexors, thereby reducing the reciprocal specific treatment that was responsible for the effects is inhibition of the tibialis anterior. Therefore, the increase in unclear. Most patients in this study had moderately severe tibialis anterior activity in this study could have been due spasticity and movement disorders; therefore, the results to reduced reciprocal inhibition. Furthermore, increased would not necessarily be similar for stroke patients with tibialis anterior activity during the pre-swing phase may mild or severe disabilities. have contributed to the decrease in soleus activity during the swing phase. Conclusions Decreased knee flexion angle during the swing phase In this study, marked changes in muscle activity were ob- due to hyperactivation of the rectus femoris is a gait abnor- served in stroke patients who underwent physical therapy mality that generally occurs in stroke patients.39, 40 BoNT- after BoNT-A administration to the ankle plantar flexors. A is often administered to the rectus femoris to overcome These changes included increased biceps femoris activ- this problem.39, 40 However, decreased muscle activity dur- ity and coactivation index of the rectus femoris and bi- ing the stance phase after BoNT-A administration to the ceps femoris during the loading response phase; further- rectus femoris has been reported.41 BoNT-A administra- more, they included increased tibialis anterior activity tion to the soleus acts on Renshaw cells at the spinal cord and decreased soleus and rectus femoris activities in the segment level, which reduces the activity of the quadriceps pre-swing and swing phases. Almost all patients in both femoris.15 Hence, there are concerns regarding potential treatment groups demonstrated improved walking speed, decreased knee stability during the stance phase follow- 16 European Journal of Physical and Rehabilitation Medicine February 2019

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, GAIT ELECTROMYOGRAPHY AFTER BOTULINUM TOXIN ADMINISTRATION FUJITA cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. and none reported gait deterioration. Physical therapy af- 17.  Picelli A, Bacciga M, Melotti C, La Marchina E, Verzini E, Ferrari ter BoNT-A administration increased the stability and leg F, et al. Combined effects of robot‑assisted gait training and botulinum swing capability of the knee. Therefore, it can possibly toxin type A on spastic equinus foot in patients with chronic stroke: a prevent deterioration in walking ability. pilot, single blind, randomized controlled trial. Eur J Phys Rehabil Med 2016;52:759–66. References 18.  Johnson CA, Burridge JH, Strike PW, Wood DE, Swain ID. The ef- fect of combined use of botulinum toxin type A and functional electric 1.  Watkins CL, Leathley MJ, Gregson JM, Moore AP, Smith TL, Sharma stimulation in the treatment of spastic drop foot after stroke: a preliminary AK. Prevalence of spasticity post stroke. Clin Rehabil 2002;16:515–22. investigation. Arch Phys Med Rehabil 2004;85:902–9. 2.  Gracies JM, Singer BJ, Dunne JW. The role of botulinum toxin injec- 19.  Chen HX, Wang W, Xiao HQ, Wang H, Ding XD. Ultrasound-guid- tions in the management of muscle overactivity of the lower limb. Disabil ed botulinum toxin injections and EMG biofeedback therapy the lower Rehabil 2007;29:1789–805. limb muscle spasm after cerebral infarction. Eur Rev Med Pharmacol Sci 3.  Simpson DM, Gracies JM, Graham HK, Miyasaki JM, Naumann M, 2015;19:1696–9. Russman B, et al.; Therapeutics and Technology Assessment Subcom- 20.  The SENIAM project. SEMG sensors; [Internet]. Available from: mittee of the American Academy of Neurology. Assessment: Botulinum http://www.seniam.org/lowerleg_ location.htm [cited 2014, Sep 1]. neurotoxin for the treatment of spasticity (an evidence-based review): re- 21.  Bayram S, Sivrioglu K, Karli N, Ozcan O. Low-dose botulinum toxin port of the Therapeutics and Technology Assessment Subcommittee of the with short-term electrical stimulation in poststroke spastic drop foot: a American Academy of Neurology. Neurology 2008;70:1691–8. preliminary study. Am J Phys Med Rehabil 2006;85:75–81. 4.  Ogawa A, Izumi S, Katayama Y, Kayama T, Suzuki N, Mori E, et al. 22.  Kamper SJ, Maher CG, Mackay G. Global rating of change scales: Japanese guidelines for the management of stroke 2015. Tokyo: Kyowaki- a review of strengths and weaknesses and considerations for design. J kaku; 2015. Japanese. Manual Manip Ther 2009;17:163–70. 5.  Foley N, Murie-Fernandez M, Speechley M, Salter K, Sequeira K, 23.  Turns LJ, Neptune RR, Kautz SA. Relationships between muscle ac- Teasell R. Does the treatment of spastic equinovarus deformity following tivity and anteroposterior ground reaction forces in hemiparetic walking. stroke with botulinum toxin increase gait velocity? A systematic review Arch Phys Med Rehabil 2007;88:1127–35. and meta-analysis. Eur J Neurol 2010;17:1419–27. 24.  Chow JW, Yablon SA, Stokic DS. Coactivation of ankle muscles dur- 6.  Kaji R, Osako Y, Suyama K, Maeda T, Uechi Y, Iwasaki M; ing stance phase of gait in patients with lower limb hypertonia after ac- GSK1358820 Spasticity Study Group. Botulinum toxin type A in post- quired brain injury. Clin Neurophysiol 2012;123:1599–605. stroke lower limb spasticity: a multicenter, double-blind, placebo-con- 25.  Lauziere S, Betschart M, Aissaoui R, Nadeau S. Understanding spa- trolled trial. J Neurol 2010;257:1330–7. tial and temporal gait asymmetries in individuals post stroke. Int J Phys 7.  Burbaud P, Wiart L, Dubos JL, Gaujard E, Debelleix X, Joseph PA, et Med Rehabil 2014;23:201. al. A randomised, double blind, placebo controlled trial of botulinum toxin 26.  Wall JC, Devlin J, Khirchof R, Lackey B. Measurement of step in the treatment of spastic foot in hemiparetic patients. J Neurol Neurosurg widths and step lengths: a comparison of measurements made directly Psychiatry 1996;61:265–9. from a grid with those made from a video recording. J Orthop Sports Phys 8.  Cioni M, Esquenazi A, Hirai B. Effects of botulinum toxin-A on gait Ther 2000;30:410–7. velocity, step length, and base of support of patients with dynamic equin- 27.  Ugbolue UC, Papi E, Kaliarntas KT, Kerr A, Earl L, Pomeroy VM, ovarus foot. Am J Phys Med Rehabil 2006;85:600–6. et al. The evaluation of an inexpensive, 2D, video based gait assessment 9.  Fujita K, Miaki H, Nakagawa T, Hori H, Kobayashi Y. The effects of system for clinical use. Gait Posture 2013;38:483–9. muscle activity and spatio temporal parameters after botulinum treatment 28.  Allen JL, Kautz SA, Neptune RR. Step length asymmetry is repre- for hemiparetic gait. Physical Therapy Japan 2016;43:477–85. [Japanese.] sentative of compensatory mechanisms used in post-stroke hemiparetic 10.  Naumann M, Jankovic J. Safety of botulinum toxin type A: a sys- walking. Gait Posture 2011;33:538–43. tematic review and meta-analysis. Curr Med Res Opin 2004;20:981–90. 29.  Field A. Discovering statistics using IBM SPSS statistics. Second edi- 11.  Gurevich T, Peretz C, Moore O, Weizmann N, Giladi N. The effect of tion. London: Sage Publications; 2005. injecting botulinum toxin type a into the calf muscles on freezing of gait 30.  Gastaldi L, Lisco G, Pastorelli S, Dimanico U. Effects of botuli- in Parkinson’s disease: a double blind placebo-controlled pilot study. Mov num neurotoxin on spatio-temporal gait parameters of patients with Disord 2007;22:880–3. chronic stroke: a prospective open-label study. Eur J Phys Rehabil Med 12.  Jonkers I, Stewart C, Spaepen A. The complementary role of the plan- 2015;51:609–18. tarflexors, hamstrings and gluteus maximus in the control of stance limb 31.  da Cunha-Filho IT, Henson H, Wankadia S, Protas EJ. Reliability of stability during gait. Gait Posture 2003;17:264–72. measures of gait performance and oxygen consumption with stroke survi- 13.  Caleo M, Antonucci F, Restani L, Mazzocchio R. A reappraisal of the vors. J Rehabil Res Dev 2003;40:19–25. central effects of botulinum neurotoxin type A: by what mechanism? J 32.  Hesse S, Krajnik J, Luecke D, Jahnke MT, Gregoric M, Mauritz Neurochem 2009;109:15–24. KH. Ankle muscle activity before and after botulinum toxin therapy for 14.  Dyer JO, Maupas E, Melo SA, Bourbonnais D, Forget R. Abnormal lower limb extensor spasticity in chronic hemiparetic patients. Stroke coactivation of knee and ankle extensors is related to changes in heterony- 1996;27:455–60. mous spinal pathways after stroke. J Neuroeng Rehabil 2011;8:41. 33.  Filippi GM, Errico P, Santarelli R, Bagolini B, Manni E. Botuli- 15.  Marchand-Pauvert V, Aymard C, Giboin LS, Dominici F, Rossi A, num A toxin effects on rat jaw muscle spindles. Acta Otolaryngol Mazzocchio R. Beyond muscular effects: depression of spinal recurrent 1993;113:400–4. inhibition after botulinum neurotoxin A. J Physiol 2013;591:1017–29. 34.  Kerzoncuf M, Bensoussan L, Delarque A, Durand J, Viton JM, Ros- 16.  Roche N, Zory R, Sauthier A, Bonnyaud C, Pradon D, Bensmail si-Durand C. Plastic changes in spinal synaptic transmission following D. Effect of rehabilitation and botulinum toxin injection on gait in botulinum toxin A in patients with post-stroke spasticity. J Rehabil Med chronic stroke patients: a randomized controlled study. J Rehabil Med 2015;47:910–6. 2015;47:31–7. 35.  Mulroy SJ, Klassen T, Gronley JK, Eberly VJ, Brown DA, Sullivan KJ. Gait parameters associated with responsiveness to treadmill training with body-weight support after stroke: an exploratory study. Phys Ther 2010;90:209–23. Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 17

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically FUJITA COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, GAIT ELECTROMYOGRAPHY AFTER BOTULINUM TOXIN ADMINISTRATION cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. 36.  Perry J. Gait analysis: normal and pathological function. Second edi- in people with stroke: a prospective observational study. Arch Phys Med tion. NJ: Slack Inc.; 2010. Rehabil 2008;89:56–61. 37.  Den Otter AR, Geurts AC, Mulder T, Duysens J. Abnormalities in 40.  Tok F, Balaban B, Yaşar E, Alaca R, Tan AK. The effects of onabotu- the temporal patterning of lower extremity muscle activity in hemiparetic linum toxin A injection into rectus femoris muscle in hemiplegic stroke gait. Gait Posture 2007;25:342–52. patients with stiff-knee gait: a placebo-controlled, nonrandomized trial. 38.  Tang SF, Hong JP, McKay WB, Tang CW, Wu PH, Chu NK. Modifi- Am J Phys Med Rehabil 2012;91:321–6. cation of altered ankle motor control after stroke using focal application of 41.  Boudarham J, Hameau S, Pradon D, Bensmail D, Roche N, Zory R. botulinum toxin type A. Clin Neurol Neurosurg 2012;114:498–501. Changes in electromyographic activity after botulinum toxin injection of 39.  Stoquart GG, Detrembleur C, Palumbo S, Deltombe T, Lejeune TM. the rectus femoris in patients with hemiparesis walking with a stiff-knee Effect of botulinum toxin injection in the rectus femoris on stiff-knee gait gait. J Electromyogr Kinesiol 2013;23:1036–43. Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material discussed in the manuscript. Article first published online: June 11, 2018. - Manuscript accepted: June 7, 2018. - Manuscript revised: May 11, 2018. - Manuscript received: January 12, 2018. 18 European Journal of Physical and Rehabilitation Medicine February 2019

COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically © 2018 EDIZIONI MINERVA MEDICA European Journal of Physical and Rehabilitation Medicine 2019 February;55(1):19-28 or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access Online version at http://www.minervamedica.it DOI: 10.23736/S1973-9087.18.05195-X to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. ORIGINAL ARTICLE Evaluation of manual ability in stroke patients in Benin: cultural adaptation and Rasch validation of the ABILHAND-Stroke questionnaire Ditouah D. NIAMA NATTA 1, 2, Charles S. BATCHO 3, 4, Gaëtan G. STOQUART 2, 5, 6, Etienne ALAGNIDÉ 1, Toussaint KPADONOU 1, Thierry M. LEJEUNE 2, 5, 6 * 1Department of Physical Medicine and Rehabilitation, National University Hospital of Cotonou, Cotonou, Benin; 2Institute of Experimental and Clinical Research, Neuro Musculo Skeletal Lab, Catholic University of Louvain, Brussels, Belgium; 3Centre Interdisciplinaire de Recherche en Réadaptation et Intégration Sociale (CIRRIS), Québec, QC, Canada; 4Department of Rehabilitation, Faculty of Medicine, University of Laval, Québec, QC, Canada; 5Department of Physical Medicine and Rehabilitation, University Clinic of Saint-Luc, Brussels, Belgium; 6Catholic University of Louvain, Louvain Bionics, Louvain-La-Neuve, Belgium *Corresponding author: Thierry Lejeune, Thierry M. Lejeune, Department of Physical Medicine and Rehabilitation, University Clinic of Saint-Luc, Av. Hip- pocrate 10, 1200 Bruxelles, Belgium. E-mail: [email protected] ABSTRACT BACKGROUND: ABILHAND is a self-reported questionnaire assessing manual ability. It was validated and calibrated using the Rasch analysis for European stroke patients. After a stroke, performing upper limb activities of daily living is influenced by personal and environmental contex- tual factors. It is thus important to conduct a contextual validation to use this questionnaire outside of Europe. AIM: The aim of this study was to perform a cross cultural validation of the ABILHAND-Stroke questionnaire for post-stroke patients living in Benin, a West-African country. DESIGN: Observational cross-sectional study. SETTING: Outpatient rehabilitation centres. POPULATION: 223 Beninese chronic stroke patients. METHODS: The experimental questionnaire was made of 59 items evaluating manual activities. Patients had to estimate their difficulty of per- forming each activity according to four response categories: impossible, very difficult, difficult and easy. For construct validity analysis, patients were also evaluated with other assessment tools: Box and Block Test, the motor subscale of the Functional Independence Measure, the Stroke Impairment Assessment Set, and ACTIVLIM-Stroke. Data were analysed with the Rasch partial credit model. RESULTS: The response categories very difficult and difficult were merged and the number of response categories was reduced from 4 to 3 (im- possible, difficult and easy). The Rasch analyses selected 16 bimanual activities that fit the Rasch model (chi square=42.35; P=0.10). The item location ranged from -1.10 to 2.24 logits. The standard error ranged from 0.15 to 0.22 logits. There is no differential item functioning between subgroups (age, sex, dexterity, affected side, time since stroke). The person separation index is 0.82. The questionnaire can measure 3 levels of manual ability, similarly to the occidental version. CONCLUSIONS: The ABILHAND-stroke is a Rasch validated, unidimensional and invariant questionnaire to assess manual ability among Beninese patients. The ordinal score can be transformed into linear score using a conversion table. CLINICAL REHABILITATION IMPACT: This assessment tool is clinically relevant in Benin, a developing country, since it requires no specific equipment or training. It should promote and standardize assessments for stroke patients in clinical practice and research in this African country. (Cite this article as: Niama Natta DD, Batcho CS, Stoquart GG, Alagnidé E, Kpadonou T, Lejeune TM. Evaluation of manual ability in stroke patients in Benin: cultural adaptation and Rasch validation of the ABILHAND-Stroke questionnaire. Eur J Phys Rehabil Med 2019;55:19-28. DOI: 10.23736/ S1973-9087.18.05195-X) Key words: Patient outcome-assessment - Surveys and questionnaires - Arm - Stroke. Stroke is the leading cause of neurological disabilities Four years post stroke, 45% of patients still present some in adults.1 Upper limb function is more frequently af- functional upper limb impairments2 that bear a negative fected, and is harder to restore than lower limb function.1 impact on autonomy and quality of life.3 Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 19

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically NIAMA NATTA COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, MANUAL ABILITY IN STROKE PATIENTS IN BENIN cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. Accurate, reliable and valid functional evaluations are Materials and methods essential in order to orientate treatments and assess their efficacy.4 The World Health Organization (WHO) devel- The study was approved by the Ethics Committee of the oped the International Classification of Functioning, Dis- University Hospital National Centre, Hubert Koutoukou ability and Health (ICF) to provide a common framework Maga (CNHU-HKM) of Cotonou. and standardized terminology.5 The ICF activity domain is essential in neurological rehabilitation, since its main An observational cross-sectional study was performed. objective is to improve patients’ functional abilities. Ac- Classical and modern measurement theories were used to tivities’ assessment is mainly based on questionnaires to assess the questionnaire psychometric qualities following determine what type of activities of daily living (ADL) the COSMIN taxonomy.12 The study was approved by the patients can perform. Several tools are available to assess committee on research ethics at the institution in which the upper limb activity, such as the Action Research Arm Test, research was conducted and any informed consent from the Wolf Motor Function Test, the Arm Motor Ability Test human subjects was obtained as required. or ABILHAND.6 The choice of an appropriate assessment tool should be based on its psychometric properties and Patients clinical relevance (time taken to administer the test and interpret its scores, cost, need for specific equipment and/ A convenience sample of patients was recruited within the or training, as well as how easily the test can be used in outpatient rehabilitation clinics in the Physical Medicine different settings).7 and Rehabilitation department of the CNHU-HKM, and in the Army Hospital of Cotonou between September 2012 The ABILHAND questionnaire was developed to to April 2016. In order to be included in this study patients evaluate manual ability within the ICF activity domain.5 had to meet the following criteria: 18 years of age or older, Manual ability is defined as the “ability of a person to time since stroke >6 months, living at home in Cotonou, perform ADL requiring the use of upper limbs”8. The and autonomous gait,8 i.e. a score of at least 5/7 for the ABILHAND questionnaire was validated using the items “toilet transfer” and “locomotion” on the Functional Rasch analysis for stroke patients8 and patients with Independence Measure (FIM).13 Patients were excluded if other pathologies.9 It specifically evaluates bimanual ac- they presented with the following criteria: major neglect tivities.8 It exhibits excellent psychometric properties and (Bells test >26/35),14 any concomitant pathology that is now recommended for assessment of upper extremity could potentially limit manual ability, comprehension and activity in research and clinical praxis.10 This assessment expression aphasia or cognitive disorders (score on the tool would be clinically relevant in developing coun- Mini-Mental State Examination <20).15 tries, since it requires no specific equipment or training. However, like most assessment tools, it was mainly vali- Development of ABILHAND-Stroke Benin questionnaire dated and calibrated for European populations. Recently, (content and cross-cultural validity) a cross-cultural validation was conducted for Brazilian populations.11 Based on the literature review of existing scales and ex- perts’ opinions, Penta et al.8 developed an initial 56-item In light of limited resources in developing countries, French questionnaire covering various manual activities. patients’ rehabilitation and functional assessment must be This version was submitted to one Beninese stroke patient as cost-effective as possible, for the patient as well as for and to 7 Beninese rehabilitation professionals (six physi- the healthcare system. Questionnaires are thus relevant cal therapists and one Physical Medicine and Rehabilita- to evaluate patients in these countries. However, to our tion specialist). They studied its content and cross-cultural knowledge, no questionnaire assessing upper limb activity validity according to a consensus-based decision-making. has been validated for the sub-Saharan population. After a Experts were asked to identify items relevant in the socio- stroke, performing upper limb ADL is influenced by per- cultural context of Benin, to modify items as needed or sonal and environmental contextual factors according to to suggest other items. Among the initial 56 items, two the ICF model.5 It is thus important to conduct a contextual items were discarded because they did not match the validation to use this questionnaire outside of Europe. The socio-cultural realities in Benin: “tearing open a pack objective of this study was to validate the ABILHAND of chips” and “preparing crepe batter”. Four items were questionnaire in a population of stroke patients living in modified: “spreading butter on a slice of bread” became Benin, a West-African Sub-Saharan country. “buttering a slice of bread;” “eating a sandwich” became “eating a sandwich (bread with filling);” “cracking nuts” 20 European Journal of Physical and Rehabilitation Medicine February 2019

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA NIAMA NATTA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, MANUAL ABILITY IN STROKE PATIENTS IN BENIN cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. became “cracking nuts (or palm nuts);” “opening a letter” tion.20 Developed by George Rasch,21 this probabilistic became “opening a letter (closed envelope).” Furthermore, model helps build a continuous, linear and unidimensional five items were added: “using a matchstick to light a fire,” scale.22 These psychometric properties contribute to the “praying with a rosary,” “tying a loincloth around the structural validity, which is one component of construct waist,” “using a toothpick to clean your teeth” and “shake validity.12 The likelihood test has been performed and the someone’s hand.” Finally, the experimental questionnaire result was significant (χ2=104,66; P<0.0001) then the par- administered to patients was made of 59 items (34 uni- tial credit model has been used. manual activities and 25 bimanual activities). These ex- perts also classified bimanual items into three difficulty Item selection levels: easy bimanual tasks (group A) were those that could be broken down into unimanual sequences; moder- Items selection was based on the following criteria that ately difficult bimanual tasks (group B) i.e. those requiring were successively checked: the use of the healthy hand and the affected hand to stabi- lize an object; and difficult bimanual tasks (group C) i.e. •  missing answers: items with a high proportion of those requiring the use of the healthy hand combined with missing answers (> 50%) were discarded, since they were a digital activity of the affected hand. The content validity considered as unfamiliar or unrepresentative activities for was also studied by analyzing the Spearman correlation most patients; between items’ difficulty levels assessed by the experts and the items’ difficulties resulting from Rasch analyses. •  ordering of response categories: patients had to an- swer the items according to four response categories. Ad- Testing procedure jacent response categories were separated by a threshold, corresponding to a location where a patient had an equal The first part consisted in collecting socio-demographics probability of scoring one of the adjacent categories. When and functional data as well as other variables: Stroke Im- subjects could not discriminate between the different cat- pairment Assessment Set (SIAS), an ordinal scale evalu- egories, thresholds were disordered.23 For disordered ating stroke-related neurological impairments (score 0 to thresholds, the chosen solution was to combine response 76);16 Box and Block Test (BBT) evaluating gross manual categories;24 dexterity;17 Geriatric Depression Scale (GDS) an ordinal scale assessing depression in elderly individuals (nor- •  scale targeting: a questionnaire is meant to target the mal ≤9, moderate depression between 10 and 19, severe specific population it is supposed to assess. Distributions depression between 20 and 30);18 motor subscale of the of individual ability levels and item difficulties were plot- Functional Independence Measure (FIM-Mo) consisting ted on a common metric logit scale to compare the loca- of the first 13 items (score 13 to 91);13 and ACTIVLIM- tion of persons and items. It allows to evaluate graphically Stroke, a Rasch validated questionnaire assessing limita- if items difficulties cover all levels of individual abili- tions in performing ADL.19 ties.25, 26 Items identified as “easy” or “impossible” by all subjects were removed as useless extreme items. Floor and The second part of the evaluation consisted in patients ceiling effects were assessed by computing the percent- completing the French version of the experimental ABIL- age of patients with a manual ability lower than the easiest HAND-Stroke Benin questionnaire. They had to self-as- item or greater than the most difficult item. This percent sess their ability to perform each of the 59 activities using must be lower than 15%;27 a four-level scale: impossible (0), very difficult,1 difficult2 and easy.3 Activities not performed in the past 3 months •  item-trait interaction: overall data fit is verified in were qualified as “no answer given”. If a patient was unable term of item-trait interaction assessed by a χ2 test. This χ2 to speak or read French properly, the assessor performed an test must be non-significant;26 interview in French or in the patient’s native language. In- deed, even though French is the official Beninese language, •  individual item and person fit to the Rasch model: there are around 70 dialects spoken in Benin. two fit statistics were computed as a function of all in- dividual responses to all items, the residual and χ2. The Rasch analyses residual is the difference between the model’s expected scores and the observed scores. The standardized residual The Rasch model has been widely used for the develop- was obtained by dividing the residual by an estimate of ment and validation of assessment tools in rehabilita- its standard deviation. The standardized residual consisted in a normal distribution with a mean of 0 and a standard deviation of 1. Persons and items presenting with fit re- siduals located within a range between -2.5 and 2.5 were Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 21

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically NIAMA NATTA COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, MANUAL ABILITY IN STROKE PATIENTS IN BENIN cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. deemed acceptable.26, 28 The software divided patients into convergent validity was studied by analyzing the corre- three incremental class intervals according to their level of lation between the manual ability assessed with ABIL- ability. The χ2 fit index was calculated as the standardized HAND-Stroke Benin and other functional ability scales: residual sum of the squares for each class interval. A sig- ACTIVLIM-Stroke, FIM-Mo, SIAS, BBS and GDS. nificant Bonferroni adjusted P value meant the items did Pearson correlation coefficient was used for continuous not fit the Rasch model;29 variables and Spearman’s rank correlation coefficient for ordinal variables. We expected patients with a high score •  local independence means there was no major cor- on the ABILHAND-Stroke Benin to also have a high relation between two items after discarding the underly- score on the ACTIVLIM-Stroke, FIM-Mo, BBT and SIAS ing trait effect. The correlation matrix analysis was used tests. Conversely, as suggested by Penta et al.,8 patients to verify this criterion. When two items had a correlation with major depression should have a low manual ability index ≥0.3, the item less contributive to the global fit index score. The correlation index was considered strong when was removed;25 ≥0.7; moderate when comprised between 0.3 and 0.7 and weak when <0.3.33 Three-way analysis of variance (3- •  unidimensionality is the property of the questionnaire way ANOVA) was used to evaluate the differences in the to measure a single attribute of the underlying trait. The manual ability score of patients according to the following unidimensional nature of the questionnaire was tested by characteristics: age, sex and affected side. The difficulties the method proposed by Smith.25 After a principal compo- of the items retained in both the Beninese and the Euro- nent analysis, the factor loadings of the first component’ pean ABILHAND questionnaires were compared using a residuals were used to determine two subsets of items that Student t-test and an ICC. were then compared by a paired t-test.25 The percentage of significant tests (P<0.05) should be below 5% for a ques- Results tionnaire to be called unidimensional; We recruited 233 patients who fitted the criteria listed •  differential item functioning (DIF). If an item fits above. Mean age was 54±9.7 years. Sample’s characteris- the Rasch model, patients’ responses should only be de- tics are summed up in Table I. termined by the subject’s level of manual ability and the difficulty level of the activity measured by the item.21 If Table I.—P atient characteristics. the response to an item is influenced by another variable, it means that persons with the same ability level give a Age in years: median[Range] 55 [28-79] different response to this item. So responses to this item Sex are biased and the item exhibits DIF.30 DIF was evaluated 149 (66.8%) with the Wright and Stone method31 via five variables. The Male 74 (33.2%) patients were divided in two subgroups based on their sex, Female affected side; on their median value of age (55 years), time Stroke type 79 (35.4%) since stroke (2 years), and BBT Score (39 blocks). Item Ischemic 33 (14.8%) location was determined in each subgroup. A Student t-test Hemorrhagic 111 (49.8%) and an Intra Class Correlation (ICC) were used to compare Undefined (brain CT-Scan not available) items locations between subgroups and for each variable. Paretic side 108 (48.4%) Dominant 115 (51.6%) Reliability index PSI Non-dominant Time since stroke 27 [6-99] Reliability was evaluated in terms of person separation in- Delay in month: Median [Range] 99 (44.4%) dex (PSI).32 It was calculated as the ratio of the unbiased ≤2 years 124 (55.6%) estimate of the sample standard deviation to the root mean >2 years 87 [30-87] square measurement error of the sample. It can estimate FIM-Mo: median [Range] 32 [0-75] the number of ability levels that can be measured by the BBT: median [Range] 65 [19-76] questionnaire. A reliability ≥ 0.80 and number of ability SIAS: median [Range] levels ≥2 are deemed acceptable.22, 32 GDS (N.=179) 9 [0-27] Median [range] 87 (48.6%) Convergent validity <10 92 (51.4%) ≥10 Construct validity was also investigated in terms of con- vergent validity with classic psychometric methods. The FIM-Mo: motor subscale of the functional independence measure; BBT: Box and Block Test; SIAS: Stroke Impairment Assessment Set; GDS: Geriatric Depression Scale. 22 European Journal of Physical and Rehabilitation Medicine February 2019

COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA MANUAL ABILITY IN STROKE PATIENTS IN BENIN This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically NIAMA NATTA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, fit, and five other items were removed because of local cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. dependency (correlation index ≥ 0.3 with other items). Probability Fit statistics Probability Person location (logit) Person location (logit) The final questionnaire included 16 items (Table II). The overall fit statistics (mean±SD) for items was -0.12±0.61 Impossible logit and for persons was -0.27±1.03 logit where fit sta- tistics are standardized to a mean of 0 and a SD of 1. The Impossible Difficult item-trait interaction showed a good global fit to the model for the scale (χ2= 42.35; degrees of freedom =32; P=0.10). Very difficult Easy The item location ranged from -1.10 to 2.24 logits, cor- Difficult responding to an interval of 3.34 logits with a mean item location of 0.00±1.04 logit. The standard error rate varied A Easy B from 0.15 to 0.22 logit. The thresholds for the different response categories ranged from -2.23 to 3.03 logits. The Figure 1.—Curve of the probabilities of response categories. A) Curve person location ranged from -2.13 to 3.78 logits, corre- with four categories of responses showing disordered thresholds for the sponding to an interval of 5.91 logits with a mean person response categories; B) curve with three categories of responses, no dis- location of 1.0±1.37 logits. ordered thresholds. In our sample, none of the patients reported all items im- Selection of items possible (0% floor effect). Nine subjects in our sample had ability greater than 3.03 logits and could answer “easy” No item was deleted because of a missing response. The to all items (6.4% ceiling effect). Figure 2 shows that the first analysis revealed that all 59 items showed disordered items difficulties cover relatively well the patients’ manual thresholds with the categories “very difficult” and “dif- abilities, even if the individual ability was overall superior ficult” being not well differentiated by patients (Figure to the difficulty of the items (Figure 2). Table II reports 1A). Thus, these two categories were merged, leading to a items calibration and fit indices. Table III underlines the three-level scale: impossible (0), difficult1 and easy.2 After relationship between the total raw scores, the scores in log- a new analysis, the response categories were properly or- its and the scores in percentages. In clinical practice, if pa- dered (Figure 1B). tient answered all questions, the ordinal score can be easily transformed in interval score, expressed in percentage. If Then the targeting of the scale was verified. It appeared that the 34 unimanual activities were labelled as easy by all patients. These items did not permit to differentiate pa- tients according to their manual ability level and were dis- carded. After several successive analyses on the remaining 25 items, four items were discarded because of their poor Table II.—Item difficulty location and fit index for the 16 items of the questionnaire. Items Difficulty (logit) SE (logit) Residuals χ2 DF P value* Bimanual implication Fastening a snap button (jacket, purse) -1.10 0.19 -0.04 5.48 2 0.06 A Opening a jar -1.04 B Spreading toothpaste on a toothbrush -1.02 0.20 -0.13 0.99 2 0.61 A Closing the zipper on trousers -0.91 A Opening a chocolate bar -0.89 0.17 -1.02 1.70 2 0.43 B Buttering a slice of bread -0.86 B Opening a letter (closed envelope) -0.74 0.20 0.55 1.17 2 0.55 B Using a matchstick to light a fire -0.24 B Opening a bottle using a bottle opener -0.22 0.2 0.93 3.88 2 0.14 B Tying a loincloth around the waist 0.20 C Buttoning trousers 0.39 0.19 -0.88 2.58 2 0.28 B Cutting meat with a fork and knife 0.66 C Sharpening a knife 0.83 0.23 -0.17 3.06 2 0.22 C Hammering a nail 1.33 C Clipping nails 1.38 0.16 -0.43 0.73 2 0.69 C Threading a needle 2.23 C 0.17 -0.38 3.72 2 0.15 0.19 0.09 0.54 2 0.76 0.15 -1.13 2.13 2 0.34 0.18 0.06 0.55 2 0.75 0.17 0.82 3.45 2 0.17 0.16 -0.97 5.69 2 0.05 0.16 0.27 2.86 2 0.24 0.17 0.11 3.78 2 0.15 DF: degrees of freedom. *P value after Bonferroni adjustment = 0.002. Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 23

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically NIAMA NATTA COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, MANUAL ABILITY IN STROKE PATIENTS IN BENIN cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. Persons Person item threshold distribution the patient did not answerer all questions, then a software (grouping set to interval length of 0.20 making 35 groups) or a website (for instance http://www.rehab-scales.org) is required to transform the ordinal score into interval score. Frequency Differential item functioning Freq Item DIF was evaluated through five variables. The Stu- Location dent t-test showed no significant difference in items’ dif- Items (logits) ficulty between each factor’ subgroup (P>0.05), and ICCs ranged from 0.64 to 0.85 (P<0.001) (Figure 3). Figure 2.—Person item distribution. The upper graph shows the fre- Unidimensionality quency of person distribution according to their ability; the graph below shows the frequency of item distribution according to their difficulty. The paired t-test significance level (4.52%) was lower than The x-axis shows items and persons location on the same scale; the y- the acceptable 5%. This validates the unidimensional na- axis shows distribution frequencies. ture of the scale. Table III.—O rdinal-to-interval score transformation table. Reliability Ordinal score Interval score in logit Interval score in The ABILHAND-Stroke Benin questionnaire showed percentage (%) good reliability. The PSI was 0.82 after discarding extreme 0 -4.37 scores. This indicates that three levels of ability can be dis- 1 -3.54 0 tinguished within the patient’ sample. 2 -2.97 9 3 -2.57 16 Content validity 4 -2.26 20 5 -1.99 23 The items’ difficulty levels assessed by the experts are 6 -1.77 26 strongly correlated with the items’ difficulties resulting 7 -1.56 29 from Rasch analyses (rho =0.88, P<0.001). 8 -1.37 31 9 -1.19 33 Convergent validity 10 -1.02 35 11 -0.86 37 The patients’ manual ability evaluated with ABILHAND- 12 -0.69 39 Stroke was moderately to strongly correlated with their 13 -0.54 41 neurological impairments, manual dexterity, and ability 14 -0.39 43 to perform ADL (P<0.001, rho and r ranged from 0.56 15 -0.23 44 to 0.89, Table IV). On the other hand, their manual ca- 16 -0.08 46 pacity was negatively, albeit discretely, correlated with 17 0.08 48 their depressive state assessed with the GDS. The 3-way 18 0.23 49 ANOVA (Table IV) showed that patients’ responses were 19 0.4 51 not influenced by age, sex or laterality of the affected side 20 0.57 53 (P>0.24). The Student-t test showed no significant differ- 21 0.75 55 ence between the item difficulties for items common to 22 0.93 57 both the Beninese and European ABILHAND question- 23 1.13 59 naires (P=0.72) and ICC=0.70. 24 1.33 61 25 1.55 63 Discussion 26 1.79 66 27 2.07 68 This study presents the validation of the ABILHAND- 28 2.37 71 Stroke questionnaire for the Beninese population. This 29 2.72 75 self-reporting questionnaire assesses 16 bimanual activi- 30 3.17 79 ties using three-response categories. Rasch analyses dem- 31 3.79 84 32 4.64 91 100 In clinical practice, if patient answered all questions, the ordinal score can be easily transformed in interval score, expressed in logit or in percentage. If the patient did not answer all questions, then a software or a website (for instance http://www.rehab-scales.org) is required to transform the ordinal score into interval score. 24 European Journal of Physical and Rehabilitation Medicine February 2019

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA NIAMA NATTA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, MANUAL ABILITY IN STROKE PATIENTS IN BENIN cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. Age 1 ICC=0.85; P<0.001 Women ICC=0.84; P<0.001 BBT 1 ICC=0.87; P<0.001 Age 2 Men BBT 2 Dominant affected ICC=0.64; P=0.007 Time 1 ICC=0.7; P=0.003 Non dominant affected Time 2 Figure 3.—ABILHAND-Stroke questionnaire invariance via seven subgroups: age (age1: ≤55 years; age2: >55 years), sex, time since stroke (time1: ≤ 2years; time 2: >2years), laterality of the affected side, motor impairment score with the SIAS (SIAS1: ≤65; SIAS2: >65), manual dexterity score of the affected side evaluated with the BBT (BBT1: ≤32; BBT2: >32); depression level based on the GDS (GDS1: <10; GDS2: ≥10) scores. Lines indicate a confidence interval of 95%. Points outside of the two lines would indicate items presenting with differential item functioning. No item of the questionnaire presents with differential functioning. Table IV.—R elationship between the manual ability score of pa- onstrated its linearity, unidimensionality, invariance and tients and population characteristics and functional assessment. reliability. ABILHAND-Stroke is clinically relevant and well adapted to the Beninese context. Statistics P value The Beninese version of the ABILHAND-Stroke ques- Age F=0.10 0.74 tionnaire is made of 16 bimanual activities. Fourteen items Sex F=1.35 0.24 are common to both Beninese and European versions and Affected side F=0.14 0.71 their hierarchical orders are strongly correlated. The other BBT rho=0.66 <0.001 two items correspond to activities added by the Beninese SIAS rho=0.60 <0.001 experts during the content validity stage: “using a match- GDS rho=-0.24 <0.001 stick to light a fire” and “tying a loincloth around the FIM-Mo rho=0.56 <0.001 waist.” ACTIVLIM-Stroke r=0.89 <0.001 The content validity is also supported by the strong cor- The correlation coefficients (r = Pearson; rho = Spearman) present the relation relation between the items difficulties and their bimanual between the patients’ manual ability and their gross manual ability (Box and requirements (Table II). The easiest item (Fastening a Block Test), neurological impairments (SIAS), depression state (GDS), functional snap button, -1.10 logit) belonged to group A, the mod- independence (FIM-Mo and Activlim). BBT: Box and Block Test; SIAS: Stroke Impairment Assessment Set; GDS: Geriatric Depression Scale; FIM-Mo: Motor subscale of the Functional Independence Measure; F: three-way (age, sex and affected side) analysis of variance assessing the effect of age, sex and affected side on the person location. Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 25

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically NIAMA NATTA COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, MANUAL ABILITY IN STROKE PATIENTS IN BENIN cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. erately difficult item (Using a matchstick to light a fire, tion when using ordinal scales.37 Nevertheless, the Rasch -0.24 logit) belonged to group B, and the most difficult method has some drawbacks. It requires specialized skills item (Threading a needle, 2.23 logit) belonged to group C. and software and a large amount of data for the question- naire development (>100).38 The ABILHAND-Stroke Benin had a lower number of items (16 items) compared to the European and Brazilian Manual ability evaluated with the ABILHAND-Stroke versions (23 items). Indeed, several items from the original Benin was well correlated with neurological impairments version were discarded because they did not fit the Rasch and manual dexterity: 36% of the variance of the manual model. There are no standard guidelines for the required ability score are explained by the SIAS, and 44% are ex- number of items in a scale, yet it contributes to the reliabil- plained by the affected hand BBT score (Table IV). Sev- ity and targeting of the assessing tool.34 Despite this low eral other authors already underlined this relationship number of items, the ABILHAND-Stroke Benin measures between the body function and structure domain and the a difficulty range of 3.34 logits, which is targeted with the activity domain of the ICF. Indeed, 36% of the manual range of manual ability of the Beninese patients. The range ability variance can be explained by the BBT score among of measure of ABILHAND-Stroke Benin is similar to the stroke patients8 and 49% in children with cerebral palsy.39 questionnaires validated in Europe (3.90 logits)8 and Bra- zil (4.11 logits).11 Activity limitations are related to the measures of a pa- tient’ manual ability. The manual ability was moderately- Miller et al. estimated that an individual can discrimi- to-strongly correlated to the global activity limitations nate up to 7 levels of response categories.35 However, evaluated with the FIM and ACTIVLIM-stroke. In our Rasch validation often leads to reducing the number of study, 79% of the manual ability variance was explained levels. For instance, all ABILHAND questionnaires, vali- by the ACTIVLIM-Stroke score (Table IV). Similarly, dated for assessing manual ability among several popula- Vandervelde et al.40 showed that 58% of the manual abil- tions and across several pathologies, included only 3 lev- ity variance could be explained by the global activity els of responses categories. Indeed, their Rasch analyses limitations among patients with neuromuscular disorders. demonstrated that patients were unable to differentiate One can speculate that ABILHAND and ACTIVLIM are more than 3 responses categories for manual ability.8 The complimentary, even though they do not assess the same number of response categories depends on the patient’s variables as suggested by Vandervelde et al..40 As a mat- discrimination capacity and the studied variable.22 ter of fact, the ACTIVLIM-stroke scale evaluates activ- ity limitations related to the upper limb and lower limb ABILHAND-Stroke Benin showed a good but relative- function whereas the ABILHAND-stroke only assesses ly low PSI (0.82) compared to the European (0.90) and activity limitations for the former. Thus, the ABILHAND- Brazilian (0.91) versions. However, the number of abil- stroke questionnaire is especially relevant in evaluations ity levels for the ABILHAND-Stroke Benin (3 levels) was and therapies targeting upper limb functions. slightly lower than the 4 levels of the European version and the 5 levels of the Brazilian version.11 The minimal In our study, hand ability measures were not correlated level that a questionnaire should be able to discriminate is with patients’ age or sex. These measures were also associ- two.22 These lower PSI values and levels of ability might ated with neither the laterality of the affected side nor the be related to the lower number of items.36 time since stroke. These results observed along with pa- tients’ scores strengthened the results of the item location The Rasch validation resulted in a linear scale. The re- analysis in terms of differential item functioning. These sults can then be submitted to arithmetic computation and results were in accordance with those reported in previous parametric statistics which reduces measurement error. studies.8 The linearity of the scale allowed the expression of results in percentages instead of the not well-known logit unit at Limitations of the study using the conversion table. In clinical practice, results ex- pressed in percentages should be clearer and more mean- This study presents two limits. The coexistence of the ingful for therapists and patients. Contrarily to the raw official language (French) and multiple dialects (±70) in score, one per cent represents the same amount of manual Benin was one difficulty and represented a potential limi- ability throughout the scale. Indeed, a five-point differ- tation for this study. To reach a large number of persons, ence on the raw score corresponds to a 25% difference at representative of the Beninese population, the French lan- the extremities of the scale and only 9% in the middle of guage was used. When patients did not speak French or the scale (Table III). This is why editors recommend cau- were unable to read French, the assessor conducted the 26 European Journal of Physical and Rehabilitation Medicine February 2019

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA NIAMA NATTA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, MANUAL ABILITY IN STROKE PATIENTS IN BENIN cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. interview in the patient’s native language. On one hand, 5.  World Health Organization. International Classification of Function- this increased the validity of the answers, but, on the other ing, Disability and Health. Geneva: World Health Organization; 2001. hand, this could have represented a potential bias. The tar- Available from: http://www.who.int/classifications/icf/en/ [cited 2019, geting is suboptimal. Indeed, similarly to the Brazilians Jan 7]. subjects,11 the mean capacity of the Beninese subjects is 6.  Lang CE, Bland MD, Bailey RR, Schaefer SY, Birkenmeier RL. Assess- 1.0±1.37 logits whereas it should ideally be between -0.5 ment of upper extremity impairment, function, and activity after stroke: and +0.5 logit.41 But, the range of the items thresholds foundations for clinical decision making. J Hand Ther 2013;26:104–14, covers well the range of the patient’s locations, and the quiz 115. observation of Figure 2 shows that patients and items are 7.  Rowland TJ, Gustafsson L. Assessments of Upper Limb Ability fol- evenly distributed. lowing Stroke: a Review. Br J Occup Ther 2008;71:427–37. 8.  Penta M, Tesio L, Arnould C, Zancan A, Thonnard JL. The ABIL- Following the COSMIN taxonomy, several psychomet- HAND questionnaire as a measure of manual ability in chronic stroke pa- ric properties were demonstrated for the ABILHAND- tients: rasch-based validation and relationship to upper limb impairment. stroke Benin questionnaire: structural, cross-cultural and Stroke 2001;32:1627–34. convergent validity contributing to construct validity, and 9.  Arnould C, Vandervelde L, Batcho CS, Penta M, Thonnard JL. internal consistency contributing to reliability. However, Can manual ability be measured with a generic ABILHAND scale? A several other psychometric properties should be further cross-sectional study conducted on six diagnostic groups. BMJ Open explored in a future study: test-retest reliability, measure- 2012;2:1–9. ment error and responsiveness. All patients evaluated in 10.  Alt Murphy M, Resteghini C, Feys P, Lamers I. An overview of sys- this study had an autonomous gait. This choice was made tematic reviews on upper extremity outcome measures after stroke. BMC to avoid the impact of locomotion on manual ability, and Neurol 2015;15:29. to recruit patients who were likely to experience several 11.  Basílio ML, de Faria-Fortini I, Magalhães LC, de Assumpção FS, de manual activities in their daily environment.22 This inclu- Carvalho AC, Teixeira-Salmela LF. Cross-cultural validity of the Brazil- sion criterion did not generate a bias favoring the recruit- ian version of the ABILHAND questionnaire for chronic stroke individu- ments of patients with greater manual ability. Indeed only als, based on Rasch analysis. J Rehabil Med 2016;48:6–13. 6.4% of the patients were able to easily perform all items 12.  Mokkink LB, Terwee CB, Patrick DL, Alonso J, Stratford PW, Knol and the patients presented with a wide range of manual DL, et al. The COSMIN checklist for assessing the methodological qual- dexterity (BBT range 0-75). ity of studies on measurement properties of health status measurement in- struments: an international Delphi study. Qual Life Res 2010;19:539–49. Conclusions 13.  Gellez-Leman MC, Colle F, Bonan I, Bradai N, Yelnik A. [Evalua- tion of the disabilities of hemiplegic patients]. Ann Readapt Med Phys After ACTIVLIM-Stroke19 and ABILOCO-Stroke,42 2005;48:361–8. French. ABILHAND-Stroke Benin does increase the number of 14.  Gauthier L, Dehaut F, Joanette Y. The Bells test: a quantitative and Rasch validated scales available in the African context. qualitative test for visual neglect. Int J Clin Neuropsychol 1989;11:49–54. Together, they will help promote and standardize assess- 15.  Crum RM, Anthony JC, Bassett SS, Folstein MF. Population-based ments for stroke patients’ in clinical practice and improve norms for the Mini-Mental State Examination by age and educational research in this population. level. JAMA 1993;269:2386–91. 16.  Liu M, Chino N, Tuji T, Masakado Y, Hase K, Kimura A. Psychomet- References ric properties of the Stroke Impairment Assessment Set (SIAS). Neurore- habil Neural Repair 2002;16:339–51. 1.  Kwakkel G, Kollen BJ, Wagenaar RC. Therapy impact on functional 17.  Mathiowetz V, Volland G, Kashman N, Weber K. Adult norms recovery in stroke rehabilitation: a critical review of the literature. Phys- for the Box and Block Test of manual dexterity. Am J Occup Ther iotherapy 1999;13:457–70. 1985;39:386–91. 2.  Broeks JG, Lankhorst GJ, Rumping K, Prevo AJ. The long-term out- 18.  Brink TL, Yesavage JA, Lum O, Heersema PH, Adey M, Rose TL. come of arm function after stroke: results of a follow-up study. Disabil Screening tests for geriatric depression. Clin Gerontol 1982;1:37–43. Rehabil 1999;21:357–64. 19.  Batcho CS, Tennant A, Thonnard JL. ACTIVLIM-Stroke: a cross- 3.  Pang MY, Harris JE, Eng JJ. A community-based upper-extremity cultural Rasch-built scale of activity limitations in patients with stroke. group exercise program improves motor function and performance of Stroke 2012;43:815–23. functional activities in chronic stroke: a randomized controlled trial. Arch 20.  Tesio L. Measuring behaviours and perceptions: rasch analysis as a Phys Med Rehabil 2006;87:1–9. tool for rehabilitation research. J Rehabil Med 2003;35:105–15. 4.  Ashford S, Slade M, Turner-Stokes L. Conceptualisation and develop- 21.  Rasch G. Probabilistic models for some intelligence and attainment ment of the arm activity measure (ArmA) for assessment of activity in the tests. Chicago, IL: University of Chicago; 1960. hemiparetic arm. Disabil Rehabil 2013;35:1513–8. 22.  Penta M, Arnould C, Decruynaere C. Développer et interprèter une échelle de mesure: Applications du modèle de Rasch. Brussels: Edition Mardaga; 2005. 23.  Andrich D. Category ordering and their utility. Rasch Meas Trans 1996;9:464–5. 24.  Zhu W, Updyke WF, Lewandowski C. Post-hoc Rasch analysis of optimal categorization of an ordered-response scale. J Outcome Meas 1997;1:286–304. 25.  Smith EV Jr. Detecting and evaluating the impact of multidimension- ality using item fit statistics and principal component analysis of residuals. J Appl Meas 2002;3:205–31. Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 27

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically NIAMA NATTA COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, MANUAL ABILITY IN STROKE PATIENTS IN BENIN cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. 26.  Pallant JF, Tennant A. An introduction to the Rasch measurement 35.  Miller GA. The magical number seven plus or minus two: some model: an example using the Hospital Anxiety and Depression Scale limits on our capacity for processing information. Psychol Rev (HADS). Br J Clin Psychol 2007;46:1–18. 1956;63:81–97. 27.  Lo C, Liang WM, Hang LW, Wu TC, Chang YJ, Chang CH. A psy- 36. Frisbie DA. Reliability of scores from teacher-made tests. Educational chometric assessment of the St. George’s respiratory questionnaire in Measurement: Issues and Practical. National council on Measurement in patients with COPD using Rasch model analysis. Health Qual Life Out- Education 1988;7:25-35. comes 2015;13:131. 37.  Grimby G, Tennant A, Tesio L. The use of raw scores from ordinal 28.  Smith RM. IPARM: Item and Person analysis with Rasch model. Chi- scales: time to end malpractice? J Rehabil Med 2012;44:97–8. cago: MESA Press; 1991. 38.  Institute for Objective Measurement (IOM). The Rasch Model as a 29.  Andrich D, editor. Rasch analysis for measurement. London: Sage Construct Validation Tool; [Internet]. Available from: http://www.rasch. Publications Ltd; 1988. org/rmt/rmt221a.htm [cited 2019, Jan 7]. 30.  Holland PH, Wainer H. Differential Item Functioning. Hillsdale, NJ: 39.  Arnould C, Penta M, Thonnard JL. Hand impairments and their re- Lawrence Erlbaum; 1993. lationship with manual ability in children with cerebral palsy. J Rehabil 31.  Wright DB, Stone MH. Best Test Design. Chicago, IL: MESA Press; Med 2007;39:708–14. 1979. 40.  Vandervelde L, Van den Bergh PY, Penta M, Thonnard JL. Validation 32.  Bond TG, Fox CM. Applying the Rasch model: fundamental mea- of the ABILHAND questionnaire to measure manual ability in children surement in the human sciences. Second edition. Mahwah, NJ: Lawrence and adults with neuromuscular disorders. J Neurol Neurosurg Psychiatry Erlbaum Associates; 2007. 2010;81:506–12. 33.  Andresen EM. Criteria for assessing the tools of disability outcomes 41.  Linacre JM. Sample size and item calibration stability. Rasch Meas research. Arch Phys Med Rehabil 2000;81(Suppl 2):S15–20. Trans 1994;7:28. 34.  Hinkin TR, Tracey JB, Enz CA. Scale construction: developing reli- 42.  Sogbossi ES, Thonnard JL, Batcho CS. Assessing locomotion abil- able and valid measurement instruments. J Hosp Tour Res (Wash DC) ity in West African stroke patients: validation of ABILOCO-Benin scale. 1997;21:100–20. Arch Phys Med Rehabil 2014;95:1470–6.e3. Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material discussed in the manuscript. Article first published online: August 29, 2018. - Manuscript accepted: August 27, 2018. - Manuscript revised: July 25, 2018. - Manuscript received: January 22, 2018. 28 European Journal of Physical and Rehabilitation Medicine February 2019

COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically © 2018 EDIZIONI MINERVA MEDICA European Journal of Physical and Rehabilitation Medicine 2019 February;55(1):29-34 or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access Online version at http://www.minervamedica.it DOI: 10.23736/S1973-9087.18.05210-3 to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. ORIGINAL ARTICLE Responsiveness and predictive validity of the Tablet-based Symbol Digit Modalities Test in patients with stroke Pei-Chi HSIAO 1, Wan-Hui YU 2, Shih-Chieh LEE 3, Mei-Hsiang CHEN 4, 5, Ching-Lin HSIEH 2, 3, 6 * 1Department of Physical Medicine and Rehabilitation, Chi Mei Medical Center, Tainan, Taiwan; 2Department of Occupational Therapy, College of Medical and Health Science, Asia University, Taichung, Taiwan; 3School of Occupational Therapy, College of Medicine, National Taiwan University, Taipei, Taiwan; 4School of Occupational Therapy, Chung Shan Medical University, Taichung, Taiwan; 5Occupational Therapy Room, Chung Shan Medical University Hospital, Taichung, Taiwan; 6Department of Physical Medicine and Rehabilitation, National Taiwan University Hospital, Taipei, Taiwan *Corresponding authors: Ching-Lin Hsieh, School of Occupational Therapy, College of Medicine, National Taiwan University, Taipei, Taiwan. E-mail: [email protected] ABSTRACT BACKGROUND: The responsiveness and predictive validity of the Tablet-based Symbol Digit Modalities Test (T-SDMT) are unknown, which limits the utility of the T-SDMT in both clinical and research settings. AIM: The purpose of this study was to examine the responsiveness and predictive validity of the T-SDMT in inpatients with stroke. DESIGN: A follow-up, repeated-assessments design. SETTING: One rehabilitation unit at a local medical center. POPULATION: A total of 50 inpatients receiving rehabilitation completed T-SDMT assessments at admission to and discharge from a rehabilita- tion ward. METHODS: The median follow-up period was 14 days. The Barthel index (BI) was assessed at discharge and was used as the criterion of the predictive validity. RESULTS: The mean changes in the T-SDMT scores between admission and discharge were statistically significant (paired t-test = 3.46, P=0.001). The T-SDMT scores showed a nearly moderate standardized response mean (0.49). A moderate association (Pearson’s r =0.47) was found between the scores of the T-SDMT at admission and those of the BI at discharge, indicating good predictive validity of the T-SDMT. CONCLUSIONS: Our results support the responsiveness and predictive validity of the T-SDMT in patients with stroke receiving rehabilitation in hospitals. CLINICAL REHABILITATION IMPACT: This study provides empirical evidence supporting the use of the T-SDMT as an outcome measure for assessing processing speed in inpatients with stroke. The scores of the T-SDMT could be used to predict basic activities of daily living function in inpatients with stroke. (Cite this article as: Hsiao PC, Yu WH, Lee SC, Chen MH, Hsieh CL. Responsiveness and predictive validity of the Tablet-based Symbol Digit Mo- dalities Test in patients with stroke. Eur J Phys Rehabil Med 2019;55:29-34. DOI: 10.23736/S1973-9087.18.05210-3) Key words: Predictive value of tests - Stroke - Rehabilitation. Deficit in processing speed (PS) is frequent and severe SDMT for assessing PS in patients with stroke.8 The T- in patients with stroke.1-3 Up to 50-70% of patients SDMT has two main merits. First, the T-SDMT is easy with stroke may experience a noticeably slowed PS.4, 5 for raters and examinees to use. Specifically, the T-SDMT Deficit in PS can hamper functional recovery and quality provides animated oral instructions and thus, raters need of life in patients with stroke.6, 7 To manage patients’ defi- little time for training. Such animated instructions were cits in PS, clinicians and researchers need a sound measure proven clarity for users.8 Thus, examinees can easily un- to monitor patients’ PS problems and changes. derstand the rules of testing. Second, the T-SDMT spon- taneously records examinees’ responses, completes scor- The Tablet-based Symbol Digit Modalities Test (T-SD- ing, and stores data. Such spontaneous scoring and data MT) was constructed on the basis of the commonly used Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 29

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, HSIAO T-SDMT IN PATIENTS WITH STROKE cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. management are particularly useful for time-pressed clini- programs (e.g., physiotherapy, occupational therapy, and/ cians, as they reduce the burden on the clinicians. These or speech therapy) 3 to 5 times every week. merits support the T-SDMT as a clinically useful measure of PS in patients with stroke. This study protocol was approved by the Research Eth- ics Committee affiliated with the medical center. Informed Some sufficient psychometric properties (e.g., test-re- consent was obtained from every participant before data test reliability, concurrent validity, and ecological valid- collection. ity) of the T-SDMT have been reported in patients with stroke8 and in patients with schizophrenia.9 However, the Procedure responsiveness and predictive validity of the T-SDMT are still unknown. The responsiveness of a PS measure is use- A research assistant (an occupational therapist) screened ful to present its ability to detect (monitor) PS change over the patients and administered the T-SDMT and the Barthel time.10, 11 Thus, whether the T-SDMT is a good outcome Index (BI) to the patients. The patients were assessed at measure to detect change in PS is unknown. Moreover, admission to and discharge from the ward. At each assess- the predictive validity is critical for a PS measure to dem- ment, the T-SDMT was administered 3 times, with suf- onstrate its ability to predict future functional outcomes ficient rest between adjacent assessments. All assessments (e.g., activities of daily living).10, 11 However, whether the were administered in a quiet room to prevent interruptions. scores of the T-SDMT are useful for predicting functional outcomes has not been examined. The unknown respon- Measures siveness and predictive validity of the T-SDMT limit the utility of the T-SDMT in clinical and research settings. T-SDMT Thus, the purpose of this study was to examine the respon- siveness and predictive validity of the T-SDMT in patients The T-SDMT has 9 symbols, each corresponding to a spe- with stroke. We hypothesized that the T-SDMT would cific Arabic numeral (1-9).8 The pairs of symbols and nu- have sufficient responsiveness and predictive validity in merals are randomly assigned each time when the stimu- patients with stroke. In addition, we examined whether re- lus is presented to the examinee on a tablet computer (i.e., peated assessments with the T-SDMT would improve the iPad) screen (Figure 1). Such a design is used to avoid a responsiveness of the measure. We hypothesized that the possible memory effect. T-SDMT would have better responsiveness with 3 assess- ments than with one or two assessments. All trials are administered using the iPad in landscape orientation. The iPad is mounted at an angle of 30º in a case. To take the test, the examinee is asked to first look at Materials and methods Participants We recruited patients with stroke who were consecutively Figure 1.—The design of the Tablet-based Symbol Digit Modalities admitted to the rehabilitation ward of a medical center in Test. Taipei, Taiwan. Inpatients were invited to participate in To respond, the examinees first look at the stimulus (e.g., square) in the this study if they met the following criteria: 1) a diagnosis middle of the screen, then look up to search for the numeral (6) cor- of ischemic stroke or cerebral hemorrhage; 2) age of at responding to the stimulus in the reference key, and finally touch the least 20 years; 3) hemiparesis or hemiplegia; and 4) abil- numeral in the 3-by-3 grid. ity to follow 3-step instructions for assessments. However, patients were excluded if they met the following crite- ria: 1) having other neurological or mental diseases (e.g., brain tumor or schizophrenia) that might influence cogni- tive function; 2) failure to complete the T-SDMT with the less affected hand or due to vision or hearing problems; 3) interval between admission and discharge assessments of less than a week; and 4) recurrence of stroke onset in the study period. All patients received regular rehabilitation 30 European Journal of Physical and Rehabilitation Medicine February 2019

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, T-SDMT IN PATIENTS WITH STROKE HSIAO cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. the symbol shown in the middle of the iPad screen, then to were considered to indicate sufficient responsiveness of search for the corresponding number in the table at the top the T-SDMT. of the screen, and last to use his/her index finger to touch the corresponding Arabic numeral on a 3-by-3 grid at the We further examined whether the number of assess- bottom of the screen. ments with the T-SDMT would improve the responsive- ness of the measure. The score of the first assessment, Before the formal test, the examinee is required to mean score of the first and second assessments, and mean complete a practice session. He/she has to respond to 5 score of 3 assessments of the T-SDMT were used, respec- practice items (with feedback showing correct or incorrect tively. The bootstrap approach was used to compare statis- responses provided by the iPad) and complete a 60-second tical differences in the responsiveness between the scores simulated test. Then a formal 90-second test is adminis- of the numbers of assessments.18-20 We drew 10,000 boot- tered. The iPad automatically records the number of cor- strap samples, each equal in size to the number of patients rect responses during the testing. A higher number of cor- observed. For each of the 10,000 samples, we calculated rect responses indicates better PS. the SRMs for different numbers of testing. Then pairwise differences of the SRMs of each sample were calculated. BI We examined whether zero was included in the 95% boot- strap percentile confidence intervals (CIs) of differences. The BI assesses patients’ performance in daily function- If zero was not included, the responsiveness of different ing (or disability).12 It contains 10 basic activities of daily numbers of assessments was considered significantly dif- living (ADL) tasks (e.g., transfers, grooming, toileting, ferent. bathing, feeding, and dressing). These tasks are rated on a 2-point (0, 1), 3-point (0, 1, 2) or 4-point (0, 1, 2, 3) Likert- Predictive validity type scale. The total score ranges from 0 to 20. A score of 0 indicates serious disability, and a score of 20 indicates Pearson’s r was used to investigate the association between no disability. The BI has sound psychometric properties the scores of the T-SDMT at admission and those of the BI in patients with stroke.13, 14 We used the score of the BI at discharge. The scores of the first assessments with the as the criterion to investigate the predictive validity of the T-SDMT were used. We hypothesized a moderate associa- T-SDMT. tion (Pearson’s r =0.40-0.60) between the two scores.21 A moderate association was considered to indicate sufficient Statistical analysis predictive validity of the T-SDMT. We first used descriptive analyses to present the charac- Results teristics of the patients. All statistical analyses were per- formed in IBM SPSS Statistics (Version 22; IBM Corp., Table I presents the sociodemographic and clinical char- Armonk, NY, USA). acteristics of the inpatients who completed the assess- ments at admission or discharge. Fifty-eight patients Responsiveness were originally recruited from October 2015 to January 2017.—Eight patients did not complete the discharge as- We first used a paired t-test to examine whether the sessments because they were discharged early without changes in scores on the T-SDMT reached statistical sig- notice. Thus, 50 patients completed both assessments and nificance. We assumed that the PS of the inpatients would they had, on average, moderate disability (mean [SD] BI significantly improve through natural recovery and tradi- Score = 13.5 [4.2]). The patients who did not complete tional rehabilitation during their hospital stay. The stan- the study were not significantly different from those who dardized response mean (SRM) was used to examine the completed the study in terms of demographic character- responsiveness of the T-SDMT between admission to and istics (i.e., sex, age, education status) (P≥0.22) or clinical discharge from the rehabilitation ward. SRM was calcu- characteristics (i.e., stroke type, side of hemiplegia, time lated by dividing the mean of the change scores by the since onset to initial evaluation, the first T-SDMT scores, standard deviation of the change. SRMs of 0.20-0.49, and BI scores at admission) (P≥0.05). However, it is noted 0.50-0.79, and ≥0.8 were considered to indicate small, that the number of male patients, who completed both as- moderate, and large responsiveness, respectively.15-17 sessments, was about twice that of female patients. The We hypothesized that the change scores of the T-SDMT patients with left hemiplegia/hemiparesis were more likely would have at least small responsiveness. Such results Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 31

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, HSIAO T-SDMT IN PATIENTS WITH STROKE cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. Table I.—Characteristics of the participants in the study. However, the 10,000 bootstrap samples indicated that the pairwise differences of the SRMs between different num- Characteristic Participants who Participants who bers of assessments were not significant (all 95% CI of completed the withdrew from differences of the SRMs included 0) (Table II). study (N.=50) the study (N.=8) Predictive validity Sex, N. (%) 33 (66.0%) 7 (87.5%) 17 (34.0%) 1 (12.5%) The scores of the T-SDMT at admission were significantly Male 57.6 (13.6) 63.4 (14.8) associated with those of the BI at discharge (Pearson’s r Female =0.47, P<0.001). The level of association was moderate. 7 (14.0%) 0 (0.0%) Age, years, mean (SD) 6 (12.0%) 1 (12.5%) Discussion 23 (46.0%) 3 (37.5%) Education status, N. (%) 12 (24.0%) 4 (50.0%) We found that the values of SRM of the T-SDMT were 2 (4.0%) 0 (0.0%) around moderate (0.49-0.58) for all different numbers (1- Elementary 3) of assessments. These results indicate that the T-SDMT Middle school 13 (26.0%) 2 (25.0%) could show nearly moderate change in PS in patients with High school or vocational high 37 (74.0%) 6 (75.0%) stroke receiving inpatient rehabilitation. A possible reason for the moderate change in PS is that our participants only re- school 15 (68.0%) 3 (37.5%) ceived inpatient rehabilitation for about one to three weeks. University 34 (30.0%) 5 (62.5%) The limited periods of follow-up may have prevented large Graduate school 0 (0%) amounts of change in PS in these participants. Thus, their 1 (2.0%) 24.5 (5-80) moderate improvement in PS seems reasonable. These ob- Stroke type, N. (%) 23 (7-85) servations support the responsiveness of the T-SDMT and 15.9 (8.8) the ability of the T-SDMT to be used an outcome measure Cerebral hemorrhage 25.3 (13.8) to detect change in PS in patients with stroke. Cerebral infarction 13.0 (1.7) 13.5 (4.2) Our results also showed non-significant pairwise dif- Side of hemiplegia, N. (%) ferences in the SRM of the T-SDMT between single and multiple assessments. These results indicate that the re- Right sponsiveness of the T-SDMT was similar regardless of the Left number of assessments. This finding does not support our Bilateral hypothesis that three assessments would render the most responsive outcome as compared to one and two assess- Time since onset to initial ments. In theory, the mean score of multiple assessments evaluation, days, median can reduce random measurement error and thus should im- (minimum-maximum) prove the responsiveness of a measure.22, 23 Our results do not support such a theory. Two reasons may explain our The first T-SDMT scores at results. First, our sample size was medium, so the effect of admission, mean (SD) multiple assessments in reducing random measurement er- ror might have been compromised. Second, the magnitude BI scores at admission, mean (SD) T-SDMT: Tablet-based Symbol Digit Modalities Test; BI: Barthel Index. to drop out. In addition, the drop-out patients’ T-SDMT scores at admission had a trend to be lower than those of the patients completing both assessments. Responsiveness Table II presents the results of the responsiveness of the T-SDMT using 1-3 assessments. The mean changes (first assessments of the patients) in the T-SDMT’s scores be- tween admission and discharge were statistically signifi- cant (paired t=3.46, P=0.001). The T-SDMT scores (first assessments) had a nearly moderate SRM (0.49). The T- SDMT scores (the mean scores of the first 2 or all 3 as- sessments) showed slightly larger SRMs (0.57-0.58). Table II.—C omparison of responsiveness of one-, two-, and three assessments of the T-SDMT (N.=50). Admission score Discharge score Difference score Paired t-test (P) SRM Mean (SD) Mean (SD) Mean (SD) (95% CI)* One assessment 25.3 (12.8) 28.5 (15.1) 3.2 (6.6) 3.46 (0.001) 0.49 (0.24-0.75) The first two assessments 25.5 (12.9) 28.8 (14.9) 3.3 (5.7) 4.06 (<0.001) 0.57 (0.39-0.85) Three assessments 25.7 (13.0) 28.9 (14.8) 3.3 (5.6) 4.10 (<0.001) 0.58 (0.36-0.89) Difference score: discharge score-admission score; SRM: standardized response mean. *The Bootstrap results showed that the 95% CIs for the SRMs of the 3 kinds of assessments overlapped each other (mean difference in SRM [95% CI] of one time vs. two times, one time vs. three times, and two times vs. three times were -0.08 [-0.27, 0.06], -0.09 [-0.34, 0.07], and -0.01 [-0.11, 0.07]). 32 European Journal of Physical and Rehabilitation Medicine February 2019

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, T-SDMT IN PATIENTS WITH STROKE HSIAO cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. of random measurement error of the T-SDMT may be so ed. Multiple (i.e., two and three) assessments did not sig- small that the effect of multiple assessments in reducing nificantly improve the responsiveness of the T-SDMT, so random measurement error was limited. Further studies a single assessment is suggested for prospective users. The recruiting large sample sizes (e.g., 100) might be needed scores of the T-SDMT seem useful for predicting basic ac- to confirm our findings. tivities of daily living function in inpatients with stroke. We found moderate association (Pearson’s r=0.47) be- References tween the scores of the T-SDMT at admission and those of the BI at discharge. This result supports our hypothesis 1.  Ballard C, Stephens S, Kenny R, Kalaria R, Tovee M, O’Brien J. Pro- and indicates that the T-SDMT has sufficient predictive file of neuropsychological deficits in older stroke survivors without de- validity. A measure with sufficient predictive validity can mentia. Dement Geriatr Cogn Disord 2003;16:52–6. be used by clinicians as their reference to determine appro- 2.  Brand C, Alber B, Fladung AK, Knauer K, König R, Oechsner A, et priate rehabilitation interventions and discharge plans for al. Cognitive performance following spontaneous subarachnoid haem- patients with stroke. Thus, our results support the predic- orrhage versus other forms of intracranial haemorrhage. Br J Neurosurg tive utility of the T-SDMT in clinical settings. 2014;28:68–80. 3.  Douven E, Köhler S, Schievink SH, van Oostenbrugge RJ, Staals J, For further investigation of predictive validity, re- Verhey FR, et al. Baseline Vascular Cognitive Impairment Predicts the searchers may use the scores of the T-SDMT to validate Course of Apathetic Symptoms After Stroke: the CASPER Study. Am J or construct a model predicting disability. In this study, Geriatr Psychiatry 2018;26:291–300. we did not control for the possible covariates (e.g., pos- 4.  Hochstenbach J, Mulder T, van Limbeek J, Donders R, Schoonderwaldt tural control and motor function) when we examined the H. Cognitive decline following stroke: a comprehensive study of cogni- association between the scores of the T-SDMT and the BI. tive decline following stroke. J Clin Exp Neuropsychol 1998;20:503–17. The covariates can also contribute to predictions of dis- 5.  Rasquin SM, Lodder J, Ponds RW, Winkens I, Jolles J, Verhey FR. ability of the patients.24-26 Further studies controlling for Cognitive functioning after stroke: a one-year follow-up study. Dement covariates may be needed to further confirm the predictive Geriatr Cogn Disord 2004;18:138–44. validity of the T-SDMT. 6.  Barker-Collo S, Feigin VL, Parag V, Lawes CM, Senior H. Auckland Stroke Outcomes Study. Part 2: cognition and functional outcomes 5 years Limitations of the study poststroke. Neurology 2010;75:1608–16. 7.  Narasimhalu K, Ang S, De Silva DA, Wong MC, Chang HM, Chia This study has three limitations. First, we only recruited KS, et al. The prognostic effects of poststroke cognitive impairment no patients in the subacute stage who were receiving hospital dementia and domain-specific cognitive impairments in nondisabled isch- rehabilitation. Second, our sample was biased. The number emic stroke patients. Stroke 2011;42:883–8. of male patients was about twice that of female patients. 8.  Tung LC, Yu WH, Lin GH, Yu TY, Wu CT, Tsai CY, et al. Develop- Patients with severe disability were rarely recruited. More- ment of a Tablet-based symbol digit modalities test for reliably assess- over, the patients with left hemiplegia/hemiparesis were ing information processing speed in patients with stroke. Disabil Rehabil more likely to drop out, and the drop-out patients’ first T- 2016;38:1952–60. SDMT scores at admission was lower than those of the 9.  Tang SF, Chen IH, Chiang HY, Wu CT, Hsueh IP, Yu WH, et al. A patients completing both assessments (Table I). Third, the comparison between the original and Tablet-based Symbol Digit Mo- follow-up periods (lengths of stay at rehabilitation wards) dalities Test in patients with schizophrenia: test-retest agreement, random were short (median =14 days). These limitations might measurement error, practice effect, and ecological validity. Psychiatry Res have limited the generalizability of the findings of respon- 2018;260:199–206. siveness and predictive validity of the T-SDMT. Further 10.  Mokkink LB, Terwee CB, Patrick DL, Alonso J, Stratford PW, Knol validation (particularly on patients at acute or chronic pa- DL, et al. The COSMIN checklist for assessing the methodological qual- tients, patients with severe disability, and female patients) ity of studies on measurement properties of health status measurement in- of our findings is warranted. struments: an international Delphi study. Qual Life Res 2010;19:539–49. 11.  Mokkink LB, Terwee CB, Patrick DL, Alonso J, Stratford PW, Knol Conclusions DL, et al. The COSMIN study reached international consensus on taxon- omy, terminology, and definitions of measurement properties for health- Our results support the responsiveness and predictive va- related patient-reported outcomes. J Clin Epidemiol 2010;63:737–45. lidity of the T-SDMT in patients with stroke receiving re- 12.  Mahoney FI, Barthel DW. Functional Evaluation: The Barthel Index. habilitation in hospitals. The utility of the T-SDMT as an Md State Med J 1965;14:61–5. outcome measure for assessing PS is preliminarily validat- 13.  Hsueh IP, Lee MM, Hsieh CL. Psychometric characteristics of the Barthel activities of daily living index in stroke patients. J Formos Med Assoc 2001;100:526–32. 14.  Hsueh IP, Lin JH, Jeng JS, Hsieh CL. Comparison of the psychomet- ric characteristics of the functional independence measure, 5 item Barthel index, and 10 item Barthel index in patients with stroke. J Neurol Neuro- surg Psychiatry 2002;73:188–90. 15.  Husted JA, Cook RJ, Farewell VT, Gladman DD. Methods for assess- ing responsiveness: a critical review and recommendations. J Clin Epide- miol 2000;53:459–68. Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 33

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, HSIAO T-SDMT IN PATIENTS WITH STROKE cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. 16.  Chen KL, Chen CT, Chou YT, Shih CL, Koh CL, Hsieh CL. Is 21.  Portney LG, Watkin MP. Foundations of clinical research: applica- the long form of the Fugl-Meyer motor scale more responsive than the tions to practice. Third edition. Upper Saddle River, NJ: Pearson Prentice short form in patients with stroke? Arch Phys Med Rehabil 2014;95: Hall; 2009. 941–9. 22.  Lu WS, Huang SL, Hsieh CL. Several trials are useful to re- 17.  Hsieh YW, Lin JH, Wang CH, Sheu CF, Hsueh IP, Hsieh CL. Dis- duce the value of minimal detectable change. Arch Phys Med Rehabil criminative, predictive and evaluative properties of the simplified stroke 2009;90:181–2, author reply 182–3. rehabilitation assessment of movement instrument in patients with stroke. 23.  Lui J, MacGillivray MK, Sawatzky BJ. Test-retest reliability and J Rehabil Med 2007;39:454–60. minimal detectable change of the SmartWheel clinical protocol. Arch 18.  Efron B. Bootstrap Methods: Another Look at the Jackknife. Ann Stat Phys Med Rehabil 2012;93:2367–72. 1979;7:1–26. 24.  Aprile I, Piazzini DB, Bertolini C, Caliandro P, Pazzaglia C, Tonali P, 19.  Huang YJ, Lin GH, Lee SC, Chen YM, Huang SL, Hsieh CL. Group- et al. Predictive variables on disability and quality of life in stroke outpa- and Individual-Level Responsiveness of the 3-Point Berg Balance Scale tients undergoing rehabilitation. Neurol Sci 2006;27:40–6. and 3-Point Postural Assessment Scale for Stroke Patients. Arch Phys 25.  Kwakkel G, Wagenaar RC, Kollen BJ, Lankhorst GJ. Predict- Med Rehabil 2018;99:529–33. ing disability in stroke—a critical review of the literature. Age Ageing 20.  Huang YJ, Chen KL, Chou YT, Hsueh IP, Hou CY, Hsieh CL. Com- 1996;25:479–89. parison of the Responsiveness of the Long-Form and Simplified Stroke 26.  Hsieh CL, Sheu CF, Hsueh IP, Wang CH. Trunk control as an early Rehabilitation Assessment of Movement: Group- and Individual-Level predictor of comprehensive activities of daily living function in stroke Analysis. Phys Ther 2015;95:1172–83. patients. Stroke 2002;33:2626–30. Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material discussed in the manuscript. Funding.—Chung Shan Medical University and Chi Mei Medical Center (CSMU-CMMC-104-06 and CMCSMU10402) and the Ministry of Science and Technology in Taiwan (106-2314-B-002-252-MY3 and 107-2314-B-002-046-MY3). Authors’ contributions.—Mei-Hsiang Chen and Ching-Lin Hsieh equally contributed to this work. Every author has made final approval of the manuscript. Pei-Chi Hsiao: drafting the article; Wan-Hui Yu: data analysis and interpretation; Shih-Chieh Lee and Mei-Hsiang Chen: critical revision of the article; Ching- Lin Hsieh: conception or design of the work. Article first published online: June 14, 2018. - Manuscript accepted: June 13, 2018. - Manuscript revised: May 23, 2018. - Manuscript received: January 29, 2018. 34 European Journal of Physical and Rehabilitation Medicine February 2019

COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically © 2018 EDIZIONI MINERVA MEDICA European Journal of Physical and Rehabilitation Medicine 2019 February;55(1):35-9 or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access Online version at http://www.minervamedica.it DOI: 10.23736/S1973-9087.18.05202-4 to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. ORIGINAL ARTICLE Effect of turning direction on Timed Up and Go test results in stroke patients Hohee SON, Chanhyun PARK * Department of Physical Therapy, College of Health Science, Catholic University of Pusan, Pusan, South Korea *Corresponding author: Chanhyun Park, Department of Physical Therapy, College of Health Science, Catholic University of Pusan, 57 Oryundaero Geum- jeonggu, Pusan, South Korea. E-mail: [email protected] ABSTRACT BACKGROUND: The Timed Up and Go (TUG) test is an assessment tool for measuring mobility in stroke patients. In stroke patients, the turn- ing direction of the affected and unaffected sides may influence turning time. AIM: The aim of this study is to investigate the effects of the turning direction according to the affected and unaffected sides of stroke patients during their Timed Up and Go (TUG) test and to define clinically salient outcomes during TUG tests performed in the clinic. DESIGN: Observational design. SETTING: Department of Physical Therapy in a rehabilitation center. POPULATION: One hundred thirteen hemiparetic stroke patients. METHODS: Stroke patients were asked to perform the TUG test by turning toward their affected and unaffected sides. Patients were divided ac- cording to gait speed, with their gait speed from the 10mWT being used. Those with a gait speed <48 m/min were assigned to the severe ambula- tory dysfunction (SAD) group, whereas those with a gait speed ≥48 m/min were assigned to the moderate ambulatory dysfunction (MAD) group. RESULTS: The TUG test results showed a longer turning time when turning with the unaffected side as the turning axis (17.10±5.69 s) than with the affected side as the turning axis (17.52±5.90 s). When the patients were divided into the MAD and SAD groups based on the 10mWT results, patients in the SAD group exhibited slightly longer times (0.55±1.11 s) than those in the MAD group (0.29±1.03 s); however, this dif- ference was not significant. CONCLUSIONS: The present study found that stroke patients showed differences in the TUG test results based on their turning direction, and less time was required when turning in the direction of the affected side than when turning in the direction of the unaffected side. CLINICAL REHABILITATION IMPACT: Turning direction can affect the results of the TUG test; it should be controlled in the execution of the TUG test in clinical settings. (Cite this article as: Son H, Park C. Effect of turning direction on Timed Up and Go test results in stroke patients. Eur J Phys Rehabil Med 2019;55:35-9. DOI: 10.23736/S1973-9087.18.05202-4) Key words: Walking - Mobility limitation - Gait - Stroke. For stroke patients, the recovery of physical functions of steps, which increases the risk of falling.3 In particular, is the most important goal in rehabilitation training. stroke patients have asymmetrical posture and movement Among physical functions, the recovery of gait-related due to paralysis of one side of the body, which has a great- functions is necessary for patients’ mobility and perfor- er impact on their ability to turn in a standing posture.4 The mance of activities of daily living (ADL); thus, gait train- ability to turn in a standing posture is closely associated ing should focus on gait speed and include all functional with gait ability.5 elements of gait. Functional elements of gait include straight-line walking and turning direction.1 However, The Timed Up and Go (TUG) test, an assessment tool many situations require turning while standing during for measuring mobility, is comprised of standing from a ADL, such as in a narrow bathroom or kitchen.2 chair, walking in a straight line for 3 m, and turning at a designated turning point.6 It is a simple and useful test for Turning while standing requires axial rotation of one examining gait and balance abilities, and it is widely used leg as well as a longer time to turn and a greater number in patients with impaired mobility due to frailty,7 stroke,8 Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 35

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, SON TUG TEST RESULTS IN STROKE PATIENTS cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. and Parkinson’s disease.9 This test is also used for assess- Assessed for eligibility ing fall risks or measuring the outcomes of interventions, (N.=138) as it has been shown to have excellent validity and reli- ability.10 Excluded (N.=10) did not meet inclusion criteria The setup for the TUG test is specific, requiring the use of a chair with an armrest and having its height set to 43 Assessed for eligibility cm.10 However, the test does not specify whether the turn- (N.=128) ing must be performed by the leg on the affected or unaf- fected side, which allows the subject to turn in the direc- 10-m walking test tion that he or she is more comfortable. A review of studies to date on the turning direction in the TUG test showed Excluded (N.=15) equivocal findings.8, 11 In 2009, Heung and Ng found that - withdrawn informed consent the height of the chair as well as the turning direction dur- ing the TUG test were key factors that had an impact on (N.=10) the results in stroke patients.11 A study by Faria et al. that - failed 10-m WT (N.=5) assessed the effects of turning direction during the TUG test in stroke patients indicated that the fear of falling had Moderate ambulatory Severe ambulatory a bigger impact than the turning direction.8 dysfunction group dysfunction group (N.=58) Accordingly, the objective of the present study was to (N.=55) investigate the effects of the turning direction according to the affected and unaffected sides of stroke patients on Timed Up and Go test their TUG test results in order to use the findings as mea- surement variables during TUG tests performed in clinical Analyzed (N.=58) Analyzed (N.=55) settings. Moreover, this study aimed to use the scores from the 10-m walking test (10mWT) to investigate the effects Figure 1.—Flow chart of patient participation in the study. of differences in gait ability on the turning direction in the TUG test. The first hypothesis was that there would be dif- whereas those with a gait speed ≥48 m/min were assigned ferences in the TUG test results when hemiplegic patients to the moderate ambulatory dysfunction (MAD) group used the leg on the affected side as the turning axis versus (Figure 1).12-14 using the leg on the unaffected side. The second hypoth- esis in this study was that differences in gait speed would This study was approved by the ethics committee of the have an effect on the turning direction during the TUG test. Catholic University of Pusan. The aims and procedures were explained to all participants before they provided in- Materials and methods formed consent. One hundred and thirteen hemiparetic stroke patients par- Measures ticipated in this study. The inclusion criteria for the present analysis were as follows: 1) patients diagnosed by a phy- The 10mWT sician as having had a stroke; 2) those who were able to communicate and follow simple commands; and 3) those A 14-m long straight walking path was created, and ex- who were able to walk at least 10 m independently (with cluding the first and last 2 m for acceleration and decel- or without walking aids); and 4) had a score of less than 2 eration, the middle 10 m was used for measuring the time points on the modified Ashworth Scale. Participants who to perform the task. The values from three measurements 1) had a Mini-Mental State Examination score <24; 2) had were used to derive the mean value, and a 30-second rest a visual or vestibular disorder; or 3) had an orthopedic dis- period was given in between the measurements. ease were excluded. Timed Up and Go test Patients were divided according to gait speed on the 10mWT. Those with a gait speed <48 m/min were as- While sitting in a chair with an armrest, the patient stood signed to the severe ambulatory dysfunction (SAD) group, up from the chair when the examiner gave the verbal 36 European Journal of Physical and Rehabilitation Medicine February 2019

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, TUG TEST RESULTS IN STROKE PATIENTS SON cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. command to “start.” After standing up from the chair, the Results patient walked 3 m to the turning point, turned, returned to the chair, and sat down. The total time to perform this The demographic characteristics of the 113 participants task was measured. In the present study, the patient was are shown in Table I. The TUG test and 10mWT results required to turn at the turning point using his or her leg on showed excellent test-retest reliability in this study. ICCs the affected side as the turning axis in three trials and the (2,1) for the TUG test and the 10mWT were 0.97 and 0.96, leg on the unaffected side as the turning axis in another set respectively (Table II). of three trials. The measurements from the three trials for each side were used to derive the mean value. The order The TUG test results showed a longer turning time of using the affected side or the unaffected side first was when turning with the unaffected side as the turning determined by the flip of a coin, and the six trials were per- axis than with the affected side as the turning axis (Ta- formed by switching from one side to the other after each ble III). Moreover, when the patients were divided into trial. Each patient was given a 30-second rest period in be- the MAD and SAD groups based on the 10mWT results, tween the measurements, and the chair used in the test had both groups showed a longer turning time in the TUG a height of 43 cm, an armrest height of 63 cm, a backrest. test results when turning with their unaffected side as the turning axis. The results were slightly longer in the SAD Statistical analysis group than in the MAD group, but the difference was not significant (Table IV). Sample size was determined using G*Power 3.1.9 soft- Table II.—T est-retest reliability of TUG and 10mWT in partici- ware and was based on the study by previously published pants. data.8, 15 The power and the alpha level were set as 0.8 and 0.05, respectively. And the effect size was set at 0.55, Variables 95% CI the calculated sample size was at least 53 subjects in each ICC P value group. Lower Upper Descriptive statistics were used to describe the demo- graphic characteristics of the participants. Chi-square tests TUG (turn affected side) 0.97 0.96 0.98 0.00 and independent t-tests were used to compare demograph- ic characteristics and the 10mWT results. To determine TUG (turn unaffected side) 0.97 0.96 0.98 0.00 the test-retest reliability of outcome measures for the TUG test, intraclass correlation coefficients (ICCs) were used. 10mWT 0.96 0.98 0.99 0.00 To determine the difference between trials and between groups of the TUG test, the paired t-test and the indepen- TUG: Timed Up and Go; 10mWT: 10-minute walking test; ICC: intraclass dent t-test were used. All data were analyzed using SPSS correlation coefficients; CI: confidence interval. Statistical Analysis software v. 22 (IBM Corp., Armonk, NY, USA) with a significance level of P<0.05. Table III.—Results of TUG test to turning direction in all partici- pants. Variables Affected side Unaffected side t P value 17.52±5.90 -4.14 0.00 TUG test, s 17.10±5.69 TUG: Timed Up and Go. Table I.—D emographic characteristics of participants including 10mWT scores. Variables Total MAD group SAD group P value (N.=113) (N.=58) (N.=55) (χ2) Age, years 53.34±7.34 60.24±7.70 58.38±6.88 0.18 Height, cm 164.74±7.50 164.19±7.64 165.33±7.38 0.42 Weight, kg 62.84±10.56 0.90 Disease duration, months 62.96±10.06 63.09±9.65 12.75±5.61 0.12 Sex, M/F 11.79±6.39 10.88±6.98 0.83 Type of stroke: ischemic/hemorrhage 35/20 0.05 Modified Ashworth Scale 73/40 38/20 23/32 0.07 Affected side, left/right 58/55 35/23 0.31 10mWT, s 0.38 0.51 0.00 52/61 0.26 28/27 14.27±5.92 24/34 19.24±4.63 9.55±1.44 Values are mean±SD or number. 10mWT: 10-minute Walking Test; MAD: moderate ambulatory dysfunction (<48 m/min walking speed); SAD: severe ambulatory dysfunction (≥48 m/min walking speed). Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 37

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, SON TUG TEST RESULTS IN STROKE PATIENTS cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. Table IV.—Results of TUG test to turning direction for MAD and SAD groups. Group Affected side Unaffected side Mean differences t P value MAD group (N.=58) 12.62±2.83 12.91±3.17 0.29±1.03 -2.15 0.04 SAD group (N.=55) 21.83±3.78 22.38±3.83 0.55±1.11 -3.68 0.00 t P value -14.70 -14.32 -1.30 0.00 0.00 0.20 TUG: Timed Up and Go; MAD: moderate ambulatory dysfunction (<48 m/min walking speed); SAD: severe ambulatory dysfunction (≥48 m/min walking speed). Discussion various patients with impaired mobility.19, 20 This test mea- sures the gait speed of a subject walking a short, straight Numerous studies have reported that the number of strides 10-m line comfortably or as fast as possible,21 and since and the time required increases when turning during gait, extensive effort is needed within a short period, this test and hemiplegic patients require more time for turning than also assesses muscle strength.22 However, functional as- other elderly patients of the same age.4, 5, 16 Although such sessment of stroke patients requires assessment of not only a difference may be attributable to various reasons, such gait speed but balance and functional mobility as well. The as asymmetric gait and weakened leg muscles, it is also TUG test is an assessment tool for measuring mobility, affected by the fear of falling.8 Turning is a motion com- balance, and risk of falling.23 This present study classified monly performed and required in ADL; however, accord- the subjects into the severe ambulatory dysfunction group ing to previous studies, the risk of falling is 7.9-fold higher and the moderate ambulatory dysfunction group based on when turning than when walking in a straight line.17 The a 10mWT test result of 48 m/min. It was found that it took TUG test, which is often used in clinical settings since it both groups significantly less time in a TUG test to turn in can easily assess walking in a straight line and turning si- the affected leg direction; while there was no significant multaneously, is comprised of sitting and standing from a differences between the groups, it took the severe ambula- chair, walking in a straight line, and turning 180°. tory dysfunction group longer to turn in the unaffected leg direction. This would be attributable to their fear of falling, The TUG test, conducted in 113 hemiplegic patients, as their gait ability was low. There was no significant differ- showed that less time was required when turning using the ence because although this study’s severe ambulatory dys- leg on the affected side, which was consistent with the re- function group comprised chronic stroke patients 6 months sults from a previous study.11 When turning, the leg on the post stroke, their physical function had recovered relatively side to which the person is turning is used as the axis, and better than that of patients in previous studies, as it took the number of steps taken by the opposite leg increases; in them only 22 seconds on average to complete a TUG test. hemiplegic patients, movement in the swing phase of the unaffected leg may be easier when turning with the leg on What is worth noting is that this study included 115 sub- the affected side. On the other hand, in 2015, a study by jects, more than other previous studies examining turning Kobayashi et al.5 reported that the time required to turn in directions in stroke patients’ TUG test, and the subjects the direction of the unaffected side was shorter. However, were divided based on the gait speed at which they could the results in the present study showed that approximately perform community ambulation. In addition, many TUG 17 seconds were needed to complete the TUG test, and 14 studies do not specify a turning direction, and accordingly, seconds were required to complete the 10mWT, whereas researchers select the turning direction of test at their dis- in the study by Kobayashi et al., approximately 35.5 sec- cretion. This study’s findings demonstrate that the turning onds were required to complete the TUG test and the gait direction had an effect on the TUG outcome. speed was 0.48 m/s on the 10mWT. It is believed that this discrepancy may be due to differences in the lower ex- In addition, if a patient understands in which direction tremity function of the study subjects. it would be easier to turn when standing and applies this to his or her daily life, the patient could then move more Gait speed is one of the most basic elements of gait func- efficiently with less fatigue. tion assessment, and the gait speed for community ambula- tion is generally considered to be 54 m/min.18 In clinical Limitations of the study practice, various assessment tools are used to easily and simply evaluate gait-related physical functions. Among Limitations in the present study include the fact that the them, the 10mWT for measuring gait speed is a tool used in stroke patients were relatively active, showing a mean 38 European Journal of Physical and Rehabilitation Medicine February 2019

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, TUG TEST RESULTS IN STROKE PATIENTS SON cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. TUG test time of 17 seconds. Moreover, only gait speed 8.  Faria CD, Teixeira-Salmela LF, Nadeau S. Effects of the direction of and the presence of paralysis were analyzed in the context turning on the timed up & go test with stroke subjects. Top Stroke Rehabil of the turning direction in the TUG test. When turning in a 2009;16:196–206. standing posture, gait ability is involved, but trunk stabil- 9.  Matinolli M, Korpelainen JT, Korpelainen R, Sotaniemi KA, Matinolli ity is also closely associated with maintaining balance.23 VM, Myllylä VV. Mobility and balance in Parkinson’s disease: a popula- Therefore, future studies should analyze various other tion-based study. Eur J Neurol 2009;16:105–11. factors, such as trunk function, postural sway, number of 10.  Ng SS, Hui-Chan CW. The timed up & go test: its reliability and as- steps, and fear of falling. sociation with lower-limb impairments and locomotor capacities in people with chronic stroke. Arch Phys Med Rehabil 2005;86:1641–7. Conclusions 11.  Heung TH, Ng SS. Effect of seat height and turning direction on the timed up and go test scores of people after stroke. J Rehabil Med The present study found that stroke patients showed differ- 2009;41:719–22. ences in the TUG test results based on their turning direc- 12.  Perry J, Garrett M, Gronley JK, Mulroy SJ. Classification of walking tion, and less time was required when turning in the direc- handicap in the stroke population. Stroke 1995;26:982–9. tion of the affected side compared to the unaffected side. 13.  Tashiro H, Isho T, Hoshi F. Relationship of dynamic balance and mo- Going forward, better instructions should be provided in bility to community ambulation following stroke. Arch Phys Med Rehabil clinical settings considering the fact that the turning direc- 2017;98:e101–2. tion can affect patients’ TUG test results. 14.  Feys P, Bibby B, Romberg A, Santoyo C, Gebara B, de Noordhout BM, et al. Within-day variability on short and long walking tests in per- References sons with multiple sclerosis. J Neurol Sci 2014;338:183–7. 15.  Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible sta- 1.  Dite W, Temple VA. Development of a clinical measure of turning for tistical power analysis program for the social, behavioral, and biomedical older adults. Am J Phys Med Rehabil 2002;81:857–66, quiz 867–8. sciences. Behav Res Methods 2007;39:175–91. 2.  Glaister BC, Bernatz GC, Klute GK, Orendurff MS. Video task 16.  Chan WN, Tsang WW. The performance of stroke survivors in turn- analysis of turning during activities of daily living. Gait Posture ing-while-walking while carrying out a concurrent cognitive task com- 2007;25:289–94. pared with controls. PLoS One 2017;12:e0189800. 3.  Harris JE, Eng JJ, Marigold DS, Tokuno CD, Louis CL. Relationship 17.  Cumming RG, Klineberg RJ. Fall frequency and characteristics and of balance and mobility to fall incidence in people with chronic stroke. the risk of hip fractures. J Am Geriatr Soc 1994;42:774–8. Phys Ther 2005;85:150–8. 18.  Lord SE, McPherson K, McNaughton HK, Rochester L, Weather- 4.  Hollands KL, Hollands MA, Zietz D, Wing AM, Wright C, van Vliet all M. Community ambulation after stroke: how important and obtain- P. Kinematics of turning 180 degrees during the timed up and go in stroke able is it and what measures appear predictive? Arch Phys Med Rehabil survivors with and without falls history. Neurorehabil Neural Repair 2004;85:234–9. 2010;24:358–67. 19.  Pearson M, Dieberg G, Smart N. Exercise as a therapy for improve- 5.  Kobayashi M, Takahashi K, Sato M, Usuda S. Association of perfor- ment of walking ability in adults with multiple sclerosis: a meta-analysis. mance of standing turns with physical impairments and walking ability in Arch Phys Med Rehabil 2015;96:1339–1348.e7. patients with hemiparetic stroke. J Phys Ther Sci 2015;27:75–8. 20.  Manji A, Amimoto K, Matsuda T, Wada Y, Inaba A, Ko S. Effects of 6.  Podsiadlo D, Richardson S. The timed “Up & Go”: a test of basic func- transcranial direct current stimulation over the supplementary motor area tional mobility for frail elderly persons. J Am Geriatr Soc 1991;39:142–8. body weight-supported treadmill gait training in hemiparetic patients after 7.  Åhlund K, Bäck M, Öberg B, Ekerstad N. Effects of comprehensive stroke. Neurosci Lett 2018;662:302–5. geriatric assessment on physical fitness in an acute medical setting for frail 21.  Dalgas U, Severinsen K, Overgaard K. Relations between 6 minute elderly patients. Clin Interv Aging 2017;12:1929–39. walking distance and 10 meter walking speed in patients with multiple sclerosis and stroke. Arch Phys Med Rehabil 2012;93:1167–72. 22.  Bonnyaud C, Pradon D, Vaugier I, Vuillerme N, Bensmail D, Roche N. Timed Up and Go test: comparison of kinematics between patients with chronic stroke and healthy subjects. Gait Posture 2016;49:258–63. 23.  Verheyden G, Nieuwboer A, De Wit L, Feys H, Schuback B, Baert I, et al. Trunk performance after stroke: an eye catching predictor of func- tional outcome. J Neurol Neurosurg Psychiatry 2007;78:694–8. Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material discussed in the manuscript. Funding.—This work was supported by the National Research Foundation of Korea (NRF) funded by the Korea government (MSIT) No. 2017R1C1B5074040. Authors’ contributions.—Hohee Son contributed to the design of the study, the analysis and interpretation of the results, and writing the content of the manu- script. Chanhyun Park collected and interpreted the data. Article first published online: July 6, 2018. - Manuscript accepted: July 5, 2018. - Manuscript revised: June 27, 2018. - Manuscript received: January 26, 2018. Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 39

COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically © 2017 EDIZIONI MINERVA MEDICA European Journal of Physical and Rehabilitation Medicine 2019 February;55(1):40-6 or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access Online version at http://www.minervamedica.it DOI: 10.23736/S1973-9087.17.04851-1 to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. ORIGINAL ARTICLE Evaluation of the occurrence and diagnose definitions for nocturnal polyuria in spinal cord injured patients during rehabilitation Annick VIAENE 1 *, Marie-Astrid DENYS 2, An-Sofie GOESSAERT 2, Jana CLAEYS 3, Ann RAES 4, Saskia ROGGEMAN 2, Karel EVERAERT 2 1Department of Physical Medicine and Orthopedic Surgery, Ghent University Hospital, Ghent, Belgium; 2Department of Urology, Ghent University Hospital, Ghent, Belgium; 3 Department of Gynecology, Ghent University Hospital, Ghent, Belgium; 4Department of Pediatric Nephrology, Ghent University Hospital, Ghent, Belgium *Corresponding author: Viaene Annick, De Pintelaan 185, 9000 Gent, Belgium. E-mail. [email protected] ABSTRACT BACKGROUND: Little is known about the occurrence of nocturnal polyuria (NP) in spinal cord injured (SCI) patients and the definitions which are preferable in this population. AIM: To determine the occurrence of NP in SCI patients during in-patient rehabilitation in the Ghent University Hospital. To study the influence of different time periods (daytime, bed rest and sleep) on the accuracy of the existing diagnose definitions for NP specifically for this type of patients. DESIGN: Retrospective study using patient records. SETTING: SCI patients during hospital-based rehabilitation between 2011 and 2014. POPULATION: Seventy-four SCI patients were selected and their records of frequency-volume charts (FVC) were examined, after exclusion of unreliable data, forty-seven patients were retained for the current study. METHODS: Retrospective study using data from FVC of either two or three days from patients with SCI. Nocturnal urine production (NUP) and nocturnal polyuria index (NPi) were calculated. RESULTS: There was a significant increase in diuresis, calculated as urine production, between day time and bed rest (P=0.008) and between day time and sleep (P=0.001). All patients showed NP during a 12-hour night time period (including both bed rest and sleep) and 39 patients showed NP during the 8-hour period of sleep. There was no significant difference in mean urine production between bed rest and sleep. Preva- lence of NP did not significantly differ between the complete or incomplete SCI patients or between patients with higher and lower SCI levels. CONCLUSIONS: This study showed that the occurrence of NP in patients with SCI is high and that it is important to consider which definitions of NP are used for diagnosis. Increase in diuresis is observed during bed rest and sleep and the diagnose is correctly estimated when nocturnal urine production definitions are used in both time periods. In accordance with what was expected, diagnose of NP was overestimated when NP index type definitions were used. CLINICAL REHABILITATION IMPACT: It is important to be aware of the frequent-occurrence of NP in SCI patients and the impact of their daily routine to the accuracy of the diagnosis of NP. More knowledge about this topic can help to avoid incontinence caused by nocturnal polyuria. (Cite this article as: Viaene A, Denys MA, Goessaert AS, Claeys J, Raes A, Roggeman S, et al. Evaluation of the occurrence and diagnose defini- tions for nocturnal polyuria in spinal cord injured patients during rehabilitation. Eur J Phys Rehabil Med 2019;55:40-6. DOI: 10.23736/S1973- 9087.17.04851-1) Key words: Spinal cord injuries - Nocturnal enuresis - Rehabilitation. Nocturia and nocturnal incontinence are bothersome general health.1, 2 Existence of nocturia is strongly related symptoms which are very often caused by nocturnal to age. 35% of men and 50% of women between 20- and polyuria (NP). NP is defined as the production of an ab- 40-years old deals with the problem of nocturia. For men normally large volume of urine during sleep. When people and women older than 70 years these percentages even in- have to wake up two or more times a night, to void, it can crease to 69% and 77% respectively.3 compromise their quality of sleep, cognitive functions and NP is a known phenomenon in spinal cord injured (SCI) pa- 40 European Journal of Physical and Rehabilitation Medicine February 2019

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, NOCTURNAL POLYURIA IN SCI PATIENTS DURING REHABILITATION VIAENE cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. tients.4 Nevertheless, it has been rarely described in litera- ple.15 Reference values of NUP>90 mL/hour (1.5 mL/min) ture and the discomfort for these patients is not yet studied. have already been cited in men aged 50-78 years old.16 Physical experts who work daily with SCI patients in NP can also be diagnosed if the percentage of the total rehabilitation centers experience that there probably is 24-hour urine volume produced during the time spent in a relationship between NP and the edemas in the lower bed (when the patient is considered to be sleeping), ex- extremities. These patients build up edemas during the ceeds the percentage of time spent in bed within these day when they are sitting in their wheelchair and when 24-hours.17 However, when this method is used for SCI the patients go to lie in bed with their legs up, the ede- patients, the time spent in bed before going to sleep (e.g. mas disappear towards the morning. This natural drainage watching TV) is also defined as sleeping time. Therefore, mechanism can increase the urine production during the to calculate NP, it is probably more accurate to exclude the night and therefore SCI patients are probably a much more time spent awake in bed from sleeping time.1 vulnerable group to suffer from NP. This phenomenon is supported by studies about lifestyle measures for the treat- NP definitions based on weight are difficult to use in ment of nocturia.5-7 patients who have a SCI because they require a specialized platform scale for wheelchair users or a scales hook for Due to NP it is necessary doing the bladder emptying patients who cannot stand up.18 more than once during the night. When intermittent cath- eterization is used, sleep is disrupted 2 or 3 times at night. Because autonomic (dys)function can occur differently When a suprapubic catheter is used, often for cervical SCI according to the level of SCI, this study will focus on NP patients with limited hand function or tetraplegics with ar- in relation to SCI above and below T5. In addition, the tificial ventilation, the burden can be less.8-10 completeness of SCI (the ASIA type) can be important.19-22 NP can result from endogenous fluid and solute shift, The aim of this research was to measure the occurrence cardiovascular and autonomic disease.11, 12 However, the of NP in SCI patients, evaluate the applicability of the dif- mechanism of NP remains unclear and little is known ferent diagnose definitions of NP for this type of patients about its exact prevalence in SCI patients. and to study the influence of the level and (in)complete- ness of the spinal lesion on NP. Although NP is descriptively defined as the produc- tion of an abnormally large volume of urine during sleep, Materials and methods there are different definitions that can be used to diagnose whether a patient has NP or not. In the current study two The study was approved by the Ghent University Hospi- definitions are applied: the Nocturnal Polyuria Index (NPi) tal review board (EC UZ Gent 2014/0256), 18th of March, and the nocturnal urine production volume (NUP). Belgian registration number: B670201420552. The decla- ration of Helsinki was followed and written informed con- Cut-off values for both NP definitions depend on the age sent for study participation was obtained from all subjects. groups for which they are validated.13 Seventy-four patient records from SCI patients who were When the NPi definition is used, NP is defined as an hospitalized for rehabilitation between 2011 and 2014, were urinary output during sleep of 20% or more of the 24-hour retrospectively analyzed. Inclusion criteria were SCI pa- total in the young and of 33% or more in the elderly. This tients in a first rehabilitation following their SCI, with useful definition assumes persons with normal 24-hour urine out- frequency-volume charts (FVC) for 2 or (preferably) 3 days put and a nocturnal sleeping duration of 8 hour.1 The differ- in a maximum period of 3 weeks with a normal fluid intake ences between “young” and “old” are taken into account in and continence during the selected 24-hour analysis period. the “mixed” definitions in which 60 years old is the cut-off Exclusion criteria were gavage, incontinence (there is no age, NPi>20% for people <60 years and >30% for people standard procedure for weighing incontinence pads), di- of 60 years and older, or in which 65 years old is the cut- uretic medication or bladder enlargement. For each patient, off age, NPi>20% for people <65 years and NPi>33% for periods of intravascular infusion and days of occasional in- people of 65 years and older. The rate of NUP is calcu- continence were excluded. Days where a patient required lated as the nocturnal urine volume during time asleep and more than 6 catheterizations and patients with a non-stan- is measured in milliliters per minute. The reference value dard record of registration were discarded from analysis. for NUP of 0.9 mL/min is based on the results of a study in young men aged 25-35 years.2, 14 However due to the FVC, age at SCI, type of SCI (level and ASIA grading) changing diurnal variation of urine production with age, were available in the patient records. Patients with reliable this reference value is not suitable for use in older peo- FVCs for 3 days were included and ideally, if available, Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 41

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, VIAENE NOCTURNAL POLYURIA IN SCI PATIENTS DURING REHABILITATION cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. 3 consecutive days were used. A minimum of 2 FVC was y) and NPi>33% (≥65y). For calculations of NUP, urinary accepted for inclusion since it was not always possible to volumes of 0.9 mL/min, 1.3 mL/min and 90mL/hour were measure the volume when a patient was away from the included. Nocturnal urine volume (NUV) as a percentage rehabilitation center. For example, absences were due to a of the total urine volume produced during 24-hours was result of occasional incontinence, a weekend at home, the also included if it exceeded the percentage of time spent in need for other investigations and examination elsewhere bed during these 24-hours. The distribution of the different and altered fluid intake due to intravenous infusion. An ad- age groups was based on those used in other studies.5-10 mission period of 3 weeks was accepted. Statistical analysis The daily routine in our rehabilitation center starts with the morning care at 08:00, patients are following therapy The SPSS program version 23 was used for statistical in the morning and in the afternoon, have their meals at analyses. Normality of the data was tested using the Sha- the restaurant at 12:00 and 17:00 and most of them go to piro-wilks test and QQ plots. A series of paired Student’s lay down in bed after diner, usually at 20:00. Most often t-tests were used to compare NP during the 3 different time they get their last bladder emptying at midnight and then periods (day time, bed rest and sleep). they go to sleep. This day schedule is different compared to other types of patients and routines because of the bed A Fisher’s Exact Test was used to study the frequency rest period without sleep between 20:00 and 00:00. distribution of the bladder emptying methods and gender in the different age classes. Patients with clean intermittent catheterization (CIC) have a frequency of 5 to 6/day with a minimum of 4/day, Significance differences in urine production between at 12:00, 16:00, 20:00, 00:00, 04:00, 08:00. Patients with a the different bladder emptying moments was evaluated us- suprapubic catheter have a recording of frequency ideally ing a Friedman test with post hoc Bonferroni correction 6 times a day. Patients who can perform the catheterization for multiple testing. independently are responsible for their own recording with nursing supervision. The daily schedule in the rehabilita- Results tion hospital was taken as the basis for daytime recordings and this was from 0800h until dinner. Most of the patients A total of 74 patients were screened, of which 47 records went to bed between 20:00 and 22:00, with most sleeping with sound FVC were retained for the study. Twenty-seven between 22:00 and 00:00. For the purpose of this study a patient records were excluded due to: gavage (2), inconti- bed rest period without sleep was taken into account be- nence (5), diuretic medication (1), frequent intravascular in- tween 20:00 and 00:00 and sleeping time was defined to fusion due to antibiotic use 3 times a week for chronic infec- be between 00:00 and 08:00 a.m. Together, bed rest and tion (1), bladder augmentation (1) and unusable data (17). sleeping time were considered as “night time” thus starting at 20:00 p.m. until 08:00 a.m. The 47 patients, 33 male and 14 females, had a usable FVC over a period of maximum 19 days. Thirty-one pa- Personal information on age, gender, date of SCI, level tients had 3 FVC within 7 days. Ten patients required more of SCI and ASIA type was recorded. than 7 days to obtain 3 Frequency-Volume charts. Six pa- tients were accepted with only 2 days of records. The use The increments used for data analysis were NPi 20%, of the bladder emptying methods differed significantly be- NPi 30%, NPi 33%, and specific age related definitions tween different age groups (P=0.017) but the distribution NPi>20% (<60 y) and >30% (≥60 y), and NPi>20% (<65 of the sexes did not (P=0.19) (Table I). For the age groups Table I.—The differences in distribution between sexes and between different bladder emptying methods for each age group (Fishers exact test, P values are respectively 0.19 and 0.017). Gender Male N. (%) <35 Age group  >50   Female N. (%) Bladder emptying method Intermittent catheterization N. (%) 10 (83.3) 35-50 11 (55.0)   Intermittent catheterization by a nurse N. (%) 2 (16.7) 9 (45.0) Suprapubic catheter N. (%) 7 (58.3) 12 (80.0) 1 (5.0) Spontaneous micturation N. (%) 0 (0.0) 3 (20.0) 3 (15.0) 4 (33.3) 7 (46.7) 13 (65.0) 1 (8.3) 1 (6.7) 3 (15.0) 6 (40.0) 1 (6.7) 42 European Journal of Physical and Rehabilitation Medicine February 2019

COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA NOCTURNAL POLYURIA IN SCI PATIENTS DURING REHABILITATION VIAENE This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access <35 years and 35-50 years intermittent cathetherization 1200 to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove,was most frequently used (respectively 58% and 47%), cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. and for patients older than 50 years the subrapubic cathed- 1000 er was most frequently used to empty the bladder (65%). Frequency 800 Urine production (mL) The mean age of patients with SCI was 46±16 years. The mean time span between the date of the SCI and the 600 first registration date at the revalidation center was 97 days +/- 88 days. 400 The age distribution of the studied patients is presented 200 in Figure 1. 0 Twenty-seven patients (57%) had an SCI level of T5 12h 16h 20h 00h 04h 08h or higher and 20 patients (43%) had an SCI level of T6 or lower. Twenty-eight (60%) patients had a complete lesion Figure 2.—Urine production at each bladder emptying moment. Boxes (ASIA type A) and 19 (40%) patients had an incomplete that have no letters in common are significantly different from each oth- lesion (5 patients were ASIA B, 8 were ASIA C and 6 were er (Friedman test with Bonferoni corrections). ASIA D). Table II.—NP calculated for the sleeping time of 8 hours and for The mean 24-hour urine volume is 2123.40±585.24 mL. a night of 12 hours. The urine volumes differed significantly between bladder emptying moments (Friedman test P<0.0001) (Figure 2).   Night of 12 hours Sleeping time of 8 hours 00:00/04:00/08:00 a.m. 04:00/08:00 a.m. The number of patients with NP for the sleeping time Bladder emptying of 8 hours (without bed rest) and for a night of 12 hours moments included: N. (%) N. (%) (including bed rest in the evening) were calculated based at the different definitions for NP (Table II). Definition to diagnose NP 47 (100) 46 (97.9) 47 (100) 36 (76.6) Table III illustrates the urine production (UP) during the NPi>20% 46 (97.9) 32 (68.1) 3 different time periods: day time, bed rest and sleep Urine NPi>30% 47 (100) 44 (3.6) production during the day (08:00h-20:00h) was signifi- NPi>33% 46 (97.9) 34 (72.3) cantly lower than those during bed rest (20:00h-00h) Stu- NPi>20% en >30% 32 (68.1) dents t-test, P=0.008). A similar significant difference was NPi>20% en >33% 44 (93.6) NA found between day time (08:00h-20:00h) and sleep (00h- NUV>50% 35 (74.5) 41 (87.2) NUP 0.9 mL/min 27 (57.4) 36 (76.6) Mean = 46,1277 NUP 1.3 mL/min 27 (57.4) Std. dev. = 15,8548 NUP 90 mL/hr N.=47 6,0 Table III.—Mean urine production ± standard deviation (SD) dur- ing day time (12-h), bed rest (4-h) and sleeping time (8-h) and P values of the paired students t-test. 4,0 Mean±SD P values } }2,0 0.907 }NUP 20-h – 00-h bed rest 4h UP 08-h – 20-h day time 12h 1.30 ± 0.44 0.008 1.66 ± 0.91 0.001 NUP 00-h – 08-h sleeping time 8h 1.69 ± 0.69 0,0 40,00 60,00 80,00 08:00h) (P=0.001). There was no significant difference in 20,00 urine production between bed rest and sleep (P=0.907). Age at SCI The rate of NUP is age dependent. The NUP 0.9 mL/ min, NUP 1.3 mL/min and NUP 90 mL/hour calculated for Figure 1.—Age distribution of the patient cohort. Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 43

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, VIAENE NOCTURNAL POLYURIA IN SCI PATIENTS DURING REHABILITATION cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. Table IV.—Number of patients with NP based on the different diagnose definitions (NUP or NPi) during night time (12h, bed rest +sleep- ing time) or sleeping time (8h), for the different age groups. Bold numbers are the values for the age groups for which the definitions are initially intended to.   NUP NUP NUP NPi>20% NPi>30% NPi>33%   0.9 mL/min 1.3 mL/min 90 mL/h N. (%) N. (%) N. (%) N. (%) Night time (12h) N. (%) N. (%) 12 (100) 12 (100) 11 (92) Age <35 years 7 (58) 15 (100) 15 (100) 15 (100) Age between 35-49 years 9 (75) 9 (75) 10 (67) 20 (100) 20 (100) 20 (100) Age ≥50 years 15 (100) 10 (67) 10 (50) 20 (100) 16 (80) 12 (100) 9 (75) 7 (58) Sleeping time (8h) 7 (58) 15 (100) 10 (67) 8 (53) Age <35 years 9 (75) 9 (75) 8 (53) 19 (95) 17 (85) 17 (85) Age between 35-49 years 13 (87) 11 (73) 12 (60) Age ≥50 years 19 (95) 16 (80) the different age groups are presented in Table IV. For the they are sitting in their wheelchair during daytime. Chang- 12-hour night time, 29 (62%) patients have NP. For the 8h ing position to a recumbent position during the night, in- sleeping time, 32(68%) patients have NP. creases intravascular volume. The increased intravascular volume also stimulates the secretion of the atrial natriuret- When NP was calculated based on NPi, according to the ic peptide and inhibits the renin-angiotensin-aldosterone age groups all patients have NP in a 12-hour night time system.1, 9, 21, 24-28 Both mechanisms increase natriuresis. period and 39 patients have NP during the 8-hour sleeping This surplus of fluid presented to the kidneys would lead time. to a higher nocturnal diuresis which then contributes to the increased solute diuresis during the night. Exploring There were no significant differences between the high- the therapeutic use of bed rest in the afternoon or in the er level SCI and the lower level SCI for any of the in- evening could be interesting for patient with incontinence crements when we compare cervical SCI with the lower during night because of NP. SCI or when we compared T5 and higher to T6 and lower. There were no significant differences between the com- The NUP-based definitions are more consistently dem- plete ASIA A patients and the incomplete ASIA B, C and onstrating NP when excess diuresis is occurring through- D patients in any of the situations. out the night. According to Van Doorn, the NUP definitions depend more so on age, which suggests that consideration Discussion of age distribution within a patient cohort is an important factor when determining the most appropriate definitions This study demonstrates that NP occurs very frequently to use.16 The age distribution in Table I is based on the in SCI patients regardless of their age, level of lesion and age cut-off limits reported by Blanker.15 Possibly, a NUP completeness of the lesion. The results of Table II show of 90 mL per hour was too harsh to use as a guide for the that, when the NPi definition is used, the number of SCI cohort in the current study. However Table IV clearly il- patients with NP tend to be overestimated. This is most lustrates that the obtained numbers closely match the cal- likely because no distinction between bed rest and sleep culations, according to the age categories. This supports is used and therefore the period of bed rest is counted as the view that definitions should be used for the population sleeping time. for which they have been validated, depending on age and hours of bed rest or sleep. There are no studies available Table III demonstrates that NP in SCI patients is not concerning the impact of the daily routine and the age to only related to sleep but is also related to resting in the the accuracy of the diagnosis of NP in SCI patients. supine position during the first stages of the night. This suggests that there are possibly different mechanisms for According to the ICS report 2002, the nocturnal urine NP in different phases of the night. The first mechanism volume is the sum of the voiding volumes during sleep. could be explained by the disturbed circadian rhythm of However, Szollar divided the 24-hour period into a 12- vasopressin which would be inadequate or even absent in hour daytime (08:00-20:00hrs) and a 12- hour night time para- and tetraplegics and therefore contribute to an in- (20:00-08:00hrs) in his 1995 study that pre-dates the ICS creased water diuresis during the night.4, 21, 23, 24 Secondly, report.4, 8 Our study demonstrates that an increase in di- it is observed that patients with spinal cord lesions experi- uresis occurs during bed rest and during sleep with the ence more fluid retention in their lower extremities when 44 European Journal of Physical and Rehabilitation Medicine February 2019

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, NOCTURNAL POLYURIA IN SCI PATIENTS DURING REHABILITATION VIAENE cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. prevalence correctly estimated by NUP definitions in both In conclusion, we might expect different mechanisms periods. As expected, this was overestimated by NPi type of NP in SCI patients. We can hypothesize reabsorption definitions. The observation of an increased bladder vol- of edema in the lower extremities with salt diuresis dur- ume at 20:00 hours (Figure 2), indicates that some patients ing bed rest but this aspect needs to be studied more in may lie down after their evening meal which takes place detail. The results of this study can help to avoid NP in between 17:00 hours and 18:00 hours. This phenomenon SCI patients. of the impact of posture has been described by Koji Yo- shimura, Kooner and Gutte.17, 22, 29 References The autonomic dysfunction resulting in loss of vascu- 1.  van Kerrebroeck P, Abrams P, Chaikin D, Donovan J, Fonda D, Jack- lar tone in the lower extremities with pooling of fluid in son S, et al.; Standardisation Sub-committee of the International Conti- the legs is the cause of a decreased intravascular volume nence Society. The standardisation of terminology in nocturia: report from during daytime. Due to the recumbence at night, a redis- the Standardisation Sub-committee of the International Continence Soci- tribution of extravascular fluid into the intravascular space ety. Neurourol Urodyn 2002;21:179–83. prompts the intravascular volume to rise and causes an in- 2.  Asplund R. The nocturnal polyuria syndrome (NPS). Gen Pharmacol crease in urine production. This phenomenon is very ap- 1995;26:1203–9. parent in cervical SCI patients.22, 30, 31 In former studies, the 3.  Bosch JL, Weiss JP. The prevalence and causes of nocturia. J Urol number of SCI patients is small and often the tetraplegic 2013;189(Suppl):S86–92. population is not compared with a paraplegic group, which 4.  Szollar S, North J, Chung J. Antidiuretic hormone levels and polyuria makes it difficult to underbuilt the effect of differences in in spinal cord injury. A preliminary report. Paraplegia 1995;33:94–7. the autonomic (dys)function between SCI patient accord- 5.  Soda T, Masui K, Okuno H, Terai A, Ogawa O, Yoshimura K. Effi- ing to the level and completeness of injury. However, in cacy of nondrug lifestyle measures for the treatment of nocturia. J Urol the current study, we found no difference based on either 2010;184:1000–4. level of SCI or on whether the patient had a complete or 6.  Barkin J. Nocturia: diagnosis and management for the primary care incomplete lesion. physicians. Can J Urol 2016;23(Suppl 1):16–9. 7.  Wein AJ. Re: The New England Research Institutes, Inc. (NERI) Noc- The third mechanism relates the diurnal blood pressure turia Advisory Conference 2012: focus on outcomes of therapy. J Urol patterns and nocturnal hypertension to the measurement of 2014;191:1053–4. diurnal urine production in SCI patients.19 Elevated noc- 8.  Hunter KF, Bharmal A, Moore KN. Long-term bladder drainage: Su- turnal pressure in autonomic neuropathy is suggested as a prapubic catheter versus other methods: a scoping review. Neurourol Uro- possible reason for nocturnal diuresis with volume deple- dyn 2013;32:944–51. [Review] tion and postural hypotension during daytime. The interac- 9.  Sorokin I, De E. Options for independent bladder management in pa- tion between other regulating mechanisms remains unclear. tients with spinal cord injury and hand function prohibiting intermittent catheterization. Neurourol Urodyn 2015;34:167–76. [Review] Limitations of the study 10.  Böthig R, Hirschfeld S, Thietje R. Quality of life and urological mor- bidity in tetraplegics with artificial ventilation managed with suprapubic The number of patients included in this study was rather or intermittent catheterisation. Spinal Cord 2012;50:247–51. small and due to various reasons, became further reduced 11.  Gulur DM, Mevcha AM, Drake MJ. Nocturia as a manifestation of when the reliability of the measurements was not certain. systemic disease. BJU Int 2011;107:702–13. Therefore, it is possible that this had an effect on the sta- 12.  Jin MH, Moon G. Practical management of nocturia in urology. In- tistical processing. Not having a gold standard for defining dian J Urol 2008;24:289–94. NP is a limitation for all NP studies. 13.  Hofmeester I, Kollen BJ, Steffens MG, Bosch JL, Drake MJ, Weiss JP, et al. Impact of the International Continence Society (ICS) report Conclusions on the standardisation of terminology in nocturia on the quality of re- ports on nocturia and nocturnal polyuria: a systematic review. BJU Int The results of this retrospective study suggest that the oc- 2015;115:520–36. currence of NP is 68% amongst SCI patients regardless of 14.  Kirkland JL, Lye M, Levy DW, Banerjee AK. Patterns of urine flow their age, level of lesion and completeness of the lesion. and electrolyte excretion in healthy elderly people. Br Med J (Clin Res Ed) 1983;287:1665–7. Increase in diuresis is observed during bed rest and 15.  Blanker MH, Bernsen RM, Bosch JL, Thomas S, Groeneveld FP, sleep and NP is correctly diagnosed when the NUP defini- Prins A, et al. Relation between nocturnal voiding frequency and noc- tions are used. As expected, diagnosis of NP was overesti- turnal urine production in older men:a population-based study. Urology mated by the NPi definitions. 2002;60:612–6. 16.  van Doorn B, Blanker MH, Kok ET, Westers P, Bosch JL. Preva- lence, incidence, and resolution of nocturnal polyuria in a longitudinal community-based study in older men: the Krimpen study. Eur Urol 2013;63:542–7. 17.  Yoshimura K, Terai A. Classification and distribution of symptomatic nocturia with special attention to duration of time in bed: a patient-based study. BJU Int 2005;95:1259–62. Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 45

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, VIAENE NOCTURNAL POLYURIA IN SCI PATIENTS DURING REHABILITATION cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. 18.  Rembratt A, Norgaard JP, Andersson KE. What is nocturnal polyuria? 25.  Lote CJ. Principles of renal physiology. 5th ed. New York: Springer; BJU Int 2002;90(Suppl 3):18–20. 2012. 19.  Goh MY, Wong EC, Millard MS, Brown DJ, O’Callaghan CJ. A ret- 26.  Mathias CJ. Orthostatic hypotension and paroxysmal hyper- rospective review of the ambulatory blood pressure patterns and diurnal tension in humans with high spinal cord injury. Prog Brain Res urine production in subgroups of spinal cord injured patients. Spinal Cord 2006;152:231–43. 2015;53:49–53. 27.  Goessaert AS, Krott L, Hoebeke P, Vande Walle J, Everaert K. Diag- 20.  Krassioukov A. Autonomic function following cervical spinal cord nosing the pathophysiologic mechanisms of nocturnal polyuria. Eur Urol injury. Respir Physiol Neurobiol 2009;169:157–64. 2015;67:283–8. 21.  Kilinç S, Akman MN, Levendoglu F, Özker R. Diurnal variation of 28.  Claydon VE, Steeves JD, Krassioukov A. Orthostatic hypotension antidiuretic hormone and urinary output in spinal cord injury. Spinal Cord following spinal cord injury: understanding clinical pathophysiology. Spi- 1999;37:332–5. nal Cord 2006;44:341–51. 22.  Kooner JS, Frankel HL, Mirando N, Peart WS, Mathias CJ. Haemo- 29.  Guite HF, Bliss MR, Mainwaring-Burton RW, Thomas JM, Drury PL. dynamic, hormonal and urinary responses to postural change in tetraple- Hypothesis: posture is one of the determinants of the circadian rhythm of gic and paraplegic man. Paraplegia 1988;26:233–7. urine flow and electrolyte excretion in elderly female patients. Age Age- 23.  Szollar SM, Dunn KL, Brandt S, Fincher J. Nocturnal polyuria and ing 1988;17:241–8. antidiuretic hormone levels in spinal cord injury. Arch Phys Med Rehabil 30.  Williams HH, Wall BM, Horan JM, Presley DN, Crofton JT, Share L, 1997;78:455–8. et al. Nonosmotic stimuli alter osmoregulation in patients with spinal cord 24.  Zahariou A, Karagiannis G, Papaioannou P, Stathi K, Michail X. The injury. J Clin Endocrinol Metab 1990;71:1536–43. use of desmopressin in the management of nocturnal enuresis in patients 31.  Leehey DJ, Picache AA, Robertson GL. Hyponatraemia in quadriple- with spinal cord injury. Eura Medicophys 2007;43:333–8. gic patients. Clin Sci (Lond) 1988;75:441–4. Conflicts of interest.—Karel Everaert: consultant, travel and research grants, principal investigator for Ferring, Allergan, Astellas. There is no conflict of interest for any of the other authors. Funding.—Viaene Annick receives an educational grant from the University Hospital Ghent. Article first published online: November 3, 2017. - Manuscript accepted: October 30, 2017. - Manuscript revised: October 13, 2017. - Manuscript received: June 6, 2017. 46 European Journal of Physical and Rehabilitation Medicine February 2019

COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically © 2018 EDIZIONI MINERVA MEDICA European Journal of Physical and Rehabilitation Medicine 2019 February;55(1):47-55 or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access Online version at http://www.minervamedica.it DOI: 10.23736/S1973-9087.18.05111-0 to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. ORIGINAL ARTICLE Outcome measures in the clinical evaluation of ambulatory Charcot-Marie-Tooth 1A subjects Laura MORI 1, 2 *, Valeria PRADA 1, 2, Alessio SIGNORI 3, Davide PAREYSON 4, Giuseppe PISCOSQUITO 4, 5, Luca PADUA 6, 7, Costanza PAZZAGLIA 7, Gian Maria FABRIZI 8, Nicola SMANIA 9, Alessandro PICELLI 9, Angelo SCHENONE 1, 2, TreSPE Study Group ‡ 1Department of Neuroscience, Rehabilitation, Ophthalmology, Genetics, Maternal and Child Health, University of Genoa, Genoa, Italy; 2San Martino University Hospital and Institute for Cancer Research and Care, Genoa, Italy; 3Division of Biostatistics, Department of Health Science (DISSAL), University of Genoa, Genoa, Italy; 4Carlo Besta Neurological Institute for Research and Care, Milan, Italy; 5Functional Neuromotor Rehabilitation Unit, Maugeri Scientific Institutes, Telese Terme, Benevento, Italy; 6Department of Geriatrics, Neurosciences, and Orthopedics, Sacred Heart Catholic University, Rome, Italy; 7Don Carlo Gnocchi Nonprofit Organization, Milan, Italy; 8Section of Neurology, Department of Neurosciences, Biomedicine, and Movement Sciences, University of Verona, Verona, Italy; 9Neuromotor and Cognitive Rehabilitation Research Center, Department of Neurosciences, Biomedicine and Movement Sciences, University of Verona, Verona, Italy ‡Members are listed at the end of the paper. *Corresponding author: Laura Mori, Department of Neuroscience, Rehabilitation, Ophthalmology, Genetics, Maternal and Child Health (DINOGMI), Uni- versity of Genoa, Largo Daneo 3, 16132 Genoa, Italy. E-mail: [email protected] ABSTRACT BACKGROUND: The outcome measures (OMs) in clinical trials for Charcot-Marie-Tooth disease (CMT) still represent an issue. A recent study highlighted that three additional clinical OMs, the 10-Meter Walk Test (10MWT), the 9-Hole Peg Test, and foot dorsal flexion dynamometry, further improve discrimination between severely and mildly affected patients. Another study has recently assessed the validity and reliability of the 6-Minute Walk Test (6MWT). AIM: The aim of this study was to identify the most useful scales in the clinical evaluation of CMT1A patients. DESIGN: Observational study of the baseline data collected in a multicenter, prospective, randomized, single blind, controlled study to evaluate the efficacy and safety of an innovative rehabilitation protocol based on treadmill training, stretching, respiratory, and proprioceptive exercises (TreSPE study) in CMT1A patients. SETTING: The outpatient service of the four Italian centers involved, which are specialized in hereditary neuropathies. POPULATION: Fifty-three subjects with a clinical and genetically confirmed diagnosis of CMT1A. METHODS: At baseline, in addition to the CMT Neuropathy Score, all subjects underwent walking evaluation (6MWT, 10MWT), balance assessment (Berg Balance Scale [BBS], Short Physical Performance Battery [SPPB]) and a subjective evaluation of quality of life (SF36) and walking ability (Walk12). RESULTS: Analyzing the baseline data, as expected, we found a strong correlation between walk and balance evaluation, proving the validity of these tests in investigating the functional impairment of CMT1A subjects. Particularly, we found that subjects with better balance control walk at higher speed and perceive less limitations in their physical activities or motor skills. This can be reconducted to the fact that ankle stability depends upon different factors such as anatomy integrity, muscle strength and proprioception. CONCLUSIONS: We identify the 6MWT, 10MWT, and SPPB as the most useful scales, in addition to the CMTNS, to evaluate the functional impairment of CMT1A patients who retain their walking capability and we suggest the use of SPPB because of its rapidity to assess balance and gait disorders in clinical settings. CLINICAL REHABILITATION IMPACT: In the clinical practice it is important to evaluate patients comprehensively but rapidly. These out- come measures can help us to correctly assess balance and walking ability in CMT1A patients. (Cite this article as: Mori L, Prada V, Signori A, Pareyson D, Piscosquito G, Padua L, et al. Outcome measures in the clinical evaluation of ambulatory Charcot-Marie-Tooth 1A subjects. Eur J Phys Rehabil Med 2019;55:47-55. DOI: 10.23736/S1973-9087.18.05111-0) Key words: Charcot-Marie-Tooth disease - Outcome assessment - Walking - Postural balance - Quality of life. Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 47

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, MORI CHARCOT-MARIE-TOOTH 1A EVALUATION cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. Charcot-Marie-Tooth (CMT) neuropathy is the most selected the 10MWT, 6MWT, Walk12, Berg Balance Scale common inherited neurological disorder. It is char- (BBS), and Short Physical Performance Battery (SPPB) as acterized by autosomal dominant, recessive or X-linked OMs. These data will be useful in planning future clini- inheritance and it includes several variants among which cal trials in CMT neuropathy, especially if focused on the the most common type is CMT1A.1, 2 In CMT1A, symp- lower limb function and gait ability. toms usually start within the 1st or 2nd decade of life with slowly progressive distal muscle weakness and atrophy, Materials and methods although cases with an earlier onset are also known. These patients often complain about gait disorders with frequent This study was carried out in four Italian centers spe- falls and difficulties in running as dorsiflexor muscles of cialized in hereditary neuropathies: the Department of the ankle and of the toes are the most affected. In addition, Neuroscience, Rehabilitation, Ophthalmology, Genetics, joint tightness, deformities and altered proprioception fur- Maternal and Child Health (DINOGMI) at the Univer- ther impair muscle function, gait and balance.1, 3, 4 To date, sity of Genoa; the Carlo Besta Neurological Institute and there is no effective therapy addressed to specifically treat Foundation for Research and Care, in Milan; Centro Santa different forms of CMT and the efficacy of rehabilitation Maria della Pace at Don Carlo Gnocchi Nonprofit Organi- is unclear.2, 5-7 However, physiotherapy may slow the pro- zation in Rome; and the Department of Neurological and gression of symptoms in CMT1A, as well as in many other Visual Science at the University of Verona. The clinical chronic degenerative neurological diseases.8-10 Either aer- trial received the approval of the Ethical Committees of obic exercises, muscle strengthening, stretching, proprio- each Center (v.n. 4/2012). ceptive treatments or their combination may be applied as possible intervention methods. Inclusion criteria were: clinical and genetically con- firmed diagnosis of CMT1A; age between 18 and 75 years; A previous, preliminary study on the use of treadmill ability to walk without support with or without ankle foot training, stretching, respiratory, and proprioceptive exer- orthesis for at least 100 meters; score at the SPPB between cises (TreSPE) on a small population of patients affected 2 and 10; ability to understand and sign the informed con- by CMT, suggested that this rehabilitation protocol is well sent form. tolerated and effective in improving balance and gait func- tion.8 Being CMT a rare disease, we planned a multicenter, Exclusion criteria were: patients affected by HNPP or prospective, randomized, single blinded and controlled CMT types other than CMT1A; vestibular affections; psy- study to evaluate the efficacy and safety of the TreSPE chiatric, cardiovascular and lung disorders or severe ar- protocol. thropathic changes in the lower limbs preventing the per- formance of exercise resistance training; other associated The outcome measures (OMs) in clinical trials and in causes of neuropathy (diabetes, endocrine disorders, vas- clinical evaluation of patients affected by CMT still rep- culitis, herniated disc, use of potentially neurotoxic mol- resent an issue, although some recent reviews and clinical ecules as antineoplastic drugs or alcohol consumption). studies addressed their significance.5, 11-14 The most com- monly used scale is the CMT Neuropathy Score (CMTNS), The protocol required the subjects to be blindly ran- that has been recently modified (CMTNS2) to reduce floor domized into two treatment groups (either TreSPE or and ceiling effects and to improve sensitivity for detect- SPE) and were evaluated at baseline (T0), after 3 months ing changes over time.15, 16 A recent study highlighted that of treatment (T1) and after 6 months (T2). We recruited 53 three additional clinical OMs (the 10-meter Walk Test subjects affected by CMT1A (32 women, 21 males, mean [10MWT], the 9-Hole Peg Test, and foot dorsal flexion age: 52.1 years, range 19-69 years), which were evaluated dynamometry), further improve discrimination between at the baseline and who met all the inclusion criteria. In severely and mildly affected patients.13 Moreover, valid- each center, a single examining researcher, neurologist or ity and reliability of the 6-Minutes Walking test (6MWT) rehabilitation specialist was responsible of the evaluation as OM in patients affected by CMT has been recently as- of all patients. sessed.17 A complete neurological and physical examination was We present the baseline data of the TreSPE study, to performed after recording a detailed medical history, in- identify the best and more useful OMs to detect and quan- cluding duration of disease considered from the first on- tify disability in CMT1A neuropathy. TreSPE is a reha- set of symptoms, the course of the disease and all medical bilitation protocol based on walking exercise; hence, we treatment taken within the previous six months. Moreover, all patients underwent walking evaluation with the 6MWT 48 European Journal of Physical and Rehabilitation Medicine February 2019

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, CHARCOT-MARIE-TOOTH 1A EVALUATION MORI cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. and the 10MWT, which were chosen as primary outcome Statistical analysis measures in the TreSPE study, and a subjective evaluation of walking ability with the Walk12 scale. Lower limbs With a statistical power of 80% and an alpha error of 5%, strength evaluation in dorsiflexion and plantiflexion was a total of 46 patients, statistical analysis of data resulted in assessed with a dynamometer (Cit Technics, Groningen, obtaining a significant correlation of 0.4 or more. This was The Netherlands) and measured as maximum voluntary considered as acceptable for the aim and kind of study we isometric contraction.18 conducted. Balance was assessed with BBS and SPPB. Disability Mean and standard deviation or median with interquar- was evaluated with the CMTNS and not with its modified tile range (IQR: 25th-75th percentile) were reported for form as the trial started before the changes in the scale continuous characteristics. Pearson’s correlation coeffi- were done. A subjective evaluation of the quality of life cients were calculated between all continuous characteris- (QoL) through the Medical Outcomes Study – Short Form tics (i.e.: 6MWT, 10MWT, Foot dorsal and plantar) while 36 (SF36) was also administered. Spearman’s correlation coefficient was adopted for ordinal outcomes (CMTNS, Walk 12, BBS, SPPB). Before the The 10MWT is a validated test in CMT disease that as- application of correlation coefficient, a graphical inspec- sesses functional mobility and walking speed in meters tion was done to define distribution of variables and detect per seconds over a short duration in a set distance.13, 18 the possible presence of outliers. Since many correlation The 6MWT is a reliable and sensitive endurance test al- coefficients were assessed, P values associated to each ready validated in other neurological diseases and recently correlation were adjusted for multiple tests according to validated also in patients affected by CMT.17, 19-24 6MWT the false-discovery rate approach. According to Evans,39 mainly evaluates ambulation ability and aerobic resis- correlation coefficients up to 0.4 were considered weak or tance. Walk12 is a generic patient-reported rating scale for mild while in the range 0.4-0.6 moderate and from 0.6 to walking difficulties in daily life that has been recently vali- 0.8 moderately strong. Unpaired Student’s t-test was used dated as a measurement of walking ability in peripheral to test if continuous features such as age, CMTNS and dis- neuropathies.25 It is used to measure the impact of neuro- ease duration were significantly different between males logical disability on walking function in different neuro- and females. Subsequently, a main component analysis logical conditions.26-28 including all clinical and SF36 outcomes was performed. Number of components was determined according to Kai- The BBS, a 14-item objective test, is a sensitive scale ser criterion based on an eigenvalue >1 and to retain a to detect subtle balance impairment and fall risk that has clinical interpretability of results. An orthogonal rotation been validated in people affected by neurological disor- according to varimax criteria was performed to make in- ders.29-31 It has also been used in the assessment of disabil- terpretation easier. Stata (v.13; StataCorp.) was used for ity in patients affected by CMT.32, 33 computation. SPPB is a widely used instrument to quantify balance Results disorders and gait impairment in elderly people and many different diseases.34-38 A total of 53 patients (32 females; 60.4%) with a mean age of 52.1 years and age ranging between 19-69 were re- A hand-held dynamometer was used to evaluate foot cruited. strength, as previously described in patients affected by CMT.18 Briefly, the dynamometer was fixed on a board po- Table I reports the demographic and clinical character- sitioned on a platform in the best possible position as to istics, as well as patient’s performances according to dif- test foot strength avoiding at the same time the bias of the ferent tests. Patients had a mean duration of CMT disease examiner’s influence. The patient laid down with the foot of 30 years without significant differences between males against the myometer which he had either to push (foot and females (P>0.05) and a median CMTNS of 10 (IQR:7- dorsiflexion) or pull (plantar flexion). Each movement was 12), which corresponds to the mild to moderate disability performed 3 times on both sides and held for 3 seconds, category.40 BMI was in a range of normal weight (24.8±4.7 with a 60 second rest between each movement. The mean kg) according to World Health Organization standards. of the 3 scores was considered. Lower limbs strength demonstrates a mean of 44.1 N for dorsiflexion and a mean of 73.8 N for plantiflexion. We also included the SF36, a subjective evaluation of the QoL, to clarify whether the expected improvement in gait and functional parameters, as seen in our preliminary study, may lead to a positive impact on patient’s daily life. Vol. 55 - No. 1 European Journal of Physical and Rehabilitation Medicine 49

This document is protected by international copyright laws. No additional reproduction is authorized. It is permitted for personal use to download and save only one file and print only one copy of this Article. It is not permitted to make additional copies (either sporadically COPYRIGHT© 2019 EDIZIONI MINERVA MEDICA or systematically, either printed or electronic) of the Article for any purpose. It is not permitted to distribute the electronic copy of the article through online internet and/or intranet file sharing systems, electronic mailing or any other means which may allow access to the Article. The use of all or any part of the Article for any Commercial Use is not permitted. The creation of derivative works from the Article is not permitted. The production of reprints for personal or commercial use is not permitted. It is not permitted to remove, MORI CHARCOT-MARIE-TOOTH 1A EVALUATION cover, overlay, obscure, block, or change any copyright notices or terms of use which the Publisher may post on the Article. It is not permitted to frame or use framing techniques to enclose any trademark, logo, or other proprietary information of the Publisher. Table I.—C linical and demographic characteristics of CMT1A Table II.—S F36 scores of CMT1A patients (N.=53). patients. SF36 components Mean±SD Patients’ characteristics Value Physical functioning 63.6±26.4 Role, Physical 60.8±42 Age at examination, years 52.1 (11.9)  (range: 19.3-73.4) Bodily pain 61.5±26.6 Sex, female, N. (%) 32 (60.4%)  General health 53.5±20 Disease duration, years 30.2±18.4  (range: 3-68) Vitality BMI, kg/m2 Social functioning 53±18.3 CMTNS 24.8±4.7  (range: 17.4-48.2) Role emotional 66±24 10MWT, s 10 (7-12)  (range: 1-20) Mental health 60.9±41.8 6MWT, m PCS 65.2±20.2 Walk12 8.4±2.5  (range: 4.7-18) MCS 42.6±9.8 BBS 430.2±87.8  (range: 250-652.4) 48.7±9.6 SPPB 29 (20-39)  (range: 12-51) Foot dorsal, N 50 (46-54)  (range: 30-56) SF36: Medical Outcomes Study – Short Form 36; PCS: Physical Composite Foot plantar, N Score; MCS: Mental Composite Score. 9 (7-11)  (range: 5-12) 44.1±32.2  (range: 0-142) 73.8±51.2  (range: 0-216) Data are reported as mean±SD and range or median (interquartile range), unless walking ability (6MWT and 10MWT) showed a moderate- otherwise stated. ly strong and significant negative correlation. Correlations BMI: Body Mass Index; CMTNS: Charcot-Marie-Tooth Neuropathy Score; were also detected between these measurements and BBS, 10MWT: 10-Meter Walk Test; 6MWT: 6-Minute Walk Test; Walk12: walking SPPB and Walk12 (Figure 1). The CMTNS and the disease ability assessment; BBA: Berg Balance Scale; SPPB: Short Physical Performance duration showed only low-mild correlation with SPPB and Battery. Walk12, but not with all the other OMs. Table II shows all SF36 scores. These were slightly Correlations between components of SF36 and all clini- lower once compared to the normative data of Italian Pop- cal performances are shown in Table IV. ulation, all eight components of SF36 and the total Physi- cal Composite Score (PCS) (CMT1A patients: 42.6±9.8; The Physical Functioning parameter showed significant Italian: 50.01±10.25).41, 42 On the other hand, the Mental correlation with all motor tests. The only significant and Composite Score (MCS) was similar (CMT1A patients: moderate correlation of the Role Physical parameter were 48.7±9.6; Italian: 46.63±10.75). with 10MWT, Walk12 and SPPB. The other components of SF36 lack any correlation, except with the Walk12, which Table III reports the correlations between different showed a moderate negative correlation with all subscales evaluations. As expected, the OMs specifically related to Table III.—Correlations between different tests. Tests CMTNS* BMI 10MWT Walk12* 6MWT BBS* SPPB* Foot dorsal Foot plantar Disease duration -0.18 0.17 0.02 0.39 -0.29 -0.09 -0.21 0.14 0.11 CMTNS P=0.40 P=0.40 P=0.91 (P=0.0062 P=0.072 P=0.53 P=0.16 P=0.37 P=0.56 BMI 10MWT -0.09 0.26 -0.026 0.11 -0.17 -0.35 -0.27 -0.28 Walk12 P=0.57 P=0.084 (P=0.86 P=0.52 P=0.29 P=0.02 P=0.09 P=0.12 6MWT BBS 0.23 0.22 -0.37 -0.31 -0.15 0.35 0.33 SPPB P=0.12 (P=0.15 P=0.029 P=0.051 P=0.29 P=0.041 P=0.061 Foot dorsal 0.39 -0.63 -0.64 -0.55 -0.50 -0.34 (P=0.01 P<0.001 P<0.001 P<0.001 P=0.0017 P=0.026 -0.45 -0.41 -0.47 -0.30 -0.13 P=0.0021 P=0.0051 P=0.0021 P=0.039 P=0.40 0.52 0.35 0.32 0.004 P<0.001 P=0.040 P=0.051 P=0.98 0.59 0.47 0.41 P<0.001 P=0.0013 P=0.007 0.39 0.38 P=0.011 P=0.017 0.72 P<0.001 P values for significance of correlation coefficients were corrected for multiple tests according to the false-discovery rate approach. CMTNS: Charcot-Marie-Tooth Neuropathy Score; BMI: Body Mass Index; 10MWT: 10-Meter Walk Test; Walk12: walking ability assessment; 6MWT: 6-Minute Walk Test; BBA: Berg Balance Scale; SPPB: Short Physical Performance Battery. *For these variables, Spearman’s rank correlation coefficient is reported. 50 European Journal of Physical and Rehabilitation Medicine February 2019