Neurorehabilitation and

Neural Repair
Volume 23 Number 2
February 2009 184-190
© 2009 The American Society of
Motor Training of Upper Extremity With Functional Neurorehabilitation
10.1177/1545968308324548

Electrical Stimulation in Early Stroke Rehabilitation http://nnr.sagepub.com

hosted at
http://online.sagepub.com

Sabine Mangold, PhD, Corina Schuster, MPTSc, Thierry Keller, PhD,

Andrea Zimmermann-Schlatter, MPTSc, and Thierry Ettlin, MD

Background. Functional electrical stimulation (FES) allows active exercises in stroke patients with upper extremity paralysis. Objective.
To investigate the effect of motor training with FES on motor recovery in acute and subacute stroke patients with severe to complete arm
and/or hand paralysis. Methods. For this pilot study, 23 acute and subacute stroke patients were randomly assigned to the intervention
(n = 12) and control group (n = 11). Distributed over 4 weeks, FES training replaced 12 conventional training sessions in the intervention
group. An Extended Barthel Index (EBI) subscore assessed the performance of activities of daily living (ADL). The Chedoke McMaster
Stroke Assessment (CMSA) measured hand and arm function and shoulder pain. The Modified Ashworth Scale (MAS) assessed resis-
tance to passive movement. Unblinded assessments were performed prior to and following the end of the training period. Results. The
EBI subscore and CMSA arm score improved significantly in both groups. The CMSA hand function improved significantly in the FES
group. Resistance to passive movement of finger and wrist flexors increased significantly in the FES group. Shoulder pain did not change
significantly. None of the outcome measures, however, demonstrated significant gain differences between the groups. Conclusions. We
did not find clear evidence for superiority or inferiority of FES. Our findings, and those of similar trials, suggest that the number of ses-
sions should be at least doubled to test for superiority of FES in these highly impaired patients and approximately 50 participants would
have to be assigned to each therapeutic intervention to find significant differences.

Keywords:   Electric stimulation; Stroke; Hemiplegia; Upper extremity; Rehabilitation

I mpaired motor control of the upper extremity is one of the

most frequent consequences of a stroke. As few as 13% of
the participants examined within the first 2 weeks after onset
have no upper extremity motor impairment,1 and 6 months
after the injury, severe motor deficits remain in 30% to 66%
of stroke survivors.2
To improve arm function, standard therapeutic interventions
are used, such as Bobath therapy,3 constraint-induced therapy,4
and a task-specific motor relearning program.5 Therapeutic
strategies that impose a higher level of practice activity in the
paretic muscles should be preferred.6 This is an ambitious
demand for patients who have very little voluntary function
of the paretic limb. One method that does not require voluntary
muscle function during motor tasks is functional electrical
stimulation (FES). A variety of therapeutic objectives have
been partially achieved through FES, such as facilitation of
voluntary movements, improvement of manual dexterity and
activities of daily living (ADL),7-11 and reduction of shoulder
pain12 and spasticity.13
However, in some studies, FES was applied in addition to
conventional therapy or was compared to a placebo therapy,
reducing the certainty as to whether FES effects can be attributed
to either the additional therapeutic time for motor training or the
application of the particular method. Thus, more studies are
necessary that provide the same amount of motor training to the
intervention and control group. Moreover, in standard rehabilitation
settings, additional time devoted to a therapeutic strategy may not
be financially feasible so integration of FES into the usual clinical
treatment schedule is more feasible than an additional application.
In this trial, we applied FES under the realistic rehabilitation
conditions of our country by not adding time to the typical course
of therapy. We examined stroke patients with severe to complete
paralysis of the arm and/or hand in the acute or subacute phase.
The effects of FES on this group of patients have not been
sufficiently examined. Accordingly, our purpose was to compare
motor recovery in severely affected stroke patients in early
rehabilitation who were treated with FES or conventional motor
training under normal clinical conditions.

From the Balgrist University Hospital, Spinal Cord Center, Zurich, Switzerland (SM, TK); Reha Rheinfelden, Rheinfelden, Switzerland (CS, AZ-S, TE); and
the Automatic Control Laboratory, Swiss Federal Institute of Technology, Zurich, Switzerland (TK). Address correspondence to Sabine Mangold, PhD, Balgrist
University Hospital, Spinal Cord Center/Research, Forchstrasse 340, 8008 Zurich, Switzerland. E-mail: smangold@balgrist.unizh.ch.

184

Methods
Participants
Participants were recruited from an inpatient rehabilitation
stroke center. The local ethics committee approved the study
protocol and informed written consent to participate in the
study was obtained from all participants. In cases of writing
disability, one witness confirmed the participant’s informed
consent.
First-time stroke survivors admitted to the rehabilitation
center were screened for eligibility. They were included in the
study if they met the following criteria: were 18 years of age
or older; had their first stroke from 2 weeks to 18 weeks from
entry into the trial; had an infarct or hemorrhage localized
from the cortex to the brainstem; had severe hemiparesis to
complete hemiplegia of the arm and/or hand (maximum values
of the Chedoke McMaster Stroke Assessment [CMSA] for the
arm and hand were 3 points); were with or without aphasia, but
had the ability to understand the study and to provide informed
consent; and had the ability to sit in a wheelchair or on a chair
with no contact to the back rest.
Participants were excluded in cases with the following: a
pace-maker or other stimulation devices; pregnancy; epilepsy;
prosthesis of bones or joints in the local region of the electrical
treatment; skin injuries, rashes, burns, or fresh scars at the sites
of stimulation; prior stroke or other brain impairments; severe
impairment of sensitivity and proprioception (according to
clinical estimation); shoulder-hand syndrome on the affected
side; constant, strong pain in the shoulder of the affected side
(CMSA value for painful shoulder, 1-3); or considerable
subluxation (>20 mm) of the affected shoulder joint.
Therapies and FES Intervention
Participants included in the 4-week training program were
randomly assigned to the intervention (FES and conventional
training) or to the control group (conventional training) using
a computer-generated randomization list.
In both groups, 3 to 5 occupational therapy sessions for the
upper extremity training were scheduled per week, each lasting
45 minutes. The number of therapies was adapted to the patient’s
resilience via a number of accompanying therapies, including
physiotherapy, speech therapy, and neuropsychological training.
In the intervention group, FES made up 3 of the occupational
therapy training sessions per week. If the patient had more
functional training sessions per week, they were carried out in
conventional therapeutic form.
Conventional occupational therapy for upper extremity
training mainly consisted of mobilization and selective move­
ments of the shoulder, hand, and arm, and grasping exercises
supported by the therapist if necessary. To a small degree,
ADL (eg, mixing cream, pouring water into a glass) and
sensory retraining were practiced, as supported by the therapist
or performed bimanually.
FES was carried out with the portable system called Compex
Motion14 using standard self-adhesive surface electrodes. The
Mangold et al / FES in Early Stroke Rehabilitation   185  
programmable FES system, including 4 current-regulated
stimulation channels, was used to generate custom-made
movement sequences. Proximal muscles (eg, anterior deltoid
muscle, m. triceps brachii) as well as distal muscles (finger
extensors and finger flexors) were integrated into the stimulation
sequence to reach, grasp, and release an object. Therapists
selected the muscles to be stimulated with respect to muscles
that met the following criteria: (a) had priority for the individual’s
current rehabilitation process, and (b) helped in the grasp
process (reach, grasp, and release an object). If necessary,
therapists manually treated spasticity, used a help arm to reduce
gravity, and provided manual assistance to support the grasp
action and to prevent movements that might cause damage such
as pathologic scapula movements. Before the FES treatment
period, stimulation thresholds for each muscle and the set-up
were determined in a 1-hour single FES session. If the patients’
arm functions changed during the 4-week FES training,
stimulation programs and supporting measures were adapted.
Each FES training session consisted of a program including
repetitive grasping functions. After the system was donned,
stimulation current amplitudes were individually set for each
muscle group to achieve sufficient functional contraction force
without pain or discomfort. The stimulation frequency was 25 Hz.
The pulse width varied between 0 and 250 microseconds to achieve
controlled muscle contraction. Every reach-grasp-release cycle
consisted of different steps triggered by a push button, eg, stretching
the arm with the hand opening, hand closing, arm moving back, etc.
Figure 1 shows a typical cycle. The participants were encouraged
to make volitional attempts to reach, open the hand, and close the
hand simultaneously with the stimulation. They could either trigger
the stimulation sequence themselves or the therapist performed this
task, indicating verbally to the participant when the next step was
initiated. Grasping objects were used such as a small bottle, ball,
tube, tin, glass, or handkerchief. The FES session lasted 45 minutes,
consisting of about 15 to 20 minutes of donning, other preparations
(eg, treating spasticity), and doffing, leaving 25 to 30 minutes for
functional training.
Assessments
For baseline comparability, the following data were
assessed for each patient: age, gender, time poststroke, affected
side, diagnosis (hemorrhagic or nonhemorrhagic stroke), and
hemineglect. Additionally, the Extended Barthel Index (EBI)
was selected to determine the level of independence with respect
to ADL and the CMSA to assess motor recovery of the paretic
upper limb. The EBI contains 16 items (each rated with 0-4
points) including ADL, mobility, bowel and bladder functions,
communication, and cognitive and social functions.
Baseline outcome measures were assessed 8 days and 1 day
before the beginning of the 4-week program. Posttreatment
assessments were performed at the end of the 4-week program.
Follow-up measurements were planned at 6 months after the
beginning of the treatment.
An EBI subscore, including 4 items focusing on upper
extremity function (Table 1), and the arm and hand function
186   Neurorehabilitation and Neural Repair

Figure 1 chosen. Shoulder pain ranges from constantly strong pain in

Typical Grasping Cycle the shoulder and other parts (1 point) to no pain in the shoulder
and no prognostic indicators (7 points). The MAS measures
nonvoluntary resistance against passive movement. It ranges
from no increase in muscle tone (0 points) to rigidity of the
affected parts in flexion or extension (5 points). The following
muscles were assessed with the MAS: finger flexors, finger
extensors, wrist flexors, wrist extensors, pronators, supinators,
elbow flexors, and elbow extensor.
The nursing staff assessed the EBI and the two trained
physiotherapists performed the CMSA and MAS. Participants
and assessors were not blinded. The reliability and validity of
the CMSA,15 EBI,16 and MAS17,18 have been confirmed.

Note: Six push-button signals trigger this typical grasping cycle to allow
an individual timing. The dashed plateau lines within one step indicate that
the stimulation remains constant until the next step is triggered. The sig-
nals are the following: (1) anterior deltoid muscle, m. triceps brachii, and
finger extensors are activated (the arm stretches forwards with hand open
to reach an object); (2) arm remains stretched and hand closes to grasp an
object; (3) hand remains closed and arm muscles relax; (4) arm stretches
to the front while fingers remain closed; (5) arm remains stretched
and hand is opened to release the object; (6) the stimulation of arm mus-
cles stops, bringing the arm passively back to the resting position. Once
completed, the cycle can be started anew.
CMSA score of the paretic arm were selected as primary
outcome measures. Each CMSA item score ranges from no
or almost no voluntary muscle activity (1 point) to normal
coordination of movement (7 points).
As secondary outcome measures, the item “shoulder pain”
of the CMSA and the Modified Ashworth Scale (MAS) were
Data Analysis
Data analysis was performed by intention-to-treat. An
analysis of kurtosis and skewness was done for age, time
poststroke, and number of occupational therapy sessions of
upper extremity training to select adequate statistical tests for
the evaluation of baseline comparability.
Baseline comparability was evaluated with the following
statistical tests. Age was tested with the independent t test
(Levene’s test for equality of variances). Time poststroke and
number of occupational therapy sessions were evaluated with
the 2-tailed Mann-Whitney U test for nonparametric data,
because they were not normally distributed. This test was also
used for EBI and CMSA scores (mean baseline values),
because they are ordinal parameters. The number of participants
in each group was compared with the χ2 test. The Pearson
2-sided χ2 test was used to evaluate proportional differences
in both groups regarding gender, diagnosis (hemorrhagic or

Table 1
Extended Barthel Index (Subscore)
Item Item Description Score
Eating and drinking Is unable, or requires a stomach feed tube and is dependent on assistance with device 0
Needs help in cutting and spreading 2
Is independent with adaptive equipment (eg, adapted cutlery handles) 3
Is independent, or requires a stomach feed tube, but can handle device independently 4
Grooming (washing face, Is unable 0
   combing, shaving, Needs help with many, but not all procedures 1
   brushing teeth) Needs minor assistance (eg, opening the toothpaste tube), or is physically independent, but needs supervision 2
Is independent with adaptive equipment (eg, extension of the hairbrush handle) 3
Is independent 4
Dressing and undressing Is unable 0
   (including tying shoes, Needs help with most, but not all articles of clothing, or needs help with all articles of clothing, but assists actively 1
   buttoning up) Needs help with only a few procedures (eg, support for tying shoes, donning orthotic devices), or is physically 2
   independent, but needs supervision
Is independent (perhaps with assistive devices such as sock aids) 4
Bathing self Is unable 0
Needs assistance with many, but not all procedures, (eg, support for transfers, is able to wash the upper part of the 1
   body, but not the lower part)
Needs minor assistance (eg, opening dispensers), or is physically independent, but needs supervision 2
Needs adaptive equipment (eg, bath chair, bath lift), but can use it independently 3
Is independent and needs no adaptive equipment 4

Table 2
Patients’ Clinical Characteristics at Baseline
and Statistical Group Differences
Group
Control FES Differences
Group Group (P Value)
n 11 12 .84
Age (years), mean (SD) 62 (16.2) 57.5 (16.7) .74
Female participant, n (%) 4 (36) 2 (17) .28
Time poststroke (weeks), 7.3 (5.8, 8.2) 6.7 (6.4, 7.3) .56
   median (quartiles)
Nonhemorrhagic stroke, n (%) 9 (82) 10 (83) .92
Right hemiparesis, n (%) 4 (36) 3 (25) .55
Dominant arm affected, n (%)a 7 (64) 4 (40) .28
Neglect, n (%) 2 (18) 6 (50) .11
EBI, median (quartiles) 25 (21, 32) 32 (26, 35) .14
Abbreviations: FES, functional electrical stimulation; EBI, Extended Barthel
Index.
a
There were 2 missing values in the FES group.
nonhemorrhagic stroke), affected side, dominant arm affected,
and hemineglect. The Wilcoxon Signed Rank test was applied
to examine pre- and posttreatment changes in each group,
comparing mean baseline values with the values at the end of
the 4-week program. The differences of functional gains
between both groups were analyzed with the Mann-Whitney U
test. All statistical analyses were performed with SPSS 11.5
for Windows (SPSS, Chicago, Illinois). Results were accepted
as statistically significant at P ≤ .05.
The minimal clinically important difference (MCID) was
defined as a change of at least 9% of the maximum score
of the scale.19 With this 9% threshold, the MCID for the
EBI subscore was 1.4 points, for each CMSA item the
score was 0.6, and for each MAS item the score was 0.45
points. Additionally, effect sizes (Cohen’s d) of the primary
outcome measures were determined for each group (within-
group changes) and for gain differences between both
groups. Cohen’s d was calculated using the means (m) and
standard deviations (s) of the pretreatment and posttreatment
values, which were respectively the gains of both groups.
The calculation was performed with the Microsoft Office
Excel 2003 software using the following formula: Cohen’s
d = m1 - m2/spooled, where spooled s the square root of [(s1² + s2²)/2].
Based on the measured effect sizes, an alpha level of 0.5, and
a test power of 0.8, the optimal group size was derived from a
table.20
Results
Baseline Data
A total of 23 participants were enrolled, 11 in the control
and 12 in the FES group. All participants completed the study.
Participants’ baseline characteristics are shown in Tables 2 and
Mangold et al / FES in Early Stroke Rehabilitation   187  
3. Baseline comparability was confirmed for all demographic
data and primary outcome measures except one: The FES
group had a significantly higher EBI subscore than the control
group (P = .013).
Therapy Sessions
Nine patients participated in 11 to 12 FES sessions. Three
participants stopped FES after 3 to 7 FES sessions. One
participant could not tolerate the stimulation intensity that was
necessary for a functional muscle contraction; another stopped
because he developed high finger flexors tonus during the
session, although not stimulated. In their remaining therapy
sessions, these participants received conventional therapy.
In total, the participants in the control group had 8 to 15
therapy sessions (median 13) and in the FES group 10 to 20
(median 13), including FES and conventional therapy sessions.
The number of therapy sessions was not significantly different
in either group (P = .38).
Movement Quality
FES generated distinct finger and arm movements, but the
therapist had to provide manual support when participants
reached to an object and opened the hand. Hand closure and
retracting the arm were not manually supported. All participants
used the help arm to help reduce gravity.
Primary Outcome Measures
The FES group showed significant improvements in all
primary outcome measures (EBI subscore, CMSA arm and
hand) from pretreatment to posttreatment (Table 3, Figure 2).
This was also true for the control group with the exception that
CMSA hand function only barely reached the level of significance
(P = .06). Gains from pretreatment to posttreatment were not
significantly different between either group (Table 3, Figure 2).
Clinical importance, according to the MCID, was found for
the following primary outcome measures (Table 3). In the
control group, the gains of the median EBI subscore (3 points)
and CMSA arm score (1 point) were clinically important. In
the FES group the EBI subscore reached this change criterion
(1.5 points). Group differences of median EBI subscore gains
(1.5 points higher in the control group) and of the CMSA arm
score gains (1 point higher in the control group) were clinically
important.
Cohen proposed labeling effect sizes of 0.2 as small, 0.5 as
medium, and 0.8 or higher as large.21 According to this
classification, pretreatment and posttreatment effect sizes of
the primary outcome measures were small to high (Table 4)
and effect sizes of gain differences between groups were small
to medium (0.27 to 0.62). Hence, for medium effect sizes, a =
0.05, and test power = 0.8, the optimum sample size would be
50 participants per group and for small effect sizes the sample
size would be 310 participants per group.20
188   Neurorehabilitation and Neural Repair

Table 3
Baseline Values, Gains From Pretreatment to Posttreatment, and Statistical Group Differences of Primary
Outcome Measures and Modified Ashworth Scale of 2 Muscles that Tend to Become Spastic
Control Group (n = 11) FES Group (n = 12)
Group Differences (P Value)
Outcome Baseline Median Gain Median Baseline Median Gain Median
Measures (Quartiles) (Quartiles) (Quartiles) (Quartiles) Baseline Gain

EBI subscore   5 (4, 5.5) 3** (1, 3.3) 6 (6, 7) 1.5** (0.8, 2)    0.013*   0.19
CMSA arm 1 (1, 2) 1* (0, 1) 2 (1, 2)   0* (0, 1) 0.73   0.57
CMSA hand 2 (2, 2) 0 (0, 1) 2.3 (1, 2.6)   0.3* (0, 0.5) 0.50 1.0
MAS finger flexors 1.5 (0.8, 2)      0 (-0.5, 0.8) 1.3 (0.9, 2)   0.5* (0, 1.1) 0.78   0.17
MAS wrist flexors   3 (1.5, 3) 0.5 (0, 1.3)   2 (1, 2.5)   0.5* (0, 1.1) 0.13   0.68

Abbreviations: FES, functional electrical stimulation; EBI, Extended Barthel Index; CMSA, Chedoke McMaster Stroke Assessment; MAS, Modified Ashworth
Scale.
*P < .05; **P < .01.

Figure 2 Secondary Outcome Measures

Box Plot of the Pretreatment and Posttreatment
Values of the EBI Subscore The baseline MAS score was 0 in 22 participants for finger
extensors, wrist extensors, and supinators. Finger flexors,
wrist flexors, and pronators had median baseline values
between 1.25 and 3, and elbow extensors had a median value
of 0 in the FES group and 0.5 in the control group. Nine
participants in each group had a baseline CMSA shoulder pain
score of 6 (without pain but having at least one prognostic
indicator such as very low arm function or abnormal positioning
of the scapula).
With the small sample size, secondary outcome measures
did not show significant changes within each group from
pretreatment to posttreatment, with the exception of a
significant and clinically important increase of the MAS of
finger and wrist flexors in the FES group (Table 3). However,
no secondary outcome measures demonstrated significant or
clinically important gain differences between either group.

Abbreviations: EBI, Extended Barthel Index; FES, functional electrical
stimulation.
Table 4
Effect Sizes (Cohen’s d) of Gains of the Primary Outcome
Measures From Pretreatment to Posttreatment
Effect Size (Cohen’s d)
Outcome Control FES
Measures Group Group
EBI subscore 1.13 0.53
CMSA arm 0.78 0.78
CMSA hand 0.63 0.29
Abbreviations: FES, functional electrical stimulation; EBI, Extended Barthel
Index; CMSA, Chedoke McMaster Stroke Assessment.
Discussion
We investigated changes of upper extremity function in
stroke participants with severe paralysis. Control participants
received conventional upper extremity motor training.
Distributed over 4 weeks, FES training replaced 12 conventional
therapy sessions in the intervention group. The total number of
therapy sessions did not differ significantly.
Primary outcome measures, ie, the EBI subscore and the
CMSA (arm and hand), improved significantly in both groups
except for the CMSA hand score in the control group (P =
.06). According to the MCID, only the gains of the EBI
subscore in the control group (3 points) and FES group (1.5
points), and the gain of the CMSA arm score in the control
group (1 point) were clinically important. Having 0 to 2 points
per EBI item at baseline, a change of 1.5 points on the EBI
subscore indicates participant fluctuation in 1 or 2 items from
either being unable to being able to perform some tasks, or
needing significant assistance to minor assistance, or requiring

minor assistance to achieving independence with adaptive
equipment (Table 1). In their baseline CMSA arm score, all
participants but one had 1 or 2 points. For participants starting
at 1 point, an increase of 1 point indicated the achievement of
the ability to resist passive shoulder abduction and to extend
the elbow with facilitation. For those who had these functions
at baseline, a change of 1 point indicated the ability to touch
the chin and the contralateral knee.
The findings are in accordance with other studies on the
affected upper extremity, showing significant improvements
with only conventional therapy or additional electrical stimulation
within several weeks after onset.7,22 The effect sizes are the same
magnitude found in comparable studies, eg, 0.65 for the Barthel
Index23 or 0.73 for the Fugl-Meyer gain scores.24 We found no
significant differences between functional improvements of the
intervention and the control group. However, there were
clinically important group differences of median EBI subscore
and CMSA arm score, both favoring the control group. In other
studies of comparable participants, group differences for similar
functional tests were inconsistent. The lack of significant group
differences in our study might not only be because of a small
group size, but also because of the low number of FES sessions.
In other studies on acute stroke patients, various frequencies of
stimulation sessions were applied, from 24 to 120 sessions.8,10,24,25
Furthermore, 25% of our participants had only 7 or less
stimulation sessions, reducing the potential effect of FES. The
donning and doffing of the system reduced the effective time for
exercises in the FES group, suggesting that control participants
had clinically important, though not significantly higher gains of
the EBI subscore and CMSA arm score. Baseline differences
might also be responsible for these results. The proportion of
participants with hemineglect was higher in the FES group,
reducing the motor recovery potential.26 The dominant hand
was more frequently affected in the control group. Because of
higher motor awareness of the dominant hand,27 the dominant
side possibly had the potential for greater improvement.
Furthermore, the control group had a significantly lower EBI
subscore than the FES group at baseline. As the EBI is a
bimanual assessment, the control group possibly had lower
values because dominant hands were more frequently affected.
This could be a reason for a better EBI subscore gain in the
control group because the nondominant, nonaffected hand
started from lower dexterity than the dominant, nonaffected
hand, resulting in a better improvement potential.
Considering that either the stimulation of antagonist muscles
might decrease spastic muscle tonus via reciprocal inhibition,
or that the stimulation fatigues spastic muscles, it was
theorized that electrical stimulation could help reduce
spasticity. However, resistance to passive movement showed
no significant gain differences between both groups in our
study. This corresponds to the results of other studies with the
same group of participants.8,25
A stroke is often associated with shoulder pain. In stroke
patients with severe sensorimotor deficits, the occurrence
increases from 18% in the first week to 29% after 1 month and
Mangold et al / FES in Early Stroke Rehabilitation   189  
to 47% after 6 months.28 At baseline, most of our participants
were without shoulder pain. There was no significant change in
shoulder pain within the groups during the 4-week program and
no significant gain difference between either group. Electrical
stimulation can be used to treat shoulder pain, but our stimulation
protocol was not adjusted to this special objective. Furthermore,
beneficial effects of FES become obvious in participants who
already have developed shoulder pain29 and not in groups
including only few participants with shoulder pain.12,30
Our study was performed with a small group of patients and
group imbalances weakened the validity of the intergroup
comparisons. Hence, the following conclusions have to be
appraised in this context. In acute, severely affected stroke
patients, our stimulation protocol (12 sessions with 30 minutes
of stimulation time) does not induce superior improvements
compared to the same amount of conventional upper extremity
motor training. We assume that within normal rehabilitation
conditions in our country, it would be necessary for people to
continue FES themselves at home. This would probably require
a simpler stimulation protocol. Based on our findings, further
studies have to be carried out in at least 50 participants
randomized to FES and conventional therapy, and additional
dose-response studies with our protocol are necessary to identify
the required minimum number of FES training sessions.
Acknowledgments
The authors thank the physicians of the Reha Rheinfelden
in Switzerland for assessing subjects for eligibility; the
occupational therapists, physiotherapists, and nursery staff for
thoroughly carrying out the treatment and assessment; and all
patients for their participation in the study. This work was
supported in part by the National Center of Competence
NCCR: Plasticity and Repair.
References
  1. Parker VM, Wade DT, Langton Hewer R. Loss of arm function after
stroke: measurement, frequency, and recovery. Int Rehabil Med. 1986;8
(2):69-73.
  2. Kwakkel G, Kollen BJ, van der Grond J, Prevo AJ. Probability of regain-
ing dexterity in the flaccid upper limb: impact of severity of paresis and
time since onset in acute stroke. Stroke. 2003;34(9):2181-2186.
  3. Bobath B. Adult hemiplegia: evaluation and treatment. London, England:
Heinemann Medical Books; 1990.
  4. Langhammer B, Stanghelle JK. Bobath or motor relearning program?
A comparison of two different approaches of physiotherapy in stroke
rehabilitation: a randomized controlled study. Clin Rehabil. 2000;14(4):
361-369.
  5. Woldag H, Hummelsheim H. Evidence-based physiotherapeutic concepts
for improving arm and hand function in stroke patients: a review. J
Neurol. 2002;249(5):518-528.
  6. Berner YN, Lif Kimchi O, Spokoiny V, Finkeltov B. The effect of electric
stimulation treatment on the functional rehabilitation of acute geriatric
patients with stroke—a preliminary study. Arch Gerontol Geriatr.
2004;39(2):125-132.
  7. Popovic MB, Popovic DB, Sinkjaer T, Stefanovic A, Schwirtlich L.
Clinical evaluation of functional electrical therapy in acute hemiplegic
subjects. J Rehabil Res Dev. 2003;40(5):443-453.
190   Neurorehabilitation and Neural Repair
  8. Alon G, Levitt A, McCarthy PA. Functional electrical stimulation
enhancement of upper extremity functional recovery during stroke reha-
bilitation: a pilot study. Neurorehabil Neural Repair. 2007;21:207-215.
  9. Popovic MR, Thrasher TA, Zivanovic V, Takaki J, Hajek V. Neuroprosthesis
for retraining reaching and grasping functions in severe hemiplegic
patients. Neuromodulation. 2005;8(1):60-74.
10. Popovic DB, Popovic MB, Sinkjaer T, Stefanovic A, Schwirtlich L.
Therapy of paretic arm in hemiplegic subjects augmented with a neural
prosthesis: a cross-over study. Can J Physiol Pharmacol. 2004;82(8-9):
749-756.
11. Chae J, Ng A, Yu DT, et al. Intramuscular electrical stimulation for shoul-
der pain in hemiplegia: does time from stroke onset predict treatment
success? Neurorehabil Neural Repair. 2007;21:561-567.
12. Ring H, Rosenthal N. Controlled study of neuroprosthetic functional
electrical stimulation in sub-acute post-stroke rehabilitation. J Rehabil
Med. 2005;37(1):32-36.
13. Popovic MR, Keller T. Modular transcutaneous functional electrical
stimulation system. Med Eng Phys. 2005;27(1):81-92.
14. Gowland C, Stratford P, Ward M, et al. Measuring physical impairment
and disability with the Chedoke-McMaster Stroke Assessment. Stroke.
1993;24(1):58-63.
15. Jansa J, Pogacnik T, Gompertz P. An evaluation of the Extended Barthel
Index with acute ischemic stroke patients. Neurorehabil Neural Repair.
2004;18(1):37-41.
16. Gregson JM, Leathley M, Moore AP, Sharma AK, Smith TL, Watkins CL.
Reliability of the Tone Assessment Scale and the Modified Ashworth
Scale as clinical tools for assessing poststroke spasticity. Arch Phys Med
Rehabil. 1999;80(9):1013-1016.
17. Pandyan AD, Johnson GR, Price CI, Curless RH, Barnes MP, Rodgers H.
A review of the properties and limitations of the Ashworth and Modified
Ashworth Scales as measures of spasticity. Clin Rehabil. 1999;13(5):
373-383.
18. Hsieh YW, Wang CH, Wu SC, Chen PC, Sheu CF, Hsieh CL. Establishing
the minimal clinically important difference of the Barthel Index in stroke
patients. Neurorehabil Neural Repair. 2007;21(3):233-238.
19. Bortz J, Lienert GA. Kurzgefasste Statistik für die klinische Forschung.
Leitfaden für die verteilungsfreie Analyse kleiner Stichproben. Berlin,
Germany: Springer-Verlag; 2003.
20. Cohen J. Statistical power analysis for the behavioral sciences. Hillsdale,
NJ: Lawrence Earlbaum Associates; 1988.
21. Yozbatiran N, Donmez B, Kayak N, Bozan O. Electrical stimulation of
wrist and fingers for sensory and functional recovery in acute hemiplegia.
Clin Rehabil. 2006;20(1):4-11.
22. Houlden H, Edwards M, McNeil J, Greenwood R. Use of the Barthel Index
and the Functional Independence Measure during early inpatient rehabilita-
tion after single incident brain injury. Clin Rehabil. 2006;20(2):153-159.
23. Chae J, Bethoux F, Bohine T, Dobos L, Davis T, Friedl A. Neuromuscular
stimulation for upper extremity motor and functional recovery in acute
hemiplegia. Stroke. 1998;29(5):975-979.
24. Powell J, Pandyan AD, Granat M, Cameron M, Stott DJ. Electrical
stimulation of wrist extensors in poststroke hemiplegia. Stroke.
1999;30(7):1384-1389.
25. Siekierka-Kleiser EM, Kleiser R, Wohlschläger AM, Freund HJ, Seitz RJ.
Quantitative assessment of recovery from motor hemineglect in acute
stroke patients. Cerebrovasc Dis. 2006;21(5-6):307-314.
26. Daprati E, Sirigu A. Laterality effects on motor awareness.
Neuropsychologia. 2002;40(8):1379-1386.
27. Ratnasabapathy Y, Broad J, Baskett J, Pledger M, Marshall J, Bonita R.
Shoulder pain in people with a stroke: a population-based study. Clin
Rehabil. 2003;17(3):304-311.
28. Church C, Price C, Pandyan AD, Huntley S, Curless R, Rodgers H.
Randomized controlled trial to evaluate the effect of surface neuromuscular
electrical stimulation to the shoulder after acute stroke. Stroke.
2006;37(12):2995-3001.
29. Chantraine A, Baribeault A, Uebelhart D, Gremion G. Shoulder pain and
dysfunction in hemiplegia: effects of functional electrical stimulation.
Arch Phys Med Rehabil. 1999;80(3):328-331.
30. Renzenbrink GJ, Ijzerman MJ. Percutaneous neuromuscular electrical stimu-
lation (P-NMES) for treating shoulder pain in chronic hemiplegia. Effects on
shoulder pain and quality of life. Clin Rehabil. 2004;18(4):359-365.