- Email: [email protected]
Effects of Mobilization and Tactile Stimulation on Chronic Upper-Limb Sensorimotor Dysfunction After Stroke
Archives of Physical Medicine and Rehabilitation journal homepage: www.archives-pmr.org Archives of Physical Medicine and Rehabilitation 2013;94:693-702
ORIGINAL ARTICLE
Effects of Mobilization and Tactile Stimulation on Chronic Upper-Limb Sensorimotor Dysfunction After Stroke Jacqueline M. Winter, PhD, Peter Crome, DSc, Julius Sim, PhD, Susan M. Hunter, PhD From the Research Institute for Social Sciences, Keele University, Keele, Staffordshire, United Kingdom.
Abstract Objective: To explore the effects of Mobilization and Tactile Stimulation (MTS) and patterns of recovery in chronic stroke (>12mo) when upper limb (UL) “performance” has reached a clear plateau. Design: Replicated single-system experimental study with 8 single cases using A-B-A design (baseline-intervention-withdrawal phases); length of baseline randomly determined; intervention phase involved 6 weeks of daily MTS to the contralesional UL. Setting: Community setting, within participants’ place of residence. Participants: Individual stroke survivors (NZ8; male-to-female ratio, 3:1; age range, 49e76y; 4 with left hemiplegia, 4 with right hemiplegia) discharged from ongoing therapy, more than 1 year post stroke (range, 14e48mo). Clinical presentations were varied across the sample. Interventions: Participants received up to 1 hour of daily (Monday to Friday) treatment with MTS to the UL for 6 weeks during the intervention (B) phase. Main Outcome Measures: Motor function (Action Research Arm Test [ARAT]) and motor impairment (Motricity Index [MI] arm section) of the UL. Results: UL performance was stable during baseline for all participants. On visual analysis, improvements in motor impairment were seen in all participants, and clinically significant improvements in motor function were seen in 4 of 8 participants during the intervention phase. Latency between onset of intervention and improvement ranged from 5 to 31 days (ARAT) and from 0 to 28 days (MI). Improvements in performance were maintained on withdrawal of the intervention. Randomization tests were not significant. Conclusions: MTS appears to improve UL motor impairment and functional activity many months, even years, after stroke onset. Improvement can be immediate, but more often there is latency between the start of intervention and improvement; recovery can be distal to proximal. Archives of Physical Medicine and Rehabilitation 2013;94:693-702 ª 2013 by the American Congress of Rehabilitation Medicine
Worldwide, 15 million people per annum have a stroke, of whom 5 million die and another 5 million have permanent disability.1 Further, stroke is the largest cause of major disability in the United Kingdom (UK).1 Upper limb (UL) dysfunction is a leading cause of loss of independence in stroke survivors.2 Rehabilitation for the hemiplegic UL is frequently short-term and limited by resources that tend to be focused on regaining balance and mobility so as to enable a more general functional recovery.3 Most recovery is reported to take place in the first 3 months after stroke,4 and patients with severe UL dysfunction are unlikely to recover high levels of manipulative skills5-7 useful for functional ability. Long-term rehabilitation for stroke survivors is No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the authors or on any organization with which the authors are associated.
uncommon, particularly beyond 6 months; however, the neuroplasticity literature suggests that the potential for sensorimotor recovery from stroke can continue across the lifetime.8 Therapy for the hemiplegic UL is varied. Where there is voluntary motor activity, approaches to treatment that involve repetition and practice of functional tasks, such as repetitive task training and the shaping activities involved in constraintinduced movement therapy,9 have been shown to improve UL motor impairment and functional activity. However, these approaches are not suitable for the many stroke survivors who have insufficient voluntary motor activity in that limb. Similarly, there are a number of stroke survivors who may have sufficient voluntary motor activity, or some functional ability in the limb, but who do not use their UL spontaneously in function.7 Handson therapeutic interventions, such as those used by most UK physiotherapists working in stroke, are complex therapeutic
0003-9993/13/$36 - see front matter ª 2013 by the American Congress of Rehabilitation Medicine http://dx.doi.org/10.1016/j.apmr.2012.11.028
694
J.M. Winter et al
interventions, many of which have not been evaluated in robust trials.10 Mobilization and Tactile Stimulation (MTS) is a module of routine therapy currently used in clinical practice to treat the contralesional UL after stroke.11 It is a complex hands-on therapeutic intervention12 that has been identified by expert neurophysiotherapists in the UK as a part of routine therapy. While not a novel intervention, MTS is a discrete module of therapy that has been modeled, described clearly, and its content summarized in a published treatment schedule,11 and generalizability of this standardized schedule has been established (S.M. Hunter, PhD, unpublished data, January 2013). MTS involves hands-on sensorimotor stimulation to the forearm and hand.13 Physiotherapeutic techniques, such as passive and accessory movements, cutaneous stimulation and proprioceptive feedback, active-assisted and active movement, and facilitation or guiding of functional patterns of movement, are individual components of MTS and are delivered in an appropriate combination. The selection of such an appropriate combination is based on the clinical reasoning of a skilled therapist according to patient presentation.11 A proof-of-principle phase I study12 of MTS demonstrated potential benefits of MTS in improving motor impairment (measured by the Motricity Index [MI] arm section) and functional ability (measured by the Action Research Arm Test [ARAT]) in the contralesional UL.14 A subsequent randomized, single-blind, phase I dose-modeling trial of MTS recommended a dose of 60 minutes daily in preference to a dose of 30 minutes, 120 minutes, or no MTS, in addition to a program of routine therapy.13 The potential effect of MTS appears to be one of priming13 the sensorimotor system for activity through sensory stimulation; mobilization of joints, soft tissues, and body segments provides proprioceptive stimulation, and cutaneous stimulation provides tactile, mechanical (pressure, stretch), and proprioceptive stimulation through mechanoreceptors in the glabrous (nonhairy) skin of the hand.15 Thus, the rationale is one of priming and/or augmenting activity in the motor execution system to facilitate the voluntary contraction of paretic muscle.13 Part of the modeling process12,16 involves the identification of appropriate target groups. While evidence suggests that MTS may be effective in subacute stroke,11,14 other subgroups of stroke survivors may also benefit. The aim of this study, therefore, was to explore the effects of MTS in chronic stroke (>12mo) when UL performance had reached a clear plateau.
Methods We used an exploratory, replicated, single-system, A-B-A randomized, multiple baseline design (also known as randomized n-of-1 design16) to identify individual responses to MTS over time in stroke survivors living with a dysfunctional contralesional UL. Single-system experimental design has been described as an accepted and appropriate means of evaluating clinical change.17-19
List of abbreviations: ARAT MI MRC MTS UK UL
Action Research Arm Test Motricity Index Medical Research Council Mobilization and Tactile Stimulation United Kingdom upper limb
Direct replication of a single-system experiment that follows a predictable pattern and produces the same result on at least 3 or 4 occasions is strong evidence of a causal relationship.18,20 Moreover, the accumulation of results across participants strongly increases the generalizability of the findings.18
Ethics The North Staffordshire Health Authority Local Research Ethics Committee granted approval for this study.
Inclusion and exclusion criteria We included adult stroke survivors, men and women 18 years or older (no upper age limit), if they (1) had observable contralesional UL dysfunction of at least 12 months’ duration; (2) had been discharged from ongoing therapy; and (3) were able to follow a simple 1-stage command using the nonparetic UL (eg, “place your hand on your head”), suggesting sufficient cognitive and communication ability to understand the study and to give consent. We excluded patients who had UL dysfunction caused by other pathologic disorders unrelated to stroke (eg, musculoskeletal disorders of the shoulder girdle), as this could confound treatment response, and those with any unstable medical condition. We used a purposive sampling strategy to ensure an equal number of left and right hemisphere lesions. We recruited participants from the follow-up stroke clinic at a local hospital, and invited additional volunteers from local stroke support groups to contact us if they wanted to participate. At the start and end of the study, for the purpose of generating potential hypotheses for future studies, we recorded individual Barthel Index21 scores as an indication of independence in activities of daily living, and individual Star Cancellation Test22 scores, to screen for unilateral spatial neglect. While a score of <44 is indicative of unilateral sensory neglect in an older population,23 the score for the Star Cancellation Test was not used as a threshold for exclusion.
Outcome measures We used the following outcome measures daily throughout all phases of the study to record performance: The MI has been reported to be a valid and reliable measure of motor impairment after stroke.24,25 It is sensitive to change and has high correlation with dynamometry.26 For the arm section, participants should be seated upright,27 and an assessment of power and active range of movement is made in the 3 subsections of pinch grip, elbow flexion, and shoulder abduction. Scoring is based on the Medical Research Council (MRC) grades for muscle power, but weighted scores are used.27 Ordinal scores for each subsection are summated, giving interval-level scores for the limb. Weighted scores for each subsection range from 0 to 33 (rather than 0e5 in the MRC scale). Table 1 details the scoring criteria. The ARAT28 is a performance test measuring gross motor function and prehension. It is sensitive,29 valid,30 and reliable.28,31 The overall test is divided into 4 subsections: grasp, grip, pinch, and gross motor skills. Each subsection is scored individually to provide an ordinal-level score. Subsection scores are summated to give a maximum score for the limb, providing www.archives-pmr.org
Mobilization and tactile stimulation in chronic stroke Scoring criteria for the MI arm section27
Table 1
695 Table 2
Subsections of the ARAT and the ARAT scoring system28
Pinch Grip Score
Criterion
Subsection
0 11 19
Grasp
33
No movement Beginnings of prehension Grips 2.5-cm cube between thumb and forefinger but unable to hold against gravity (ie, drops cube when lifted) Grips 2.5-cm cube, held between thumb and forefinger, against gravity, but not against weak pull (ie, can hold the cube in the air but it is easily dislodged) Grips 2.5-cm cube, held between thumb and forefinger, against pull, but weaker than the other side Normal pinch grip
Score
Criterion
0 9 14 19
No movement Palpable contraction in muscle but no movement Movement seen but not full range/not against gravity Movement full range against gravity but not against resistance Movement against resistance but weaker than the other side Normal power
22
26
Elbow Flexion and Shoulder Abduction
25 33
data that can be assumed to have interval properties.31 Table 2 details the scoring criteria. Throughout the study, we adopted a standardized approach to the ARAT.32 An increase of at least 5.7 points on the ARAT scale (a 10% increase) has been considered as clinically significant in previous studies.33,34 A portable ARAT box was custom-made according to the original description of the equipment.28
Procedure We gave a full explanation and information sheet to all interested volunteers, and a minimum of 24 hours to fully consider any decision to participate35 before gaining written informed consent and undertaking screening. We informed the relevant medical team of the individual’s participation. Data were collected over 14 months. The lead researcher recorded all outcomes and provided all therapy except during periods of absence, when an alternate researcher covered this. Both researchers were qualified physiotherapists with extensive experience of working in stroke rehabilitation and trained in using the assessment tools, thus competent in delivering the therapy, recording its content on the treatment schedule,11 and scoring all outcome measures. We collected data daily (Monday to Friday) in the participant’s place of residence, attempting to record measurements at the same time of day to avoid instability resulting from diurnal variation.18,36 Baseline phase We randomly allocated the length of the baseline (A1) phase, as this was essential for the randomization tests used to analyze the data from the ARAT and MI scores. Before any recruitment, an independent researcher used a computer-generated list www.archives-pmr.org
Task
Maximum Score*
Pick up a 10-cm wooden cube 18 Pick up a 2.5-cm wooden cube Pick up a 5-cm wooden cube Pick up a 7.5-cm wooden cube Pick up a cricket ball Pick up a stone 102.51cm Grip Pour water from glass to glass 12 Pick up a 2.25-cm tube Pick up a tube 116cm Place 3.5-cm washer over bolt Pinch Pick up a ball bearing between 3rd 18 finger and thumb Pick up a marble between 1st finger and thumb Pick up a ball bearing between 2nd finger and thumb Pick up a ball bearing between 1st finger and thumb Pick up a marble between 3rd finger and thumb Pick up a marble between 2nd finger and thumb Gross movement Place hand behind head 9 Place hand on top of head Take hand to mouth Total score 57 * Scoring criteria for each item: 0, unable to complete any part of the task; 1, able to perform part of the task; 2, completed task but with great difficulty or took an extraordinary length of time; 3, completed task normally.
(http://www.random.org/nform.html) to randomly assign participants to length of baseline, ranging from 10 to 15 working days (Monday to Friday). Numbers on the list were placed in an opaque envelope and sequentially allocated to each participant recruited. Both the treating therapist and participant were blinded to length of baseline, and opened the envelope on day 9 of the baseline (A1) phase. We recorded baseline measurements of performance daily for the allocated duration of this phase, up to a maximum of 15 days. Intervention phase The intervention (B) phase began immediately after the baseline phase and lasted for 6 weeks, with data collected on Monday to Friday each week (5d), thus generating 30 data points (ARAT and MI scores). After daily measurement of performance, participants received up to 1 hour of treatment with MTS to the hemiplegic UL. The exact content of each therapy session was recorded on a treatment sheet.11 Withdrawal phase The withdrawal (A2) phase began immediately after the intervention (B) phase and lasted for 2 weeks (10d of data collection, Monday to Friday each week), generating 10 data points (ARAT and MI scores). We withdrew all MTS intervention during this phase, requesting all participants not to attempt to
696
J.M. Winter et al
perform this treatment themselves. On completion of this phase, participants gave assurances that they had complied with this instruction.
Data analysis Single-system studies are analyzed primarily by visual inspection of plotted data points, to determine observable changes in level (change in value of the performance measure at the point of transition between phases), trend (direction in which the series of scores is progressing), slope (gradient of the trend), and variability (fluctuation of scores within any phase) of the data between phases.18 Change scores were calculated as the difference between stable baseline scores and the maximum score attained in the intervention phase. Confidence in visual interpretation of data is enhanced if it agrees with statistical analysis.18 However, statistical analysis of Table 3
single-system studies is limited because of the potential for autocorrelation between the data points.37 Because of the chronicity (12mo) of UL dysfunction in all participants, and the knowledge that regular therapy had long been discontinued, we anticipated baselines would be stable and potentially constant. This precluded the use of statistical analyses such as the c statistic,37 since constant data in the baseline prevent the necessary calculation.19 Interrupted time-series analysis, based on an autoregressive moving average approach, requires large numbers of data points37dfor example, 50 to 100 points per data collection phase.38 Such analysis was not possible because the number of data points collected per phase in this study was much smaller than this. Randomization tests are an alternative, nonparametric test that can be applied to single or large group studies when the assumptions of parametric tests are untenable.39,40 In
Participant profiles on entry to the study
Time Initial Participant Age Paretic Post Stroke Barthel No. Sex (y) Side (mo) Score Stroke Type/CT Scan Report 1
M
61
L
18
9
2
F
76
R
14
17
3
M
61
R
27
15
4
M
49
L
36
16
5
M
67
R
48
9
6
M
58
L
36
17
7
F
68
R
48
18
8
M
65
L
14
12
Initial UL Presentation
Partial anterior circulation stroke
Increased tone throughout UL. Loss of range of movement throughout all UL joints. Self-reported pain and stiffness of shoulder and pain in hand. NA Nearly full range of movement with no signs of increased tone but limb described as “heavy” and “dragging me down.” Total anterior circulation stroke. Significant increased tone in hand, Infarction in territory of middle overuse of shoulder in all attempts at cerebral artery movement. No attempt to use UL in functional tasks. Severe dysphasia. NA Significant increased tone in hand, overuse of shoulder, attempts made at use of limb in functional tasks, severe stiffness felt on passive movements of all finger joints and wrist joint. NA Increased tone in hand, painful shoulder, no attempts to use hand functionally, loss of range of movement in all joints of hand and wrist. Severe dysphasia. Hemorrhagic Overuse of shoulder, decreased strength throughout UL, very poor coordination, limb easily tires with activity, marked loss of sensation and altered sensation including proprioception, hot/cold, and light touch. Cerebral infarct Using arm well in functional activity but reports “heavy” feeling with altered sensation and pain. Very little light touch sensation or temperature sensation and poor proprioception. Mild dysphasia. Infarct over left cerebellar hemisphere Pain in shoulder, incipient return of active and right basal ganglia; lacunar movement in hand, wrist, and elbow, infarct over left external capsule increased tone in hand with loss of range of movement at wrist joint.
Abbreviations: CT, computed tomography; F, female; L, left; M, male; NA, not available; R, right.
www.archives-pmr.org
Mobilization and tactile stimulation in chronic stroke single-system studies, these tests can be used to identify the difference between mean values in different phases of the study, as long as the point at which intervention begins is randomly allocated.39 We performed these tests using appropriate syntax in SPSS version 14.0.a The syntax calculated a “count of arrangements”dthe number of possible permutations or arrangements that would produce a mean difference greater than or equal to the observed mean difference. On this basis, the statistical significance of the observed mean difference can be determined. Accordingly, we based supporting statistical data analysis on randomization tests between the baseline (A1) phase and the intervention (B) phase alone. Statistical significance was set at P.05.
Results All 8 participants completed all phases of the study series with no adverse effects. This was considered to be a positive result. Table 3 summarizes individual participant profiles on entry to the study. Absence of computed tomography scan results accounts for the missing data in table 3. Table 4 shows the mean amount of treatment received by each participant during the intervention (B) phase.
Motricity Index (arm section) Figure 1 shows the total MI arm section scores for all participants. Visual inspection of total MI scores across phases showed a positive change in trend and change in slope for all participants. All participants showed a period of latency before a change in total score, ranging from 0 days (participant 8) to 28 days (participant 1) after the introduction of the intervention. Improvements occurred in the pinch subsection for 7 participants, and all participants improved in either or both of the elbow and shoulder subsections, from stable baseline to the end of the intervention phase. Total change scores ranged from 9 to 37. Table 5 summarizes MI subsection and total change scores for each participant. In summary, total MI arm section scores improved for all participants in the intervention phase; no further improvement or deterioration was evident in the withdrawal phase.
Table 4 Mean treatment time per participant during the intervention phase
Participant No.
Total Intervention Time (h)
Range of Treatment Session Times (min)
Mean Treatment Session Time (min)
1 2 3 4 5 6 7 8
21.46 12.50 18.33 19.25 18.41 17.91 16.08 16.83
30e60 30e40 25e45 30e50 30e55 35e40 30e55 30e40
51.40 35.70 37.93 42.77 44.20 43.00 41.95 36.07
www.archives-pmr.org
697
Action Research Arm Test Figure 2 shows the total ARAT scores for each participant. Visual inspection of the plotted data showed a positive change in trend and a change in slope for 5 participants; participant 2 had attained the maximum score during the baseline phase and so could not improve further. All participants showed a period of latency before a change in total score was evident. These periods of latency ranged from 5 days (participant 7) to 31 days (participant 3) after introduction of the intervention. Table 6 summarizes the ARAT subsection and total change scores for each participant. Improvements were seen in at least 1 subsection for 7 participants. Total change scores ranged from 1 to 24. No improvement was seen for participant 2. Change scores reflected a greater magnitude of improvement in combined distal activity scores (pinch, grip, grasp) than in gross movement scores. However, a greater number of participants showed some improvement, albeit small, in gross movement scores. In conclusion, while total ARAT scores improved for 7 participants in the intervention phase, we only consider the changes to be clinically significant in 4 of the 8 participants (participants 1, 6, 7, and 8). No further improvement or deterioration was evident in the withdrawal phase in 7 patients, but 1 participant (6) showed deterioration in scores during this final phase. While we found a positive change in mean difference in scores between the baseline (A1) and intervention (B) phases for all but 1 participant (participant 2), randomization tests showed that the level of change was not statistically significant.
Variability There was little variability in the scores achieved by the participants. Negative dips in the MI (participants 3, 4, and 5) and ARAT (participants 4 and 5) scores, seen in figures 1 and 2, coincided with and could potentially be accounted for by changes to anticonvulsive medication (participant 3), poorly controlled blood glucose levels (participant 4), and the presence of flu symptoms (participant 5), which were reported in field notes.
Discussion We recruited participants to this study on the basis of their having persistent UL dysfunction at least 12 months after the onset of their stroke. ARAT scores for all participants remained stable throughout baseline, confirming that recovery in the UL at this stage post stroke appeared to have stabilized, with no further spontaneous recovery. Therefore, since further spontaneous improvement in performance was highly unlikely for this stage post stroke, we could be more confident that any changes in performance during the intervention phase were attributable to the intervention, particularly if this result was replicated across cases. Total MI (arm section) and ARAT scores achieved during the initial baseline (A1) phase did indeed remain consistent, reflecting stability of impairment and activity limitation. However, all 8 participants subsequently showed improvement in UL motor impairment (MI arm section) with the application of the intervention, and 4 participants demonstrated clinically significant improvement in ARAT scores, suggesting an effect of the MTS. Of particular interest is the finding that distal activity (MI pinch; ARAT pinch, grip, grasp) appeared to improve more than
J.M. Winter et al
Motricity Index arm section total scores
698
Day Fig 1
Total MI arm section scores for participants 1 through 8.
proximal/gross activity. This is perhaps unsurprising since MTS intervention is targeted at the hand; however, this provides evidence that distal recovery can occur independently of proximal
recovery. Nevertheless, it is also of interest that a distal-focused intervention also appeared to impact positively on proximal recovery in the chronic stage post stroke. www.archives-pmr.org
Mobilization and tactile stimulation in chronic stroke
699
Table 5 Summary of visual changes in total MI (arm) subsection scores and improvements in subsection scores for each participant, showing scores at end of baseline phase and maximum score attained during intervention phase Participant No.
Change in Trend
Change in Slope
Improvement in Pinch Subsection
Improvement in Elbow Subsection
Improvement in Shoulder Subsection
Total No. of Points Increased
1 2 3 4 5 6 7 8
Yes Yes Yes Yes Yes Yes Yes Yes
Yes Yes Yes Yes Yes Yes Yes Yes
19 7 4 4 15 0 7 4
0 8 0 5 5 0 19 5
11 0 6 0 0 11 11 5
30 15 10 9 20 11 37 14
(0e19) (26e33) (22e26) (22e26) (11e26) (26e26) (26e33) (22e26)
(14e14) (25e33) (14e14) (14e19) (9e14) (25e25) (14e33) (14e19)
(14e25) (25e25) (19e25) (19e19) (9e9) (14e25) (14e25) (14e19)
NOTE. The 2 figures in parentheses represent, respectively, the score at end of the baseline phase and the maximum score in the intervention phase; the improvement score is calculated by the subtraction of the first of these from the second.
When considering overall ARAT and MI (arm section) results in the context of participant profiles, the greatest change in ARAT score was achieved by participants presenting with self-reported sensory deficits (participants 6 and 7). In contrast, the least change in ARAT score was achieved by participants with marked distal increased tone or spasticity of the UL (participants 3, 4, and 5). However, this pattern is not replicated in the MI results, and this requires further exploration. Participants with higher initial scores did not appear to make greater improvements than those with lower entry-level scores on either the ARAT or the MI. No obvious patterns relating initial scores to final scores were seen. Therefore, we suggest that entrylevel score cannot be used as a prognostic indicator of outcome. The relationship between site of lesion or type of stroke and UL recovery in response to MTS cannot be commented on from the data generated from this single-system study. However, these variables warrant further investigation to identify the optimum target population for a larger trial of MTS after stroke. Immediate change in performance was only seen in 1 participant (8) (change of level of MI score), while change was delayed in all other participants. The length of latency ranged from 0 to 28 days for MI scores and 5 to 31 days for ARAT scores. All improvements followed a stepwise pattern, with periods of no improvement (plateau) followed by improvement. The duration of these periods varied between participants. In light of the recovery patterns seen, the inclusion of a second intervention (B) phase (A-B-A-B design) might have added further information on the emerging patterns of UL recovery. Overall, scores achieved throughout the B phase remained unchanged during the withdrawal phase, apart from participant 6, whose ARAT score decreased by 3 points at the end of this phase. This participant had reported high levels of sensory impairment, and further exploration of the impact of sensory impairment in the hand on recovery is warranted. Understanding patterns of recovery after stroke is fundamentally important to therapists to ensure that they design rehabilitation programs that maximize opportunities for recovery. All participants in this study improved with MTS, and a clear, common pattern of recovery emerged: a prolonged initial period of latency followed by stepwise periods of improvement and stabilization/plateauing. This pattern was the same for all participants, regardless of clinical presentation. This knowledge is important for clinicians because key decisions can be made regarding ongoing treatment plans, and goal setting that reflects www.archives-pmr.org
these periods of recovery and latency can be supported. Furthermore, it may inform the timing of discharge from ongoing therapy. The intensity of this therapy and the sensory bombardment of the central nervous system required participants to focus attention on the treated hand; this involved great cognitive effort, which may be a challenge to some patients receiving this intervention. However, for the participants in this study, who were able to maintain focus, positive outcomes were seen.
Study limitations The risk of subjectivity and poor reliability in visual interpretation of data has been acknowledged41; hence the recommendation that, wherever possible, statistical tests are used to support visual analysis. Visual inspection of the data has also been reported to be limited in that only effects that can be clearly and easily seen as significant are considereddsuch that weak but positive treatment effects may be overlooked17dand serial dependency of the data may affect their interpretation.42 These are further reasons not to rely on visual inspection alone. In contrast, it can be argued that changes that are sufficiently marked to be discernible through visual analysis are likely to be clinically significant,19 and statistical tests may find a statistically significant change that is of no clinical significance. Therefore, clinical significance of changes in performance must be considered alongside any statistical significance of any change in performance score. Randomization tests were used in this study to support visual analysis. However, there are difficulties in using statistical tests that rely on differences in mean scores when there are long periods of latency of recovery and when recovery occurs late in the intervention phase. If large changes in score constitute the last few data points of the intervention phase after a long period of latency, they will have little influence on the mean score in this phase, and the between-phase statistical test may therefore be nonsignificant. However, while this may seem to be weak evidence of a treatment effect, clinical experience and participant reports suggest that the changes were clinically significant, and the use of statistical analysis may not be essential to infer an effect.39 We chose not to include a measure of UL sensation as an outcome measure because of reported unreliability of the various tests.14 This, however, was a limitation of the study, particularly since 2 participants (6 and 7) with marked self-reported sensory deficits appeared to make early and great improvements. Future studies of MTS should include a measure of sensory impairment.
J.M. Winter et al
ARAT total scores
700
Day Fig 2
Total ARAT scores for participants 1 through 8.
The potential for investigator and measurement bias is considered as a potential limitation to single-system studies. In a single-system study, it is not appropriate to blind the assessor to
allocation to a treatment group because treatment groups are not part of the design. However, if the assessor had been blind to the stage of the design (A or B phases) and the point of transition www.archives-pmr.org
Mobilization and tactile stimulation in chronic stroke
701
Table 6 Summary of visual changes in total ARAT scores and improvements in scores for ARAT subsections for each participant, showing score at end of baseline phase and maximum score attained in intervention phase Participant No.
Change in Trend e Total Score
Change in Slope e Total Score
Improvement in Pinch Subsection
Improvement in Grip Subsection
Improvement in Grasp Subsection
Improvement in Gross Movement Subsection
Total No. of Points Increased
1 2 3 4 5 6 7 8
Yes No No No Yes Yes Yes Yes
Yes No No No Yes Yes Yes Yes
3 0 0 0 0 4 11 4
2 0 0 0 1 4 4 3
2 0 0 0 0 5 6 4
1 0 1 2 2 2 3 1
8 0 1 2 3 15 24 12
(6e9) (18e18) (6e6) (6e6) (0e0) (13e17) (6e17) (2e6)
(4e6) (12e12) (4e4) (4e4) (0e1) (8e12) (8e12) (4e7)
(6e8) (18e18) (6e6) (6e6) (0e0) (13e18) (12e18) (7e11)
(3e4) (9e9) (3e4) (4e6) (1e3) (7e9) (6e9) (3e4)
NOTE. The 2 figures in parentheses represent, respectively, the score at end of the baseline phase and the maximum score in the intervention phase; the improvement score is calculated by the subtraction of the first of these from the second.
between phases, this would have increased the internal validity of the results. However, blinding was not feasible within the resources for this study. Most single-system studies include a phase in which the intervention is withdrawn,17 and this is sometimes considered to be reflective of the withdrawal of treatment that occurs as a matter of course in clinical practice, which is frequently after a course of 6 weeks. The withdrawal phase serves a purpose in identifying whether changes in performance are maintained when the intervention ceases, a potential measure of the carryover effects of therapy. Withholding or withdrawing treatment in single-system studies nonetheless has potential ethical implications.43 In this study, however, all participants had been discharged from therapy, so there was no issue of denying routine intervention. On this basis, and in the light of the consent provided by the participants, we considered that withdrawal of treatment was justifiable.
Conclusions This exploratory study has generated hypotheses for further study and identified several important clinical messages:
these improvements can be carried over and maintained for at least 2 weeks after withdrawal of the intervention. We now need to explore the long-term effects of MTS on UL recovery. The application of MTS, a hands-on intervention used in routine clinical practice to prime and/or augment the motor execution system in order to enhance or facilitate voluntary activation of muscle, appears to provide beneficial effects with no negative side effects. Not only should it continue to be used in routine therapy practice after stroke for patients with varying clinical presentations, its effectiveness should be explored in ongoing studies. The results of this study contribute, along with other preliminary studies of MTS, to the identification of an appropriate target population and dose of intervention for future trials of MTS. The ability of ongoing improvements in UL performance to be made and sustained more than a year post stroke suggests that therapists should give consideration to where and when their interventions are best provided. They may need to reflect on whether services are really best placed in the acute hospital environment or whether future service redesigns may be better focused on rehabilitation in the community.
A 6-week daily program of therapist-led mobilization of joints and soft tissues coupled with tactile stimulation of the hand and forearm (MTS)da module of conventional therapydappears to improve motor impairment and motor activity. Clinically significant recovery of persistent UL dysfunction is possible even 12 months or more post stroke. Distal UL recovery can occur with targeted treatment independently of proximal recovery after stroke. MTS as a distally focused intervention may still impact positively on proximal recovery. The pattern of recovery at this stage is likely to be characterized by periods of improvement followed by plateaus of performance. A period of latency after the commencement of MTS is to be expected. The effectiveness of sensory stimulation in treatment, while as yet unclear, should not be dismissed.
Supplier
These replicated results suggest that improvements in UL performance may occur more than a year post stroke, and that
We thank West Midlands Stroke Research Network for assistance in recruitment.
www.archives-pmr.org
a. SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.
Keywords Musculoskeletal manipulations; Rehabilitation; Sensation; Stroke; Therapy; Upper extremity
Corresponding author Susan M. Hunter, PhD, School of Health and Rehabilitation, Keele University, Keele, Staffordshire, ST5 5BG, UK. E-mail address: [email protected].
Acknowledgment
702
References 1. Mackay J, Mensah G. Atlas of heart disease and stroke. Part 3: the burdendglobal burden of stroke. World Health Organization; 2004. Available at: http://www.who.int/cardiovascular_diseases/en/cvd_atlas_ 15_burden_stroke.pdf. Accessed February 21, 2012. 2. Jorgensen HS, Nakayama H, Pedersen PM, Kamersgaard L, Raaschou HO, Olsen TS. Epidemiology of stroke-related disability. Clin Geriatr Med 1999;15:785-99. 3. Carr J, Shepherd R. Neurological rehabilitation: optimizing motor performance. London: Butterworth-Heinmann; 1998. 4. Wade D, Skilbeck CE, Hewer R. Predicting Barthel ADL score at 6 months after an acute stroke. Arch Phys Med Rehabil 1983;64:24-8. 5. Wade DT, Langton-Hewer R, Wood VA, Skilbeck CE, Ismail HM. The hemiplegic arm after stroke: measurement and recovery. J Neurol Neurosurg Psychiatry 1983;46:521-4. 6. Parker VM, Wade DT, Langton-Hewer R. Loss of arm function after stroke: measurement, frequency and recovery. Int Rehabil Med 1986; 8:69-73. 7. Broeks JG, Lankhorst GJ, Rumping K, Prevo AJ. The long-term outcome of arm function after stroke: results of a follow up study. Disabil Rehabil 1999;21:357-64. 8. Nudo RJ, Friel KM. Cortical plasticity after stroke: implications for rehabilitation. Rev Neurol (Paris) 1999;155:713-7. 9. Sirtori V, Corbetta D, Moja L, Gatti R. Constraint-induced movement therapy for upper extremities in stroke patients. Cochrane Database Syst Rev 2009;(4):CD004433. 10. Winter J, Hunter S, Sim J, Crome P. Hands-on therapy interventions for upper limb motor dysfunction following stroke. Cochrane Database Syst Rev 2011;(6):CD006609. 11. Hunter SM, Crome P, Sim J, Donaldson C, Pomeroy VM. Development of treatment schedules for research: a structured review to identify methodologies used and a worked example of ‘mobilisation and tactile stimulation’ for stroke patients. Physiotherapy 2006;92: 195-207. 12. Medical Research Council (MRC). A framework for development and evaluation of RCTs for complex interventions to improve health. London: MRC Health Services and Public Health Research Board; 2000. 13. Hunter SM, Hammett L, Ball S, et al. Dose-response study of mobilisation and tactile stimulation for the upper extremity early after stroke: a phase I trial. Neurorehabil Neural Repair 2011;25: 314-22. 14. Hunter SM, Crome P, Sim J, Pomeroy VM. Effects of Mobilization and Tactile Stimulation on recovery of the hemiplegic upper limb: a series of replicated single-system studies. Arch Phys Med Rehabil 2008;89:2003-10. 15. Wing AM, Haggard P, Randall Flanagan J. Hand and brain. The neurophysiology and psychology of hand movements. San Diego: Academic Pr; 1996. 16. Medical Research Council (MRC). Developing and evaluating complex interventions: new guidance. London: MRC; 2000. Available at: http:// www.mrc.ac.uk/Utilities/Documentrecord/index.htm?dZMRC004871. Accessed September 10, 2012. 17. Kazdin AE. Single-case research designs: methods for clinical and applied settings. New York: Oxford University Pr; 1982. 18. Barlow D, Hersen M. Single case experimental designs: strategies for studying behavior change. 2nd ed. Boston: Allyn and Bacon; 1984. 19. Ottenbacher KJ. Evaluating clinical change: strategies for occupational and physical therapists. Baltimore: Williams & Wilkins; 1986. 20. Robson C. Real world research: a resource for social scientists and practitioner-researchers. 2nd ed. Oxford: Blackwell; 2002.
J.M. Winter et al 21. Mahoney FI, Barthel DW. Functional evaluation: the Barthel Index. Md State Med J 1965;14:61-5. 22. Wilson B, Cockburn J, Halligan PW. Behavioural Inattention Test. Bury St Edmunds: Thames Valley Test Company; 1987. 23. Friedman PJ. The Star Cancellation Test in acute stroke. Clin Rehabil 1992;6:23-30. 24. Collen F, Wade D, Bradshaw C. Mobility after stroke: reliability of measures of impairment and disability. Int Disabil Stud 1990;12:6-9. 25. Collin C, Wade D. Assessing motor impairment after stroke: a pilot reliability study. J Neurol Neurosurg Psychiatry 1990;53:576-9. 26. Bohannon R. Motricity Index scores are valid indicators of paretic upper extremity strength following stroke. J Phys Ther Sci 1999;11: 59-61. 27. Wade D. Measurement in neurological rehabilitation. Oxford: Oxford University Pr; 1992. 28. Lyle RC. A performance test for assessment of upper limb function in physical rehabilitation treatment and research. Int J Rehabil Res 1981; 4:483-92. 29. Wade D. Measuring arm impairment and disability after stroke. Int Disabil Stud 1989;11:89-92. 30. Hsieh C-L, Hsueh I-P, Chiang F-M, Lin P-H. Inter-rater reliability and validity of the Action Research Arm Test in stroke patients. Age Ageing 1998;27:107-13. 31. Hsueh I-P, Lee M-M, Hsieh C-L. The Action Research Arm Test: is it necessary for patients being tested to sit at a standardized table? Clin Rehabil 2002;16:382-8. 32. Donaldson C, Tallis R, Pomeroy V. Outcome measures in neurophysiotherapy for the arm and hand: have we lost our grip? Clin Rehabil 2006;20:459-60. 33. van der Lee JH, Wagenaar RC, Lankhorst GJ, Vogelaar TW, Deville´ WL, Bouter LM. Forced use of the upper extremity in chronic stroke patients: results from a single-blind randomized clinical trial. Stroke 1999;30:2369-75. 34. Donaldson C, Tallis RC, Miller S, Sunderland A, Lemon R, Pomeroy VM. Effects of conventional physical therapy and functional strength training on upper limb motor recovery after stroke: a randomised phase II study. Neurorehabil Neural Repair 2009;23:389-97. 35. Medical Research Council (MRC). MRC guidelines for good clinical practice in clinical trials. London: MRC Health Services and Public Health Research Board; 1998. 36. Sim J, Wright C. Research in health care: concepts, designs and methods. Cheltenham: Nelson Thornes; 2000. 37. Tryon W. A simplified time series analysis for evaluating treatment interventions. J Appl Behav Anal 1982;15:423-9. 38. Hartmann DP, Gottman JM, Jones RR, Gardner W, Kazdin AE, Vaught R. Interrupted time-series analysis and its application to behavioural data. J Appl Behav Anal 1980;13:543-59. 39. Todman J, Dugard P. Single-case and small-n experimental designs: a practical guide to randomization tests. Mahwah: Lawrence Erlbaum; 2001. 40. Gorman BS, Allison DB. Statistical alternatives for single-case designs. In: Franklin RD, Allison DB, Gorman BS, editors. Design and analysis of single-case research. Mahwah: Lawrence Erlbaum; 1997. p 159-214. 41. Richards SB, Taylor RL, Ramasamy R, Richards RY. Single subject research: applications in educational and clinical settings. San Diego: Singular Publishing; 1999. 42. Matyas TA, Greenwood KM. Serial dependency in single-case time series. In: Franklin RD, Allison DB, Gorman BS, editors. Design and analysis of single-case research. Mahwah: Lawrence Erlbaum; 1997. p 215-43. 43. Sim J. The ethics of single system (nZ1) research. Physiother Theory Pract 1994;10:111-22.
www.archives-pmr.org
ORIGINAL ARTICLE
Effects of Mobilization and Tactile Stimulation on Chronic Upper-Limb Sensorimotor Dysfunction After Stroke Jacqueline M. Winter, PhD, Peter Crome, DSc, Julius Sim, PhD, Susan M. Hunter, PhD From the Research Institute for Social Sciences, Keele University, Keele, Staffordshire, United Kingdom.
Abstract Objective: To explore the effects of Mobilization and Tactile Stimulation (MTS) and patterns of recovery in chronic stroke (>12mo) when upper limb (UL) “performance” has reached a clear plateau. Design: Replicated single-system experimental study with 8 single cases using A-B-A design (baseline-intervention-withdrawal phases); length of baseline randomly determined; intervention phase involved 6 weeks of daily MTS to the contralesional UL. Setting: Community setting, within participants’ place of residence. Participants: Individual stroke survivors (NZ8; male-to-female ratio, 3:1; age range, 49e76y; 4 with left hemiplegia, 4 with right hemiplegia) discharged from ongoing therapy, more than 1 year post stroke (range, 14e48mo). Clinical presentations were varied across the sample. Interventions: Participants received up to 1 hour of daily (Monday to Friday) treatment with MTS to the UL for 6 weeks during the intervention (B) phase. Main Outcome Measures: Motor function (Action Research Arm Test [ARAT]) and motor impairment (Motricity Index [MI] arm section) of the UL. Results: UL performance was stable during baseline for all participants. On visual analysis, improvements in motor impairment were seen in all participants, and clinically significant improvements in motor function were seen in 4 of 8 participants during the intervention phase. Latency between onset of intervention and improvement ranged from 5 to 31 days (ARAT) and from 0 to 28 days (MI). Improvements in performance were maintained on withdrawal of the intervention. Randomization tests were not significant. Conclusions: MTS appears to improve UL motor impairment and functional activity many months, even years, after stroke onset. Improvement can be immediate, but more often there is latency between the start of intervention and improvement; recovery can be distal to proximal. Archives of Physical Medicine and Rehabilitation 2013;94:693-702 ª 2013 by the American Congress of Rehabilitation Medicine
Worldwide, 15 million people per annum have a stroke, of whom 5 million die and another 5 million have permanent disability.1 Further, stroke is the largest cause of major disability in the United Kingdom (UK).1 Upper limb (UL) dysfunction is a leading cause of loss of independence in stroke survivors.2 Rehabilitation for the hemiplegic UL is frequently short-term and limited by resources that tend to be focused on regaining balance and mobility so as to enable a more general functional recovery.3 Most recovery is reported to take place in the first 3 months after stroke,4 and patients with severe UL dysfunction are unlikely to recover high levels of manipulative skills5-7 useful for functional ability. Long-term rehabilitation for stroke survivors is No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the authors or on any organization with which the authors are associated.
uncommon, particularly beyond 6 months; however, the neuroplasticity literature suggests that the potential for sensorimotor recovery from stroke can continue across the lifetime.8 Therapy for the hemiplegic UL is varied. Where there is voluntary motor activity, approaches to treatment that involve repetition and practice of functional tasks, such as repetitive task training and the shaping activities involved in constraintinduced movement therapy,9 have been shown to improve UL motor impairment and functional activity. However, these approaches are not suitable for the many stroke survivors who have insufficient voluntary motor activity in that limb. Similarly, there are a number of stroke survivors who may have sufficient voluntary motor activity, or some functional ability in the limb, but who do not use their UL spontaneously in function.7 Handson therapeutic interventions, such as those used by most UK physiotherapists working in stroke, are complex therapeutic
0003-9993/13/$36 - see front matter ª 2013 by the American Congress of Rehabilitation Medicine http://dx.doi.org/10.1016/j.apmr.2012.11.028
694
J.M. Winter et al
interventions, many of which have not been evaluated in robust trials.10 Mobilization and Tactile Stimulation (MTS) is a module of routine therapy currently used in clinical practice to treat the contralesional UL after stroke.11 It is a complex hands-on therapeutic intervention12 that has been identified by expert neurophysiotherapists in the UK as a part of routine therapy. While not a novel intervention, MTS is a discrete module of therapy that has been modeled, described clearly, and its content summarized in a published treatment schedule,11 and generalizability of this standardized schedule has been established (S.M. Hunter, PhD, unpublished data, January 2013). MTS involves hands-on sensorimotor stimulation to the forearm and hand.13 Physiotherapeutic techniques, such as passive and accessory movements, cutaneous stimulation and proprioceptive feedback, active-assisted and active movement, and facilitation or guiding of functional patterns of movement, are individual components of MTS and are delivered in an appropriate combination. The selection of such an appropriate combination is based on the clinical reasoning of a skilled therapist according to patient presentation.11 A proof-of-principle phase I study12 of MTS demonstrated potential benefits of MTS in improving motor impairment (measured by the Motricity Index [MI] arm section) and functional ability (measured by the Action Research Arm Test [ARAT]) in the contralesional UL.14 A subsequent randomized, single-blind, phase I dose-modeling trial of MTS recommended a dose of 60 minutes daily in preference to a dose of 30 minutes, 120 minutes, or no MTS, in addition to a program of routine therapy.13 The potential effect of MTS appears to be one of priming13 the sensorimotor system for activity through sensory stimulation; mobilization of joints, soft tissues, and body segments provides proprioceptive stimulation, and cutaneous stimulation provides tactile, mechanical (pressure, stretch), and proprioceptive stimulation through mechanoreceptors in the glabrous (nonhairy) skin of the hand.15 Thus, the rationale is one of priming and/or augmenting activity in the motor execution system to facilitate the voluntary contraction of paretic muscle.13 Part of the modeling process12,16 involves the identification of appropriate target groups. While evidence suggests that MTS may be effective in subacute stroke,11,14 other subgroups of stroke survivors may also benefit. The aim of this study, therefore, was to explore the effects of MTS in chronic stroke (>12mo) when UL performance had reached a clear plateau.
Methods We used an exploratory, replicated, single-system, A-B-A randomized, multiple baseline design (also known as randomized n-of-1 design16) to identify individual responses to MTS over time in stroke survivors living with a dysfunctional contralesional UL. Single-system experimental design has been described as an accepted and appropriate means of evaluating clinical change.17-19
List of abbreviations: ARAT MI MRC MTS UK UL
Action Research Arm Test Motricity Index Medical Research Council Mobilization and Tactile Stimulation United Kingdom upper limb
Direct replication of a single-system experiment that follows a predictable pattern and produces the same result on at least 3 or 4 occasions is strong evidence of a causal relationship.18,20 Moreover, the accumulation of results across participants strongly increases the generalizability of the findings.18
Ethics The North Staffordshire Health Authority Local Research Ethics Committee granted approval for this study.
Inclusion and exclusion criteria We included adult stroke survivors, men and women 18 years or older (no upper age limit), if they (1) had observable contralesional UL dysfunction of at least 12 months’ duration; (2) had been discharged from ongoing therapy; and (3) were able to follow a simple 1-stage command using the nonparetic UL (eg, “place your hand on your head”), suggesting sufficient cognitive and communication ability to understand the study and to give consent. We excluded patients who had UL dysfunction caused by other pathologic disorders unrelated to stroke (eg, musculoskeletal disorders of the shoulder girdle), as this could confound treatment response, and those with any unstable medical condition. We used a purposive sampling strategy to ensure an equal number of left and right hemisphere lesions. We recruited participants from the follow-up stroke clinic at a local hospital, and invited additional volunteers from local stroke support groups to contact us if they wanted to participate. At the start and end of the study, for the purpose of generating potential hypotheses for future studies, we recorded individual Barthel Index21 scores as an indication of independence in activities of daily living, and individual Star Cancellation Test22 scores, to screen for unilateral spatial neglect. While a score of <44 is indicative of unilateral sensory neglect in an older population,23 the score for the Star Cancellation Test was not used as a threshold for exclusion.
Outcome measures We used the following outcome measures daily throughout all phases of the study to record performance: The MI has been reported to be a valid and reliable measure of motor impairment after stroke.24,25 It is sensitive to change and has high correlation with dynamometry.26 For the arm section, participants should be seated upright,27 and an assessment of power and active range of movement is made in the 3 subsections of pinch grip, elbow flexion, and shoulder abduction. Scoring is based on the Medical Research Council (MRC) grades for muscle power, but weighted scores are used.27 Ordinal scores for each subsection are summated, giving interval-level scores for the limb. Weighted scores for each subsection range from 0 to 33 (rather than 0e5 in the MRC scale). Table 1 details the scoring criteria. The ARAT28 is a performance test measuring gross motor function and prehension. It is sensitive,29 valid,30 and reliable.28,31 The overall test is divided into 4 subsections: grasp, grip, pinch, and gross motor skills. Each subsection is scored individually to provide an ordinal-level score. Subsection scores are summated to give a maximum score for the limb, providing www.archives-pmr.org
Mobilization and tactile stimulation in chronic stroke Scoring criteria for the MI arm section27
Table 1
695 Table 2
Subsections of the ARAT and the ARAT scoring system28
Pinch Grip Score
Criterion
Subsection
0 11 19
Grasp
33
No movement Beginnings of prehension Grips 2.5-cm cube between thumb and forefinger but unable to hold against gravity (ie, drops cube when lifted) Grips 2.5-cm cube, held between thumb and forefinger, against gravity, but not against weak pull (ie, can hold the cube in the air but it is easily dislodged) Grips 2.5-cm cube, held between thumb and forefinger, against pull, but weaker than the other side Normal pinch grip
Score
Criterion
0 9 14 19
No movement Palpable contraction in muscle but no movement Movement seen but not full range/not against gravity Movement full range against gravity but not against resistance Movement against resistance but weaker than the other side Normal power
22
26
Elbow Flexion and Shoulder Abduction
25 33
data that can be assumed to have interval properties.31 Table 2 details the scoring criteria. Throughout the study, we adopted a standardized approach to the ARAT.32 An increase of at least 5.7 points on the ARAT scale (a 10% increase) has been considered as clinically significant in previous studies.33,34 A portable ARAT box was custom-made according to the original description of the equipment.28
Procedure We gave a full explanation and information sheet to all interested volunteers, and a minimum of 24 hours to fully consider any decision to participate35 before gaining written informed consent and undertaking screening. We informed the relevant medical team of the individual’s participation. Data were collected over 14 months. The lead researcher recorded all outcomes and provided all therapy except during periods of absence, when an alternate researcher covered this. Both researchers were qualified physiotherapists with extensive experience of working in stroke rehabilitation and trained in using the assessment tools, thus competent in delivering the therapy, recording its content on the treatment schedule,11 and scoring all outcome measures. We collected data daily (Monday to Friday) in the participant’s place of residence, attempting to record measurements at the same time of day to avoid instability resulting from diurnal variation.18,36 Baseline phase We randomly allocated the length of the baseline (A1) phase, as this was essential for the randomization tests used to analyze the data from the ARAT and MI scores. Before any recruitment, an independent researcher used a computer-generated list www.archives-pmr.org
Task
Maximum Score*
Pick up a 10-cm wooden cube 18 Pick up a 2.5-cm wooden cube Pick up a 5-cm wooden cube Pick up a 7.5-cm wooden cube Pick up a cricket ball Pick up a stone 102.51cm Grip Pour water from glass to glass 12 Pick up a 2.25-cm tube Pick up a tube 116cm Place 3.5-cm washer over bolt Pinch Pick up a ball bearing between 3rd 18 finger and thumb Pick up a marble between 1st finger and thumb Pick up a ball bearing between 2nd finger and thumb Pick up a ball bearing between 1st finger and thumb Pick up a marble between 3rd finger and thumb Pick up a marble between 2nd finger and thumb Gross movement Place hand behind head 9 Place hand on top of head Take hand to mouth Total score 57 * Scoring criteria for each item: 0, unable to complete any part of the task; 1, able to perform part of the task; 2, completed task but with great difficulty or took an extraordinary length of time; 3, completed task normally.
(http://www.random.org/nform.html) to randomly assign participants to length of baseline, ranging from 10 to 15 working days (Monday to Friday). Numbers on the list were placed in an opaque envelope and sequentially allocated to each participant recruited. Both the treating therapist and participant were blinded to length of baseline, and opened the envelope on day 9 of the baseline (A1) phase. We recorded baseline measurements of performance daily for the allocated duration of this phase, up to a maximum of 15 days. Intervention phase The intervention (B) phase began immediately after the baseline phase and lasted for 6 weeks, with data collected on Monday to Friday each week (5d), thus generating 30 data points (ARAT and MI scores). After daily measurement of performance, participants received up to 1 hour of treatment with MTS to the hemiplegic UL. The exact content of each therapy session was recorded on a treatment sheet.11 Withdrawal phase The withdrawal (A2) phase began immediately after the intervention (B) phase and lasted for 2 weeks (10d of data collection, Monday to Friday each week), generating 10 data points (ARAT and MI scores). We withdrew all MTS intervention during this phase, requesting all participants not to attempt to
696
J.M. Winter et al
perform this treatment themselves. On completion of this phase, participants gave assurances that they had complied with this instruction.
Data analysis Single-system studies are analyzed primarily by visual inspection of plotted data points, to determine observable changes in level (change in value of the performance measure at the point of transition between phases), trend (direction in which the series of scores is progressing), slope (gradient of the trend), and variability (fluctuation of scores within any phase) of the data between phases.18 Change scores were calculated as the difference between stable baseline scores and the maximum score attained in the intervention phase. Confidence in visual interpretation of data is enhanced if it agrees with statistical analysis.18 However, statistical analysis of Table 3
single-system studies is limited because of the potential for autocorrelation between the data points.37 Because of the chronicity (12mo) of UL dysfunction in all participants, and the knowledge that regular therapy had long been discontinued, we anticipated baselines would be stable and potentially constant. This precluded the use of statistical analyses such as the c statistic,37 since constant data in the baseline prevent the necessary calculation.19 Interrupted time-series analysis, based on an autoregressive moving average approach, requires large numbers of data points37dfor example, 50 to 100 points per data collection phase.38 Such analysis was not possible because the number of data points collected per phase in this study was much smaller than this. Randomization tests are an alternative, nonparametric test that can be applied to single or large group studies when the assumptions of parametric tests are untenable.39,40 In
Participant profiles on entry to the study
Time Initial Participant Age Paretic Post Stroke Barthel No. Sex (y) Side (mo) Score Stroke Type/CT Scan Report 1
M
61
L
18
9
2
F
76
R
14
17
3
M
61
R
27
15
4
M
49
L
36
16
5
M
67
R
48
9
6
M
58
L
36
17
7
F
68
R
48
18
8
M
65
L
14
12
Initial UL Presentation
Partial anterior circulation stroke
Increased tone throughout UL. Loss of range of movement throughout all UL joints. Self-reported pain and stiffness of shoulder and pain in hand. NA Nearly full range of movement with no signs of increased tone but limb described as “heavy” and “dragging me down.” Total anterior circulation stroke. Significant increased tone in hand, Infarction in territory of middle overuse of shoulder in all attempts at cerebral artery movement. No attempt to use UL in functional tasks. Severe dysphasia. NA Significant increased tone in hand, overuse of shoulder, attempts made at use of limb in functional tasks, severe stiffness felt on passive movements of all finger joints and wrist joint. NA Increased tone in hand, painful shoulder, no attempts to use hand functionally, loss of range of movement in all joints of hand and wrist. Severe dysphasia. Hemorrhagic Overuse of shoulder, decreased strength throughout UL, very poor coordination, limb easily tires with activity, marked loss of sensation and altered sensation including proprioception, hot/cold, and light touch. Cerebral infarct Using arm well in functional activity but reports “heavy” feeling with altered sensation and pain. Very little light touch sensation or temperature sensation and poor proprioception. Mild dysphasia. Infarct over left cerebellar hemisphere Pain in shoulder, incipient return of active and right basal ganglia; lacunar movement in hand, wrist, and elbow, infarct over left external capsule increased tone in hand with loss of range of movement at wrist joint.
Abbreviations: CT, computed tomography; F, female; L, left; M, male; NA, not available; R, right.
www.archives-pmr.org
Mobilization and tactile stimulation in chronic stroke single-system studies, these tests can be used to identify the difference between mean values in different phases of the study, as long as the point at which intervention begins is randomly allocated.39 We performed these tests using appropriate syntax in SPSS version 14.0.a The syntax calculated a “count of arrangements”dthe number of possible permutations or arrangements that would produce a mean difference greater than or equal to the observed mean difference. On this basis, the statistical significance of the observed mean difference can be determined. Accordingly, we based supporting statistical data analysis on randomization tests between the baseline (A1) phase and the intervention (B) phase alone. Statistical significance was set at P.05.
Results All 8 participants completed all phases of the study series with no adverse effects. This was considered to be a positive result. Table 3 summarizes individual participant profiles on entry to the study. Absence of computed tomography scan results accounts for the missing data in table 3. Table 4 shows the mean amount of treatment received by each participant during the intervention (B) phase.
Motricity Index (arm section) Figure 1 shows the total MI arm section scores for all participants. Visual inspection of total MI scores across phases showed a positive change in trend and change in slope for all participants. All participants showed a period of latency before a change in total score, ranging from 0 days (participant 8) to 28 days (participant 1) after the introduction of the intervention. Improvements occurred in the pinch subsection for 7 participants, and all participants improved in either or both of the elbow and shoulder subsections, from stable baseline to the end of the intervention phase. Total change scores ranged from 9 to 37. Table 5 summarizes MI subsection and total change scores for each participant. In summary, total MI arm section scores improved for all participants in the intervention phase; no further improvement or deterioration was evident in the withdrawal phase.
Table 4 Mean treatment time per participant during the intervention phase
Participant No.
Total Intervention Time (h)
Range of Treatment Session Times (min)
Mean Treatment Session Time (min)
1 2 3 4 5 6 7 8
21.46 12.50 18.33 19.25 18.41 17.91 16.08 16.83
30e60 30e40 25e45 30e50 30e55 35e40 30e55 30e40
51.40 35.70 37.93 42.77 44.20 43.00 41.95 36.07
www.archives-pmr.org
697
Action Research Arm Test Figure 2 shows the total ARAT scores for each participant. Visual inspection of the plotted data showed a positive change in trend and a change in slope for 5 participants; participant 2 had attained the maximum score during the baseline phase and so could not improve further. All participants showed a period of latency before a change in total score was evident. These periods of latency ranged from 5 days (participant 7) to 31 days (participant 3) after introduction of the intervention. Table 6 summarizes the ARAT subsection and total change scores for each participant. Improvements were seen in at least 1 subsection for 7 participants. Total change scores ranged from 1 to 24. No improvement was seen for participant 2. Change scores reflected a greater magnitude of improvement in combined distal activity scores (pinch, grip, grasp) than in gross movement scores. However, a greater number of participants showed some improvement, albeit small, in gross movement scores. In conclusion, while total ARAT scores improved for 7 participants in the intervention phase, we only consider the changes to be clinically significant in 4 of the 8 participants (participants 1, 6, 7, and 8). No further improvement or deterioration was evident in the withdrawal phase in 7 patients, but 1 participant (6) showed deterioration in scores during this final phase. While we found a positive change in mean difference in scores between the baseline (A1) and intervention (B) phases for all but 1 participant (participant 2), randomization tests showed that the level of change was not statistically significant.
Variability There was little variability in the scores achieved by the participants. Negative dips in the MI (participants 3, 4, and 5) and ARAT (participants 4 and 5) scores, seen in figures 1 and 2, coincided with and could potentially be accounted for by changes to anticonvulsive medication (participant 3), poorly controlled blood glucose levels (participant 4), and the presence of flu symptoms (participant 5), which were reported in field notes.
Discussion We recruited participants to this study on the basis of their having persistent UL dysfunction at least 12 months after the onset of their stroke. ARAT scores for all participants remained stable throughout baseline, confirming that recovery in the UL at this stage post stroke appeared to have stabilized, with no further spontaneous recovery. Therefore, since further spontaneous improvement in performance was highly unlikely for this stage post stroke, we could be more confident that any changes in performance during the intervention phase were attributable to the intervention, particularly if this result was replicated across cases. Total MI (arm section) and ARAT scores achieved during the initial baseline (A1) phase did indeed remain consistent, reflecting stability of impairment and activity limitation. However, all 8 participants subsequently showed improvement in UL motor impairment (MI arm section) with the application of the intervention, and 4 participants demonstrated clinically significant improvement in ARAT scores, suggesting an effect of the MTS. Of particular interest is the finding that distal activity (MI pinch; ARAT pinch, grip, grasp) appeared to improve more than
J.M. Winter et al
Motricity Index arm section total scores
698
Day Fig 1
Total MI arm section scores for participants 1 through 8.
proximal/gross activity. This is perhaps unsurprising since MTS intervention is targeted at the hand; however, this provides evidence that distal recovery can occur independently of proximal
recovery. Nevertheless, it is also of interest that a distal-focused intervention also appeared to impact positively on proximal recovery in the chronic stage post stroke. www.archives-pmr.org
Mobilization and tactile stimulation in chronic stroke
699
Table 5 Summary of visual changes in total MI (arm) subsection scores and improvements in subsection scores for each participant, showing scores at end of baseline phase and maximum score attained during intervention phase Participant No.
Change in Trend
Change in Slope
Improvement in Pinch Subsection
Improvement in Elbow Subsection
Improvement in Shoulder Subsection
Total No. of Points Increased
1 2 3 4 5 6 7 8
Yes Yes Yes Yes Yes Yes Yes Yes
Yes Yes Yes Yes Yes Yes Yes Yes
19 7 4 4 15 0 7 4
0 8 0 5 5 0 19 5
11 0 6 0 0 11 11 5
30 15 10 9 20 11 37 14
(0e19) (26e33) (22e26) (22e26) (11e26) (26e26) (26e33) (22e26)
(14e14) (25e33) (14e14) (14e19) (9e14) (25e25) (14e33) (14e19)
(14e25) (25e25) (19e25) (19e19) (9e9) (14e25) (14e25) (14e19)
NOTE. The 2 figures in parentheses represent, respectively, the score at end of the baseline phase and the maximum score in the intervention phase; the improvement score is calculated by the subtraction of the first of these from the second.
When considering overall ARAT and MI (arm section) results in the context of participant profiles, the greatest change in ARAT score was achieved by participants presenting with self-reported sensory deficits (participants 6 and 7). In contrast, the least change in ARAT score was achieved by participants with marked distal increased tone or spasticity of the UL (participants 3, 4, and 5). However, this pattern is not replicated in the MI results, and this requires further exploration. Participants with higher initial scores did not appear to make greater improvements than those with lower entry-level scores on either the ARAT or the MI. No obvious patterns relating initial scores to final scores were seen. Therefore, we suggest that entrylevel score cannot be used as a prognostic indicator of outcome. The relationship between site of lesion or type of stroke and UL recovery in response to MTS cannot be commented on from the data generated from this single-system study. However, these variables warrant further investigation to identify the optimum target population for a larger trial of MTS after stroke. Immediate change in performance was only seen in 1 participant (8) (change of level of MI score), while change was delayed in all other participants. The length of latency ranged from 0 to 28 days for MI scores and 5 to 31 days for ARAT scores. All improvements followed a stepwise pattern, with periods of no improvement (plateau) followed by improvement. The duration of these periods varied between participants. In light of the recovery patterns seen, the inclusion of a second intervention (B) phase (A-B-A-B design) might have added further information on the emerging patterns of UL recovery. Overall, scores achieved throughout the B phase remained unchanged during the withdrawal phase, apart from participant 6, whose ARAT score decreased by 3 points at the end of this phase. This participant had reported high levels of sensory impairment, and further exploration of the impact of sensory impairment in the hand on recovery is warranted. Understanding patterns of recovery after stroke is fundamentally important to therapists to ensure that they design rehabilitation programs that maximize opportunities for recovery. All participants in this study improved with MTS, and a clear, common pattern of recovery emerged: a prolonged initial period of latency followed by stepwise periods of improvement and stabilization/plateauing. This pattern was the same for all participants, regardless of clinical presentation. This knowledge is important for clinicians because key decisions can be made regarding ongoing treatment plans, and goal setting that reflects www.archives-pmr.org
these periods of recovery and latency can be supported. Furthermore, it may inform the timing of discharge from ongoing therapy. The intensity of this therapy and the sensory bombardment of the central nervous system required participants to focus attention on the treated hand; this involved great cognitive effort, which may be a challenge to some patients receiving this intervention. However, for the participants in this study, who were able to maintain focus, positive outcomes were seen.
Study limitations The risk of subjectivity and poor reliability in visual interpretation of data has been acknowledged41; hence the recommendation that, wherever possible, statistical tests are used to support visual analysis. Visual inspection of the data has also been reported to be limited in that only effects that can be clearly and easily seen as significant are considereddsuch that weak but positive treatment effects may be overlooked17dand serial dependency of the data may affect their interpretation.42 These are further reasons not to rely on visual inspection alone. In contrast, it can be argued that changes that are sufficiently marked to be discernible through visual analysis are likely to be clinically significant,19 and statistical tests may find a statistically significant change that is of no clinical significance. Therefore, clinical significance of changes in performance must be considered alongside any statistical significance of any change in performance score. Randomization tests were used in this study to support visual analysis. However, there are difficulties in using statistical tests that rely on differences in mean scores when there are long periods of latency of recovery and when recovery occurs late in the intervention phase. If large changes in score constitute the last few data points of the intervention phase after a long period of latency, they will have little influence on the mean score in this phase, and the between-phase statistical test may therefore be nonsignificant. However, while this may seem to be weak evidence of a treatment effect, clinical experience and participant reports suggest that the changes were clinically significant, and the use of statistical analysis may not be essential to infer an effect.39 We chose not to include a measure of UL sensation as an outcome measure because of reported unreliability of the various tests.14 This, however, was a limitation of the study, particularly since 2 participants (6 and 7) with marked self-reported sensory deficits appeared to make early and great improvements. Future studies of MTS should include a measure of sensory impairment.
J.M. Winter et al
ARAT total scores
700
Day Fig 2
Total ARAT scores for participants 1 through 8.
The potential for investigator and measurement bias is considered as a potential limitation to single-system studies. In a single-system study, it is not appropriate to blind the assessor to
allocation to a treatment group because treatment groups are not part of the design. However, if the assessor had been blind to the stage of the design (A or B phases) and the point of transition www.archives-pmr.org
Mobilization and tactile stimulation in chronic stroke
701
Table 6 Summary of visual changes in total ARAT scores and improvements in scores for ARAT subsections for each participant, showing score at end of baseline phase and maximum score attained in intervention phase Participant No.
Change in Trend e Total Score
Change in Slope e Total Score
Improvement in Pinch Subsection
Improvement in Grip Subsection
Improvement in Grasp Subsection
Improvement in Gross Movement Subsection
Total No. of Points Increased
1 2 3 4 5 6 7 8
Yes No No No Yes Yes Yes Yes
Yes No No No Yes Yes Yes Yes
3 0 0 0 0 4 11 4
2 0 0 0 1 4 4 3
2 0 0 0 0 5 6 4
1 0 1 2 2 2 3 1
8 0 1 2 3 15 24 12
(6e9) (18e18) (6e6) (6e6) (0e0) (13e17) (6e17) (2e6)
(4e6) (12e12) (4e4) (4e4) (0e1) (8e12) (8e12) (4e7)
(6e8) (18e18) (6e6) (6e6) (0e0) (13e18) (12e18) (7e11)
(3e4) (9e9) (3e4) (4e6) (1e3) (7e9) (6e9) (3e4)
NOTE. The 2 figures in parentheses represent, respectively, the score at end of the baseline phase and the maximum score in the intervention phase; the improvement score is calculated by the subtraction of the first of these from the second.
between phases, this would have increased the internal validity of the results. However, blinding was not feasible within the resources for this study. Most single-system studies include a phase in which the intervention is withdrawn,17 and this is sometimes considered to be reflective of the withdrawal of treatment that occurs as a matter of course in clinical practice, which is frequently after a course of 6 weeks. The withdrawal phase serves a purpose in identifying whether changes in performance are maintained when the intervention ceases, a potential measure of the carryover effects of therapy. Withholding or withdrawing treatment in single-system studies nonetheless has potential ethical implications.43 In this study, however, all participants had been discharged from therapy, so there was no issue of denying routine intervention. On this basis, and in the light of the consent provided by the participants, we considered that withdrawal of treatment was justifiable.
Conclusions This exploratory study has generated hypotheses for further study and identified several important clinical messages:
these improvements can be carried over and maintained for at least 2 weeks after withdrawal of the intervention. We now need to explore the long-term effects of MTS on UL recovery. The application of MTS, a hands-on intervention used in routine clinical practice to prime and/or augment the motor execution system in order to enhance or facilitate voluntary activation of muscle, appears to provide beneficial effects with no negative side effects. Not only should it continue to be used in routine therapy practice after stroke for patients with varying clinical presentations, its effectiveness should be explored in ongoing studies. The results of this study contribute, along with other preliminary studies of MTS, to the identification of an appropriate target population and dose of intervention for future trials of MTS. The ability of ongoing improvements in UL performance to be made and sustained more than a year post stroke suggests that therapists should give consideration to where and when their interventions are best provided. They may need to reflect on whether services are really best placed in the acute hospital environment or whether future service redesigns may be better focused on rehabilitation in the community.
A 6-week daily program of therapist-led mobilization of joints and soft tissues coupled with tactile stimulation of the hand and forearm (MTS)da module of conventional therapydappears to improve motor impairment and motor activity. Clinically significant recovery of persistent UL dysfunction is possible even 12 months or more post stroke. Distal UL recovery can occur with targeted treatment independently of proximal recovery after stroke. MTS as a distally focused intervention may still impact positively on proximal recovery. The pattern of recovery at this stage is likely to be characterized by periods of improvement followed by plateaus of performance. A period of latency after the commencement of MTS is to be expected. The effectiveness of sensory stimulation in treatment, while as yet unclear, should not be dismissed.
Supplier
These replicated results suggest that improvements in UL performance may occur more than a year post stroke, and that
We thank West Midlands Stroke Research Network for assistance in recruitment.
www.archives-pmr.org
a. SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.
Keywords Musculoskeletal manipulations; Rehabilitation; Sensation; Stroke; Therapy; Upper extremity
Corresponding author Susan M. Hunter, PhD, School of Health and Rehabilitation, Keele University, Keele, Staffordshire, ST5 5BG, UK. E-mail address: [email protected].
Acknowledgment
702
References 1. Mackay J, Mensah G. Atlas of heart disease and stroke. Part 3: the burdendglobal burden of stroke. World Health Organization; 2004. Available at: http://www.who.int/cardiovascular_diseases/en/cvd_atlas_ 15_burden_stroke.pdf. Accessed February 21, 2012. 2. Jorgensen HS, Nakayama H, Pedersen PM, Kamersgaard L, Raaschou HO, Olsen TS. Epidemiology of stroke-related disability. Clin Geriatr Med 1999;15:785-99. 3. Carr J, Shepherd R. Neurological rehabilitation: optimizing motor performance. London: Butterworth-Heinmann; 1998. 4. Wade D, Skilbeck CE, Hewer R. Predicting Barthel ADL score at 6 months after an acute stroke. Arch Phys Med Rehabil 1983;64:24-8. 5. Wade DT, Langton-Hewer R, Wood VA, Skilbeck CE, Ismail HM. The hemiplegic arm after stroke: measurement and recovery. J Neurol Neurosurg Psychiatry 1983;46:521-4. 6. Parker VM, Wade DT, Langton-Hewer R. Loss of arm function after stroke: measurement, frequency and recovery. Int Rehabil Med 1986; 8:69-73. 7. Broeks JG, Lankhorst GJ, Rumping K, Prevo AJ. The long-term outcome of arm function after stroke: results of a follow up study. Disabil Rehabil 1999;21:357-64. 8. Nudo RJ, Friel KM. Cortical plasticity after stroke: implications for rehabilitation. Rev Neurol (Paris) 1999;155:713-7. 9. Sirtori V, Corbetta D, Moja L, Gatti R. Constraint-induced movement therapy for upper extremities in stroke patients. Cochrane Database Syst Rev 2009;(4):CD004433. 10. Winter J, Hunter S, Sim J, Crome P. Hands-on therapy interventions for upper limb motor dysfunction following stroke. Cochrane Database Syst Rev 2011;(6):CD006609. 11. Hunter SM, Crome P, Sim J, Donaldson C, Pomeroy VM. Development of treatment schedules for research: a structured review to identify methodologies used and a worked example of ‘mobilisation and tactile stimulation’ for stroke patients. Physiotherapy 2006;92: 195-207. 12. Medical Research Council (MRC). A framework for development and evaluation of RCTs for complex interventions to improve health. London: MRC Health Services and Public Health Research Board; 2000. 13. Hunter SM, Hammett L, Ball S, et al. Dose-response study of mobilisation and tactile stimulation for the upper extremity early after stroke: a phase I trial. Neurorehabil Neural Repair 2011;25: 314-22. 14. Hunter SM, Crome P, Sim J, Pomeroy VM. Effects of Mobilization and Tactile Stimulation on recovery of the hemiplegic upper limb: a series of replicated single-system studies. Arch Phys Med Rehabil 2008;89:2003-10. 15. Wing AM, Haggard P, Randall Flanagan J. Hand and brain. The neurophysiology and psychology of hand movements. San Diego: Academic Pr; 1996. 16. Medical Research Council (MRC). Developing and evaluating complex interventions: new guidance. London: MRC; 2000. Available at: http:// www.mrc.ac.uk/Utilities/Documentrecord/index.htm?dZMRC004871. Accessed September 10, 2012. 17. Kazdin AE. Single-case research designs: methods for clinical and applied settings. New York: Oxford University Pr; 1982. 18. Barlow D, Hersen M. Single case experimental designs: strategies for studying behavior change. 2nd ed. Boston: Allyn and Bacon; 1984. 19. Ottenbacher KJ. Evaluating clinical change: strategies for occupational and physical therapists. Baltimore: Williams & Wilkins; 1986. 20. Robson C. Real world research: a resource for social scientists and practitioner-researchers. 2nd ed. Oxford: Blackwell; 2002.
J.M. Winter et al 21. Mahoney FI, Barthel DW. Functional evaluation: the Barthel Index. Md State Med J 1965;14:61-5. 22. Wilson B, Cockburn J, Halligan PW. Behavioural Inattention Test. Bury St Edmunds: Thames Valley Test Company; 1987. 23. Friedman PJ. The Star Cancellation Test in acute stroke. Clin Rehabil 1992;6:23-30. 24. Collen F, Wade D, Bradshaw C. Mobility after stroke: reliability of measures of impairment and disability. Int Disabil Stud 1990;12:6-9. 25. Collin C, Wade D. Assessing motor impairment after stroke: a pilot reliability study. J Neurol Neurosurg Psychiatry 1990;53:576-9. 26. Bohannon R. Motricity Index scores are valid indicators of paretic upper extremity strength following stroke. J Phys Ther Sci 1999;11: 59-61. 27. Wade D. Measurement in neurological rehabilitation. Oxford: Oxford University Pr; 1992. 28. Lyle RC. A performance test for assessment of upper limb function in physical rehabilitation treatment and research. Int J Rehabil Res 1981; 4:483-92. 29. Wade D. Measuring arm impairment and disability after stroke. Int Disabil Stud 1989;11:89-92. 30. Hsieh C-L, Hsueh I-P, Chiang F-M, Lin P-H. Inter-rater reliability and validity of the Action Research Arm Test in stroke patients. Age Ageing 1998;27:107-13. 31. Hsueh I-P, Lee M-M, Hsieh C-L. The Action Research Arm Test: is it necessary for patients being tested to sit at a standardized table? Clin Rehabil 2002;16:382-8. 32. Donaldson C, Tallis R, Pomeroy V. Outcome measures in neurophysiotherapy for the arm and hand: have we lost our grip? Clin Rehabil 2006;20:459-60. 33. van der Lee JH, Wagenaar RC, Lankhorst GJ, Vogelaar TW, Deville´ WL, Bouter LM. Forced use of the upper extremity in chronic stroke patients: results from a single-blind randomized clinical trial. Stroke 1999;30:2369-75. 34. Donaldson C, Tallis RC, Miller S, Sunderland A, Lemon R, Pomeroy VM. Effects of conventional physical therapy and functional strength training on upper limb motor recovery after stroke: a randomised phase II study. Neurorehabil Neural Repair 2009;23:389-97. 35. Medical Research Council (MRC). MRC guidelines for good clinical practice in clinical trials. London: MRC Health Services and Public Health Research Board; 1998. 36. Sim J, Wright C. Research in health care: concepts, designs and methods. Cheltenham: Nelson Thornes; 2000. 37. Tryon W. A simplified time series analysis for evaluating treatment interventions. J Appl Behav Anal 1982;15:423-9. 38. Hartmann DP, Gottman JM, Jones RR, Gardner W, Kazdin AE, Vaught R. Interrupted time-series analysis and its application to behavioural data. J Appl Behav Anal 1980;13:543-59. 39. Todman J, Dugard P. Single-case and small-n experimental designs: a practical guide to randomization tests. Mahwah: Lawrence Erlbaum; 2001. 40. Gorman BS, Allison DB. Statistical alternatives for single-case designs. In: Franklin RD, Allison DB, Gorman BS, editors. Design and analysis of single-case research. Mahwah: Lawrence Erlbaum; 1997. p 159-214. 41. Richards SB, Taylor RL, Ramasamy R, Richards RY. Single subject research: applications in educational and clinical settings. San Diego: Singular Publishing; 1999. 42. Matyas TA, Greenwood KM. Serial dependency in single-case time series. In: Franklin RD, Allison DB, Gorman BS, editors. Design and analysis of single-case research. Mahwah: Lawrence Erlbaum; 1997. p 215-43. 43. Sim J. The ethics of single system (nZ1) research. Physiother Theory Pract 1994;10:111-22.
www.archives-pmr.org