Annals of Physical and Rehabilitation Medicine 64 (2021) 101388

Available online at

ScienceDirect
www.sciencedirect.com

Review

Functional electrical stimulation of the peroneal nerve improves

post-stroke gait speed when combined with physiotherapy.
A systematic review and meta-analysis
Maira Jaqueline da Cunha a,b, Katia Daniele Rech a,b, Ana Paula Salazar a,b,
Aline Souza Pagnussat a,b,c,*
a
Rehabilitation Sciences Graduate Program, Universidade Federal de Ciências da Saúde de Porto Alegre (UFCSPA), 245, Sarmento Leite Street, 90050-170 Porto
Alegre, RS, Brazil
b
Movement Analysis and Neurological Rehabilitation Laboratory, UFCSPA, Porto Alegre, RS, Brazil
c
Health Sciences Graduate Program, Universidade Federal de Ciências da Saúde de Porto Alegre (UFCSPA), Porto Alegre, RS, Brazil

A R T I C L E I N F O A B S T R A C T

Article history: Background: Functional electrical stimulation (FES) applied to the paretic peroneal nerve has positive
Received 3 May 2019 clinical effects on foot drop secondary to stroke.
Accepted 29 March 2020 Objective: To evaluate the effectiveness of FES applied to the paretic peroneal nerve on gait speed, active
ankle dorsiflexion mobility, balance, and functional mobility.
Keywords: Methods: Electronic databases were searched for articles published from inception to January 2020. We
Stroke rehabilitation included randomized controlled trials or crossover trials focused on determining the effects of FES
Lower extremity
combined or not with other therapies in individuals with foot drop after stroke. Characteristics of
Electric stimulation
Neurologic gait disorders
studies, participants, comparison groups, interventions, and outcomes were extracted. Statistical
heterogeneity was assessed with the I2 statistic.
Results: We included 14 studies providing data for 1115 participants. FES did not enhance gait speed as
compared with conventional treatments (i.e., supervised/unsupervised exercises and regular activities
at home). FES combined with supervised exercises (i.e., physiotherapy) was better than supervised
exercises alone for improving gait speed. We found no effect of FES combined with unsupervised
exercises and inconclusive effects when FES was combined with regular activities at home. When FES
was compared with conventional treatments, it improved ankle dorsiflexion, balance and functional
mobility, albeit with high heterogeneity for these last 2 outcomes.
Conclusions: This meta-analysis revealed low quality of evidence for positive effects of FES on gait speed
when combined with physiotherapy. FES can improve ankle dorsiflexion, balance, and functional
mobility. However, considering the low quality of evidence and the high heterogeneity, these results
must be interpreted carefully.
C 2020 Elsevier Masson SAS. All rights reserved.

1. Introduction dorsiflexion during the swing phase of gait [4]. Consequently, foot
drop contributes to disruption in weight acceptance and weight
Foot drop is a common impairment after stroke. This condition transfer; reduces walking speed, efficiency and stability of gait [5]
is related to high degrees of motor impairment, weakness or lack of and increases the risk of falling [6].
voluntary control of ankle dorsiflexors and increased spasticity of Electrical stimulation is a therapeutic method able to reduce post-
plantar flexors [1–3]. Foot drop interferes with the initial foot stroke walking dysfunctions [7–9]. These benefits have been
contact at the beginning of the stance phase and hinders ankle demonstrated for some types of stimulation: transcutaneous electrical
nerve stimulation (TENS) [9], functional electrical stimulation (FES) on
the peroneal nerve and FES on tibialis anterior muscle [8]. However,
* Corresponding author at: Rehabilitation Sciences Graduate Program, Universi- some evidence suggests that FES is more effective than TENS [10]. FES
dade Federal de Ciências da Saúde de Porto Alegre (UFCSPA), 245, Sarmento Leite is able to stimulate voluntary muscle activity, reduce foot drop,
Street, 90050-170 Porto Alegre, RS, Brazil.
E-mail address: [email protected] (A.S. Pagnussat).
decrease spasticity, and lead to long-term sensorimotor cortical

https://doi.org/10.1016/j.rehab.2020.03.012
1877-0657/ C 2020 Elsevier Masson SAS. All rights reserved.
2 M. Jaqueline da Cunha et al. / Annals of Physical and Rehabilitation Medicine 64 (2021) 101388

reorganization [11,12]. When FES is applied on the peroneal nerve and 2.3. Study selection
synchronized with gait phases [13], it allows for a more precise ankle
dorsiflexion/eversion during gait [13] as compared with FES applied on Inclusion criteria were reports of RCTs or crossover trials
the tibialis anterior. FES of the peroneal nerve could also facilitate hip focused on determining the effects of FES applied to the paretic
and knee flexion during the swing phase of gait and lead to a walking peroneal nerve of post-stroke individuals with foot drop. The
pattern closer to the normal [11,14]. The stimulus can be applied with stimulus could be applied using FES conventional equipment or
conventional electrical stimulation equipment used in current clinical foot drop stimulators with implanted or non-implanted electrodes.
practice or by foot drop stimulator devices with tilt sensors or foot For crossover trials, we analyzed only data from the first period and
switches. Several foot-drop stimulators are commercially available considered them a parallel-group trial.
[14,15]. We included articles that combined or did not combine FES
Some studies have demonstrated that FES applied to the paretic with other therapies (supervised exercises, gait training, treadmill
peroneal nerve has positive clinical effects on foot drop secondary training, AFO, unsupervised exercises, regular activities at home
to stroke [16–18]. A previous systematic review showed that FES and others). Studies must have investigated the effect of FES on gait
could increase gait speed in post-stroke individuals as compared speed, active ankle dorsiflexion mobility, balance and functional
with ankle foot orthoses (AFO) [13]. However, this study included mobility.
only one randomized controlled trial (RCT). A more recent Comparator groups were sham FES, combined or not with other
systematic review included 6 RCTs and showed that both foot therapies, and conventional physiotherapy (treadmill gait training,
drop stimulation devices and AFO seem to be effective and robot-assisted gait training, gait re-education, neurodevelopmen-
‘‘equivalent’’ for increasing gait speed after stroke [19]. These tal treatment, proprioceptive neuromuscular facilitation, conven-
authors argue that probably a combination of physiotherapy and tional walking devices such as AFO and orthopedic shoes).
devices would be necessary to improve gait speed in this We included studies written in English, Spanish and Portu-
population. From the available literature and considering the lack guese. We excluded studies that enrolled infant stroke survivors as
of a meta-analysis exploring the effects of FES combined with well as theses or articles published only as abstracts, including
conventional therapy, we cannot make evidence-based inferences conference proceedings.
about the clinical benefits of FES of the peroneal nerve to treat foot Two reviewers (MJC and KDR) independently screened studies
drop after stroke. New data from RCTs that applied a stimulus on based on title and abstracts. The same reviewers deleted duplicates
the peroneal nerve by conventional equipment or implanted and independently assessed the full texts. When different
electrodes [17,18,20–22], combined or not with a conventional assessments from the same trial were reported in separate
therapy, must be considered. These data must be synthesized in a publications, we considered them as a single publication.
systematic review with meta-analysis to add value to the Disagreements were solved by consensus or with a third reviewer
previously published studies and explore the effectiveness of (APS).
FES on the peroneal nerve in post-stroke individuals.
For this purpose, we performed a systematic review of RCTs and 2.4. Data extraction and assessment of characteristics of trials
crossover trials to verify whether FES applied to the paretic
peroneal nerve, combined or not with conventional therapy, could We extracted data on methodological characteristics, number of
enhance gait speed in post-stroke individuals with foot drop. participants, comparison groups, interventions and outcomes by using
Second, we investigated the effect of this therapy on active ankle standardized forms. When studies did not have enough data available
dorsiflexion mobility, balance, and functional mobility. Our or we needed clarification, we contacted authors by e-mail.
findings may provide solid evidence to empower post-stroke
individuals and support clinicians in their decision-making when 2.5. Outcomes
dealing with post-stroke gait rehabilitation.
Our primary outcome was gait speed, which was assessed
during the 10-m Walk Test (10MWT) with stopwatches, inertial
2. Methods sensors or optoelectronic systems. Gait speed was calculated as the
ratio between distance and time spent (seconds) to walk 10 m.
2.1. Protocol and registration When authors presented data as the ‘‘time to complete the test’’,
we transformed values in meters per seconds (m/s) by using the
This systematic review and meta-analysis was conducted ratio of distance to time.
according to the Preferred Reporting Items for Systematic Reviews Secondary outcomes were:
and Meta-Analysis (PRISMA) guidelines [23]. The protocol was
registered at the International Prospective Register of Systematic  active ankle dorsiflexion mobility (data collected during gait
Reviews, PROSPERO (No. CRD42019127552) and can be accessed assessment by using optoelectronic systems);
at http://www.crd.york.ac.uk/PROSPERO/display_record.  balance assessed by the Berg Balance Scale (BBS) and;
php?ID=CRD42019127552.  functional mobility assessed by the Timed Up and Go test (TUG).

2.2. Literature search Although authors also collected data from the time participants
were receiving the FES stimulation, we analyzed data collected
The following electronic databases were searched for articles only when stimulation was off.
published from inception to January 2020: MEDLINE via PubMed,
Embase, Cochrane Central Register of Controlled Literature Search 2.6. Quality assessments
Trials (CENTRAL) and Physiotherapy Evidence Database (PEDro). In
addition, we screened studies from reference lists of manuscripts PEDro scores from the Physiotherapy Evidence Database were
included in this review and checked unpublished or ongoing used to assess the quality of studies (www.pedro.org.au). The
studies in clinical trial registries (ClinicalTrials.gov). The search PEDro scale contains 11 items and, except for item one, pertaining
strategy was adjusted to each database. Appendix 1 shows the to external validity, each satisfied item contributes one point to the
complete search strategy. total score (range 0–10) [24]. Scores 9 or 10 were considered
 M. Jaqueline da Cunha et al. / Annals of Physical and Rehabilitation Medicine 64 (2021) 101388 3

excellent quality; 6 to 8 good quality; 4 or 5 fair quality; met the eligibility criteria and were included. Thirteen studies
and < 4 poor quality. Scores 0 to 5 correspond to high risk of bias were included in the meta-analysis, providing data from 1115 par-
and 6 to 10 low risk of bias [25]. We used the grading of ticipants. Eleven studies reported data on gait speed and expressed
recommendation, assessment, development and evaluation data in m/s. One study [18] described gait speed as time spent to
(GRADE) system to assess the quality of the body of evidence [26]. walk 10 m. We contacted the author and performed the procedure
to estimate data, converting them to the same scale (m/s). One
2.7. Data analysis study was not included in the meta-analysis [29] because it used
the same sample presented by Bethoux (2014) [30]. Another
Between-group meta-analyses were conducted for quantitative 2 studies presented outcome data from the same trial, and we
analysis. Data were obtained by comparing the change from post- considered them as a single publication [16,22]. These studies
to pre-intervention. Data were pooled by using a random effects published by Kottink and colleagues in 2007 and 2012 described
model for quantitative synthesis [27]. Outcomes are presented as data on gait speed from the same participants. Thus, we considered
the weighted mean difference (MD) or standardized mean data from only Kottink and colleagues (2007) because they
differences (SMD) with 95% confidence intervals (95% CIs). reported the complete sample.
Statistical heterogeneity was assessed with the I2 statistic, with
values classified as low heterogeneity (I2 < 25%), moderate 3.2. Participants
heterogeneity (I2 25% to 50%) and high heterogeneity (I2 > 50%)
[27]. P < 0.05 was considered statistically significant. Considering Mean age of participants ranged from 45 [17] to 72 years [31]. The
that studies had varied types of protocols, a sensitivity analysis was mean time since stroke ranged from < 1 month [32] to almost
conducted to reduce the statistical heterogeneity. We considered 108 months. Twelve studies included post-stroke individuals in the
the time after stroke (acute/subacute and chronic) and the type of chronic phase [16–18,22,29,30,33–37]. Three trials included parti-
protocol applied (FES combined with supervised or unsupervised cipants in both acute and subacute phases [31,32,38]. Six studies
exercises and regular activities at home). All analyses were included both ischemic and hemorrhagic stroke [17,18,31,
conducted with R v3.3.3 (package metaphor v2.0-0) [28]. 35,37,38]. Other trials did not describe the type of stroke.

3. Results 3.3. Interventions

Table 1 depicts all information about outcomes, measurement Twelve studies used peroneal nerve devices [16–18,22,30–
tools and characteristics of participants such as age, sex and time 35,37,38] and 2 used conventional FES on the peroneal nerve
after stroke. [20,21]. Only one study used implanted electrodes to stimulate the
peroneal nerve [16,22].
3.1. Flow of articles in the review In 3 studies, participants did not perform specific exercises
while wearing the stimulation devices. They were instructed
Fig. 1 shows the flow of articles in the review. The electronic simply to follow their regular activities at home (such as regular
search strategy identified 2328 studies. After titles and abstracts walking activities throughout the day) [22,30,34]. Four studies
screening, 46 manuscripts were considered relevant and 14 trials prescribed a routine of exercises focusing on gait training while

Table 1
Summary of included studies.

Study Design Participants Protocol Outcome measures

Frequency and duration for the Characteristics

different groups

Bae et al., 2014 RCT n = 20 IG IG Tempo-spatial gait

Age (year) 30 min a day, 3 days/wk, during Robot-assisted gait training plus FES to parameters
IG: 45.4  19.7 5 wks stimulate the peroneal (device) Dynamic angular gait
CG: 52.0  16.1 + + parameters
Sex: 13M/7F 30 min a day, 3 days/wk, during Conventional physiotherapy Modified Motor
Type of stroke: I/H 5 wks CG Assessment Scale
Severity: N/S CG Robot-assisted gait training TUG
Time since stroke: >6 mo 30 min a day, 3 days/wk, during + BBS
(chronic) 5 wks Conventional physiotherapy Timing: 0; 5 wks
Follow-up: not reported +
30 min a day, 3 days/wk, during
5 wks
Bethoux et al., 2014 RCT n = 495 IG IG 10M Walk Test
Bethoux et al., Age (year) All day at home during 6 mo Participants used the FES to stimulate the Activities of Daily Living
2015a IG: 63.87  11.33 CG peroneal (device) at home Stroke Impact Scale
CG: 64.3  12.01 All day at home during 6 mo CG Serious adverse event
Sex: 304M/191F Participants used the AFO device at home 6-Min Walk Test
Type of stroke: N/S GaitRite Functional
Severity: N/S Ambulation Profile
Time since stroke: > 6 mo Modified Emory Functional
(chronic) Ambulation Profile
Follow-up: 26 weeks BBS
TUG
Gait quality
Stroke-Specific Quality of
Life
Timing: 0; 24; 48a wk
4 M. Jaqueline da Cunha et al. / Annals of Physical and Rehabilitation Medicine 64 (2021) 101388

Table 1 (Continued )

Study Design Participants Protocol Outcome measures

Frequency and duration for the Characteristics

different groups

Burridge et al., RCT n = 32 IG IG Gait speed

1997 Age (year) All day at home during 6 wks Participants used the FES to stimulate the Physiological Cost Index
IG: 52.3  14.3 CG peroneal (device) on a full-time basis for all Timing: 0; 4; 12 wks
CG: 61.3  8.6 10 physiotherapy sessions walking activities throughout the day
Sex: 23M/9F during the first month of the +
Type of stroke: N/S trial. Each session lasted for 1 h Conventional physiotherapy
Severity: N/S CG
Time since stroke: > 6 mo Conventional physiotherapy
(chronic)
Follow-up: 6 wks
Everaert et al., Cross-over n = 69 IG IG Figure-of-8 Walking Speed
2013 RCT Age (year) All day at home during 6 wks Participants used the FES to stimulate the Physiological Cost Index
IG: 57.1  12.9 CG peroneal (device) daily at home and for 10M Walk Test
CG: 55.6  11.9 All day at home during 6 wks walking in the community Mobility Index
Sex = 51M, 18F CG Perceived Safety Level
Type of stroke = n/s Participants used the AFO daily at home and Device Preference
Severity: not reported for walking in the community Timing: 0; 6 wks
Time since stroke: > 6 mo
(chronic)
Follow-up: not reported
Hwand et al., RCT n = 30 IG IG TUG
2015 Age (year) 30 min a day/4 wks Treadmill training plus FES to stimulate the BBS
IG: 50.0  7.55 + peroneal (device) 10M Walk Test
GC: 49.47  5.01 30 min twice per day/4 wks + Muscle structure of the
Sex: 17M/13F CG Conventional physiotherapy tibialis anterior
Type of stroke: I/H 30 min a day/4 wks CG Timing: 0; 4 wks
Severity: not reported + Treadmill training plus placebo of FES to
Time since stroke: > 6 mo 30 min twice per day/4 wks stimulate the peroneal (device)
(chronic) +
Follow-up: not reported Conventional physiotherapy
Kluding et al., RCT n = 197 IG IG 10M Walk Test
2013 Age (year) All day at home during 30 wks Participants used the FES to stimulate the comfortable/fast speed
IG: 60.71  12.24 + peroneal (device) at home Fugl-Meyer Assessment-
CG: 61.58  10.98 8 sessions during the first 6 wks + Lower Limb
Sex: 118M/79F CG Physical therapy focused on education on TUG
Type of stroke: I/H All day at home during 30 wks device use, gait training, and an 6-min Walk Test
Severity: N/S + individualized home exercise program BBS
Time since stroke: > 6 mo 8 sessions during the first 6 wks CG Functional reach test
(chronic) Participants used the AFO at home Stroke Impact Scale
Follow-up: not reported + Activity monitoring
Physical therapy focused on education on Timing: 0; 6; 12; 30 wks
device use, gait training, and an
individualized home exercise program
Participants received surface sensory
stimulation with TENS device
Kottink et al., RCT n = 23 IG IG Tempo-spatial gait
2012 Age (year) All day at home during 26 wks Participants used the FES to stimulate the parameters
IG: 55.6  13.16 CG peroneal (implantable) at home Dynamic angular gait
CG: 53.31  10.55 All day at home during 26 wks CG parameters
Sex: 15M/8F Participants used the conventional device Timing: 0; 4; 8; 12; 26 wks
Type of stroke: N/S at home
Severity: N/S
Time since stroke: > 6 mo
(chronic)
Follow-up: not reported
Kottink et al., RCT n = 29 IG IG 6-min Walk Test
2007 Age (year) All day at home during 26 wks Participants used the FES to stimulate the 10M Walk Test
IG: 55.2  11.36 CG peroneal (implantable) at home Physical activity
CG: 53.31  9.87 All day at home during 26 wks CG Timing: 0; 4; 8; 12; 26 wks
Sex: 20M/9F Participants used the conventional device
Type of stroke: N/S at home
Severity: N/S
Time since stroke: > 6 mo
(chronic)
Follow-up: N/S
 M. Jaqueline da Cunha et al. / Annals of Physical and Rehabilitation Medicine 64 (2021) 101388 5

Table 1 (Continued )

Study Design Participants Protocol Outcome measures

Frequency and duration for the Characteristics

different groups

Morone et al., RCT n = 20 IG IG 10M Walk Test

2012 Age (year) 40 min a day, 5 days/wk during Walking training plus FES to stimulate the Functional Ambulation
IG: 53.3  14.6 at 12 wks peroneal (device) Classification
CG: 61.2  16.2 + + Barthel Index
Sex: no report 40 min a day, 5 days/wk during Conventional physiotherapy Mobility Index
Severity: N/S at 12 wks CG Medical Research Council
Type of stroke: N/S CG Walking training plus AFO Canadian Neurological
Time since stroke: < 1 mo 40 min a day, 5 days/wk during + Scale
(acute) 12 wks Conventional physiotherapy Ashworth scale
Follow-up: N/S + Timing: 0; 4; 8; 12 wks
40 min a day, 5 days/wk during
12 wks
Mitsutake et al., RCT n = 23 IG IG 10M Walk Test
2019 Age (year) 20 min a day, all days during Walking plus FES to stimulate the peroneal Timing: 0; 2; wks
IG: 61.45  11.51 2 wks (device)
CG: 68.75  10.59 + +
Sex: 18M, 5F 40 min a day, all days during Conventional physiotherapy plus NMES
Severity: moderate 2 wks (20 min)
Type of stroke: I/H CG CG
Time since stroke: < 6 mo 20 min a day, all days during Training plus NMES
(sub-acute) 2 wks +
Follow-up: N/S + Conventional physiotherapy
40 min a day, all days during
2 wks
Salisbury et al., RCT n = 16 IG IG 10M Walk Test
2013 Age (year) All day during 12 wks Participants used the FES to stimulate the Functional Ambulation
IG: 72.9.3  13.9 + peroneal (device) at home Classification
CG: 52.6  17.2 20 min a day, 5 days/wk during + Stroke Impact scale
Sex = 6M, 10F 12 wks Routine gait re-education Perception of change in
Severity: N/S CG CG walking
Type of stroke = I/H All day during at 12 wks Participants used the AFO at home Timing: 0; 12 wks
Time since stroke: + +
> 1 < 3 mo (sub-acute) 20 min a day, 5 days/wk during Routine gait re-education
Follow-up: N/S 12 wks
Sheffler et al., RCT n = 110 IG IG Tempo-spatial gait
2015 Age (years) All day during 12 wks Participants used the FES peroneal nerve at parameters
IG: 52.8  12.2 + home Dynamic angular gait
CG: 53.2  10.1 1 h functional training (first + parameters
Sex = 67M, 43F 5 wks) + 8 h per day device Functional training Timing: 0; 12; 24; 36 wks
Severity: N/S usage GC
Type of stroke = I/H 2 h functional training (last Participants used the AFO at home
Time since stroke: > 6 mo 7 wks) + 8 h per day device +
(chronic) usage Functional training
Follow-up: 12 and 24 wks CG
All day during 12 wks
+
1 h functional training (first
5 wks) + 8 h per day device
usage
2 h functional training (last
7 wks) + 8 h per day device
usage
Park et al., RCT n = 30 IG IG Tempo-spatial gait
2017 Age (year) 30 min, 1 day FES on peroneal nerve parameters
IG: 64.2  9.7 CG + Timing: 0; 4 days
CG: 63.5  8.6 30 min, 1 day Conventional physiotherapy
Sex = N/S CG
Type of stroke = N/S TENS on sural nerve
Severity: N/S +
Time since stroke: > 6 mo Conventional physiotherapy
(chronic)
Follow-up: N/S
Sharif et al., RCT n = 38 IG IG Fugl-Meyer Assessment-
2017 Age (year): N/S 30 min a day, 5 days/wk during FES on peroneal nerve during swing phase Lower Limb
Sex: N/S 6 wks of walking Modified Ashworth Scale
Type of stroke: N/S CG + BBS
Severity: N/S 10 min a day, 5 days/wk during Physiotherapy and occupational therapy TUG
Time since stroke: > 6 mo 6 wks CG Gait Dynamic Index
(chronic) Conventional electrical stimulation on Timing: 0; 3; 6 wks
Follow-up: not reported peroneal nerve
+
Physiotherapy and occupational therapy
RCT: randomized controlled trial; N/S, not stated; IG: intervention group; CG: control group; M: male; F: remale; I: ischemic; H: hemorrhagic; wk(s): weeks; mo: months;
FES: functional electrical stimulation; AFO: ankle foot orthosis; TENS: transcutaneous electric nerve stimulation; TUG: Time Up and Go; BBS: Berg Balance Scale.
a
Study included only in the systematic review and not in the meta-analysis.
6 M. Jaqueline da Cunha et al. / Annals of Physical and Rehabilitation Medicine 64 (2021) 101388

Fig. 1. Flow of articles in the study. RCT: randomised controlled trial.

using the stimulation devices [31,33,35,37]. In these studies, the The duration of sessions ranged from 20 to 60 min
authors compared FES stimulation with AFO devices. [17,18,31,37,38]. In some studies, participants received the
In 3 studies, FES was associated with conventional physiother- stimulation at home during the entire day [16,22,29–31,33–35,37].
apy and compared with AFO [32] or with other types of stimulation Frequency of FES stimulation ranged from 1 [21] to 5
(TENS and neuromuscular electrical stimulation [NMES]) [20,31,32,38] days per week or every day at home for 4 to
[20,21,38]. Two studies combined FES with treadmill gait training 30 weeks. The intervention period studied ranged from 2 weeks
and compared to sham plus treadmill gait training [17,18]. [38] to 30 weeks [35] or just 1 day.

Table 2
PEDro scores of included studies.

Study Random Concealed Groups Participant Therapist Assessor Adequate Intention- Between-group Point estimate Total
allocation allocation similar at blinding blinding blinding follow-up to-treat difference and variability (0 to 10)
baseline analysis reported reported

Bae et al., 2014 Y Y Y N N N Y N Y Y 6

Bethoux et al., 2014 Y N Y N N N N Y Y Y 5
Bethoux et al., 2015a Y N Y N N N N Y Y Y 5
Burridge et al., 1997 Y Y N N N N Y N Y Y 5
Everaert et al., 2013 Y Y Y N N N Y N Y Y 6
Hwang et al., 2015 Y N Y N N Y Y N Y Y 7
Klulding et al., 2013 Y N N N N Y N Y Y Y 5
Kottink et al., 2007 Y Y Y N N N Y Y Y Y 7
Kottink et al., 2012 Y Y Y N N N N N Y Y 5
Mitsutake et al., 2019 Y N Y N N Y Y N Y Y 6
Morone et al., 2012 Y N Y N N Y N N Y Y 5
Park et al., 2017 Y N Y N N N N N Y Y 4
Salisbury et al., 2013 Y Y Y N N N Y Y Y Y 7
Sharif et al., 2017 Y N Y N N N N N Y Y 4
Sheffer et al., 2015 Y Y Y N N N N Y Y Y 6
Median (min-max) 5 (4–7)
a
This study was not included in the meta-analysis because the follow-up was higher than in other studies.
 M. Jaqueline da Cunha et al. / Annals of Physical and Rehabilitation Medicine 64 (2021) 101388 7

Table 3
Quality Assessment (GRADE).

Quality assessment (GRADE) Summary of findings

No. of Risk of bias Inconsistency Indirectness Imprecision Publication Overall Study event rates (%) Risk difference with FES
participants bias certainty of on peroneal nerve
(studies) evidence
Follow-up With conventional With FES on
therapy peroneal nerve

Gait speed
1077 Serious a,b,c Serious Not serious Not serious None 540 537 SMD 0.06 SD higher
(12 RCTs) LOW (0.4 lower to 0.52 higher)
Ankle dorsiflexion angle
151 Serious a,e Not serious Not serious Serious d
None 78 73 MD 3.3 higher
(3 RCTs) LOW (1.48 higher to 5.12 higher)
Balance (assessed with Berg Balance Scale)
780 Serious a,f Serious Not serious Not serious None 395 385 MD 2.76 higher
(5 RCTs) LOW (0.64 higher to 4.88 higher)
Functional mobility (assessed with Time Up and Go)
780 Serious a,h Serious g Not serious Not serious None 395 385 MD 3.19 lower
(5 RCTs) LOW (5.76 lower to 0.62 lower)

CI: confidence interval; SMD: standardized mean difference; MD: mean difference. dStudies did not blind assessors. e2 of 4 studies did not use intention-to-treat analysis.
a
Studies did not blind participants and therapists.
b
9 of 11 studies did not blind assessors.
c
6 of 11 studies did not use intention-to-treat analysis.
c
Insufficient number of studies.
e
Assessor was not blinded. Not all studies used intention-to-treat analysis.
f
2 of 5 studies did not blind assessors and did not use intention to treat analysis.
g
High heterogeneity and the confidence intervals do not overlap.

3.4. Quality of studies 3.6. Active ankle dorsiflexion mobility

The median PEDro score was 5 (range 4 to 7), so studies Three studies (n = 151) assessed the effect of FES on active ankle
presented fair quality and high risk of bias (Table 2). All trials were dorsiflexion mobility of the paretic limb and were included in the
randomized, showed between-group differences, and reported meta-analysis [16,17,37]. These studies applied FES combined or
point estimate and variability. Thirteen studies had similar groups not with supervised exercises. FES could improve active ankle
at baseline [17,18,20,21,29–32,34,37,38]. Seven trials concealed dorsiflexion as compared with conventional treatment [MD = 3.30
allocation lists [16,17,22,31,33,34,37] and 7 had adequate follow- (95% CI: 1.48 to 5.12; I2 0%, P = 0.0007)] (Fig. 4A).
up [17,18,22,31,33,34,38]. Only 4 studies blinded assessors
[18,32,35,38], and 6 studies used intention-to-treat analysis 3.7. Balance and functional mobility
[22,29–31,35,37]. No study blinded therapists or participants.
Quality of evidence was low for all outcomes. Table 3 presents the Five studies (n = 780) were included in this meta-analysis
evidence profile according to the GRADE system. [17,18,20,30,35]. These studies applied FES combined or not with
supervised exercises. FES could improve balance (evaluated by the
3.5. Outcomes measures BBS) as compared with conventional treatments [MD = 2.76 (95%
CI: 0.64 to 4.88; I2 90%, P = 0.011)]. However, the heterogeneity
3.5.1. Gait speed was high (Fig. 4B).
Twelve studies (n = 1077) evaluated gait speed and were included The same 5 studies (n = 780) evaluated functional mobility,
in the meta-analysis. FES did not enhance gait speed as compared assessed by the time to complete the TUG test, and were included
with conventional treatment [SMD = 0.092 (95% CI: 0.34 to 0.53; I2 in the meta-analysis [17,18,20,30,35]. FES induced functional
89%, P = 0.68)] (Fig. 2A). Sensitivity analysis was performed mobility improvements as compared with conventional treat-
considering the time after the stroke. Groups did not differ in gait ments [MD = 3.19 (95% CI: 5.76 to 0.62; I2 84%, P = 0.015)].
speed: 11 studies included chronic post-stroke individuals (n = 1008) However, the heterogeneity was high (Fig. 4C).
[SMD = 0.013 (95% CI: 0.49 to 0.52; I2 91%, P = 0.96)] and only
3 studies included participants in the acute/subacute phase (n = 69) 4. Discussion
[SMD = 0.41 (95% CI: 0.073 to 0.89; I2 0%, P = 0.096)].
Considering that studies included in this review applied varied This systematic review with meta-analysis aimed to determine
types of protocols, we conducted other sensitivity analyses. Four the effectiveness of FES applied to the paretic peroneal nerve of
studies combined FES with supervised programs of exercise (i.e., post-stroke individuals with foot drop. We analyzed the effect of
physiotherapy). A sensitivity analysis showed that FES combined FES combined or not with other therapies on gait speed, active
with physiotherapy could increase gait speed as compared with ankle dorsiflexion mobility, balance, and functional mobility.
physiotherapy alone (n = 133) [SMD = 0.51 (95% CI: 0.16 to 0.86; I2 Regardless of the duration of electrical stimulation, FES could
0%, P = 0.0042)] (Fig. 3). Four studies combined FES with improve gait speed only if it was associated with supervised
unsupervised home exercises (n = 355), and 3 studies used FES exercises (i.e., physiotherapy).
in regular activities at home (n = 589). A sensitivity analysis did not Some evidence indicates that FES can induce electrophysiologi-
reveal a significant difference between groups [SMD = 0.02 (95% cal modifications over time (e.g., increase maximal voluntary
CI: 0.23 to 0.19; I2 0%, P = 0.849) and SMD = 0.28 (95% CI: 1.53 contraction and motor-evoked potential), so FES could improve
to 0.96; I2 95%, P = 0.653)] (Fig. 3). voluntary motor control [11,39]. However, changes in walking
8 M. Jaqueline da Cunha et al. / Annals of Physical and Rehabilitation Medicine 64 (2021) 101388

Fig. 2. Effect of functional electrical stimulation (FES) of the peroneal nerve on gait speed (A). Sensitivity analysis considering acute/subacute and chronic stroke (B).

speed may be smaller than electrophysiological modifications changes for gait speed of 0.10 to 0.18 m/s [46,47]. Most studies
because walking requires synchronized activities of several that associated FES with physiotherapy reached a mean difference
muscles [11]. When combined with physiotherapy, FES could of 0.08 to 0.19 m/s. Studies that combined FES with unsupervised
improve gait speed, stability, and functional mobility [40,41]. A exercises and regular activities at home were ineffective in
possible explanation for this result is the improvement in ankle improving gait speed, and most reported minimal clinical
dorsiflexion movement. Once the ankle mobility increases, the changes < 0.10 m/s. Furthermore, the high heterogeneity of
initial foot contact on the ground changes from anterior to treatments may have influenced the results. Studies used an
posterior [42,43]. The improvement in foot clearance would imprecise control of the volume of unsupervised exercises and
ameliorate the forward progression and stability during locomo- regular activities at home, but supervised exercises (i.e., physio-
tion and could also improve gait speed. therapy) included strategies to improve posture, gait, balance and
Gait speed is the leading indicator of function, level of disability, functional mobility — outcomes assessed in this review.
and survival and is used to classify ambulation status after stroke Additional critical points are noteworthy when we analysed
[44,45]. The literature reports minimal clinically significant between-group differences in the studies. For instance, we need to
 M. Jaqueline da Cunha et al. / Annals of Physical and Rehabilitation Medicine 64 (2021) 101388 9

Fig. 3. Sensitivity analysis of protocols of treatment: unsupervised home exercises, physiotherapy, and regular activities at home.

consider the control groups. In 7 studies, FES was compared with part the improvements in balance and functional mobility.
AFO [16,22,29–32,34,35,37]. Two studies compared FES with Additionally, improvements in ankle mobility and tibialis anterior
treadmill gait training [17,18] and 2 compared it with conventional activity would facilitate the ankle strategy necessary to keep the
stimulation (TENS and NMES) [20,21,38]. Only 1 study compared standing to sitting, turning and transferring movements, as
FES with conventional physiotherapy [33]. Thus, differences in assessed by the BBS. Another hypothesis is that FES ameliorates
therapy delivered to the control group may have influenced the forward progression and stability during gait [42,43] and increases
results of gait speed. gait speed [13,19], and these improvements would also affect the
Post-stroke individuals usually present reduced or absent time to complete the TUG test. Minimal clinically significant
dorsiflexion during the heel strike and mid-swing phases of gait changes expected in the time to complete the TUG test range from
due to the loss of motor control [48]. Our results showed that FES 4 to 9 s [54]. Five studies were included in the meta-analysis for
improved the active ankle dorsiflexion angle during gait. Of note, this outcome. Three studies reported a mean difference between
2 studies included in the meta-analysis compared FES with AFO pre- and post-treatment greater than 4 s [17,18,20]. However, the
[16,37]: 1 study used a custom-molded hinged AFO with plantar heterogeneity was high in BBS and TUG analyses (90% for BBS and
flexion blocker and was fabricated with conventional techniques 84% for TUG). Also, 3 studies included in the TUG analysis included
[37] and the other study did not specify the type of device [16]. AFO gait training as part of the physiotherapy treatment
is commonly used as standard care to promote mobility and [17,18,20]. Considering that part of the TUG performance depends
independence [49]. This type of orthosis holds the ankle in neutral on gait velocity, this type of task-specific training [19] may have
position, maintains the foot in the swing phase of gait and avoids skewed the results of functional mobility.
the forefoot contact on the floor. AFO corrects the foot drop via a To the best of our knowledge, this is the first up-to-date and
passive mechanism and does not involve neuromuscular, spinal or comprehensive systematic review with meta-analysis that asses-
brain circuits [32,50]. For this reason, FES may be more effective sed FES applied to the peroneal nerve and investigated its effect on
than AFO in increasing active ankle dorsiflexion mobility. gait performance, active ankle dorsiflexion mobility, balance and
Impaired balance and functional mobility are also common functional mobility. This study has some limitations. For example,
after stroke and may interfere in the performance of daily life the diversity in time since the stroke, motor impairment, protocols
activities [51,52]. FES had positive effects on balance, as evaluated of treatment, and outcomes cannot be ignored. All these elements
by the BBS. The BBS is a good tool to assess static and dynamic can increase the heterogeneity and decrease the validity of our
aspects of balance and provides quantitative information about results. Although we stipulated extensive and specific eligibility
equilibrium and risk of falls [53]. Additionally, our meta-analysis criteria, the GRADE showed a high risk of bias of the studies
showed that FES was more effective than conventional treatments included in the review. The PEDro score was 5, reported as low-
in improving functional mobility. All studies included in this quality evidence. Most studies did not conduct intention-to-treat
review assessed functional mobility by the TUG test. The TUG test analyses, adequate follow-up or blinding. However, blinding
is a complex task, comprising several elements, including gait participants and therapists is difficult because the electrical
performance, mobility, and balance. By repetitive muscle contrac- stimulus promotes visible ankle dorsiflexion. Another important
tions, FES could increase sensory inputs to the brain and contribute limitation is the significant heterogeneity in gait speed, balance,
to the motor relearning [11,13]. This reasoning could explain in and functional mobility meta-analyses. Possibly, the high hetero-
10 M. Jaqueline da Cunha et al. / Annals of Physical and Rehabilitation Medicine 64 (2021) 101388

Fig. 4. Effect of FES of the peroneal nerve on active ankle dorsiflexion mobility (A), balance (B), and functional mobility (C).

geneity is due to differences in protocols and sample characte- Future studies may emphasize gait training and other challenging
ristics. Furthermore, most studies did not present detailed activities during physiotherapy combined with FES in order to obtain
information on the type of stroke and its severity. better outcomes. Further research with adequate methodological
quality is still necessary to determine the exact effects of FES on gait
speed, ankle dorsiflexion mobility, balance, and functional mobility.
5. Conclusions
Funding

Our meta-analysis showed positive effects of FES on the This work was partially supported by a grant from the Brazilian agency
peroneal nerve to improve gait speed when combined with CAPES (Coordenação de Aperfeiçoamento de Pessoal de Nı´vel Superior).
supervised exercises (i.e., physiotherapy). FES combined with
Disclosure of interest
unsupervised home exercises had no effect on gait speed. However,
taking into account the high heterogeneity in our analysis, we The authors declare that they have no competing interest.
could not determine the benefits of FES combined with regular
activities at home for improving gait speed. Additionally, our Appendix A. Supplementary data
findings showed a low quality of evidence for positive effects of FES
on active ankle dorsiflexion mobility, with high heterogeneity for Supplementary data associated with this article can be found, in
balance and functional mobility. the online version, at https://doi.org/10.1016/j.rehab.2020.03.012.
 M. Jaqueline da Cunha et al. / Annals of Physical and Rehabilitation Medicine 64 (2021) 101388 11

[28] Viechtbauer W. Conducting Meta-Analyses in R with the metafor Package.

References
J Stat Software 2010;36:1–48.
[29] Bethoux F, Rogers HL, Nolan KJ, Abrams GM, Annaswamy T, Brandstater M, et al.
[1] Chisholm AE, Perry SD, McIlroy WE. Correlations between ankle-foot impair- Long-Term Follow-up to a Randomized Controlled Trial Comparing Peroneal
ments and dropped foot gait deviations among stroke survivors. Clin Biomech Nerve Functional Electrical Stimulation to an Ankle Foot Orthosis for Patients
(Bristol Avon) 2013;28:1049–54. With Chronic Stroke. Neurorehabil Neural Repair 2015;29:911–22.
[2] Stewart JD. Foot drop: where, why and what to do? Pract Neurol [30] Bethoux F, Rogers HL, Nolan KJ, Abrams GM, Annaswamy TM, Brandstater M,
2008;8:158–69. et al. The effects of peroneal nerve functional electrical stimulation versus
[3] Pittock SJ, Moore AP, Hardiman O, Ehler E, Kovac M, Bojakowski J, et al. A ankle-foot orthosis in patients with chronic stroke: a randomized controlled
double-blind randomised placebo-controlled evaluation of three doses of trial. Neurorehabil Neural Repair 2014;28:688–97.
botulinum toxin type A (Dysport) in the treatment of spastic equinovarus [31] Salisbury L, Shiels J, Todd I, Dennis M. A feasibility study to investigate the
deformity after stroke. Cerebrovasc Dis 2003;15:289–300. clinical application of functional electrical stimulation (FES), for dropped foot,
[4] Duncan PW. Stroke disability. Phys Ther 1994;74:399–407. during the sub-acute phase of stroke - A randomized controlled trial. Physio-
[5] Nolan KJ, Yarossi M. Weight transfer analysis in adults with hemiplegia using ther Theory Pract 2013;29:31–40.
ankle foot orthosis. Prosthet Orthot Int 2011;35:45–53. [32] Morone G, Fusco A, Di Capua P, Coiro P, Pratesi L. Walking training with foot
[6] Tyson SF, Crow JL, Connell L, Winward C, Hillier S. Sensory impairments of the drop stimulator controlled by a tilt sensor to improve walking outcomes: a
lower limb after stroke: a pooled analysis of individual patient data. Top Stroke randomized controlled pilot study in patients with stroke in subacute phase.
Rehabil 2013;20:441–9. Stroke Res Treat 2012;2012:523564.
[7] Pereira S, Mehta S, McIntyre A, Lobo L, Teasell RW. Functional electrical [33] Burridge JH, Taylor PN, Hagan SA, Wood DE, Swain ID. The effects of common
stimulation for improving gait in persons with chronic stroke. Top Stroke peroneal stimulation on the effort and speed of walking: a randomized
Rehabil 2012;19:491–8. controlled trial with chronic hemiplegic patients. Clin Rehabil
[8] Hong Z, Sui M, Zhuang Z, Liu H, Zheng X, Cai C, et al. Effectiveness of 1997;11:201–10.
Neuromuscular Electrical Stimulation on Lower Limbs of Patients With Hemi- [34] Everaert DG, Stein RB, Abrams GM, Dromerick AW, Francisco GE, Hafner BJ,
plegia After Chronic Stroke: A Systematic Review. Arch Phys Med Rehabil et al. Effect of a foot-drop stimulator and ankle-foot orthosis on walking
2018;99. 1011-1022 e1011. performance after stroke: a multicenter randomized controlled trial. Neuro-
[9] Lin S, Sun Q, Wang H, Xie G. Influence of transcutaneous electrical nerve rehabil Neural Repair 2013;27:579–91.
stimulation on spasticity, balance, and walking speed in stroke patients: a [35] Kluding PM, Dunning K, O’Dell MW, Wu SS, Ginosian J, Feld J, et al. Foot drop
systematic review and meta-analysis. J Rehabil Med 2018;50:3–7. stimulation versus ankle foot orthosis after stroke: 30-week outcomes. Stroke
[10] Robbins SM, Houghton PE, Woodbury MG, Brown JL. The therapeutic effect of 2013;44:1660–9.
functional and transcutaneous electric stimulation on improving gait speed in [36] Parker MG, Tyson SF, Weightman AP, Abbott B, Emsley R, Mansell W. Perceptual
stroke patients: a meta-analysis. Arch Phys Med Rehabil 2006;87:853–9. control models of pursuit manual tracking demonstrate individual specificity
[11] Everaert DG, Thompson AK, Chong SL, Stein RB. Does functional electrical and parameter consistency. Atten Percept Psychophys 2017;79:2523–37.
stimulation for foot drop strengthen corticospinal connections? Neurorehabil [37] Sheffler LR, Taylor PN, Bailey SN, Gunzler DD, Buurke JH, IJzerman MJ, et al.
Neural Repair 2010;24:168–77. Surface peroneal nerve stimulation in lower limb hemiparesis: effect on
[12] Thompson AK, Lapallo B, Duffield M, Abel BM, Pomerantz F. Repetitive quantitative gait parameters. Am J Phys Med Rehabil 2015;94:341–57.
common peroneal nerve stimulation increases ankle dorsiflexor motor evoked [38] Mitsutake T, Sakamoto M, Horikawa E. The effects of electromyography-
potentials in incomplete spinal cord lesions. Exp Brain Res 2011;210:143–52. triggered neuromuscular electrical stimulation plus tilt sensor functional
[13] Kottink AI, Oostendorp LJ, Buurke JH, Nene AV, Hermens HJ, Ijzerman MJ. The electrical stimulation training on gait performance in patients with subacute
orthotic effect of functional electrical stimulation on the improvement of stroke: a randomized controlled pilot trial. Int J Rehabil Res 2019;42:358–64.
walking in stroke patients with a dropped foot: a systematic review. Artif [39] Knash ME, Kido A, Gorassini M, Chan KM, Stein RB. Electrical stimulation of the
Organs 2004;28:577–86. human common peroneal nerve elicits lasting facilitation of cortical motor-
[14] Stein RB, Everaert DG, Thompson AK, Chong SL, Whittaker M, Robertson J, et al. evoked potentials. Exp Brain Res 2003;153:366–77.
Long-term therapeutic and orthotic effects of a foot drop stimulator on [40] Harris-Love ML, Macko RF, Whitall J, Forrester LW. Improved hemiparetic
walking performance in progressive and nonprogressive neurological disor- muscle activation in treadmill versus overground walking. Neurorehabil
ders. Neurorehabil Neural Repair 2010;24:152–67. Neural Repair 2004;18:154–60.
[15] Taylor PN, Burridge JH, Dunkerley AL, Wood DE, Norton JA, Singleton C, et al. [41] Peurala SH, Tarkka IM, Pitkanen K, Sivenius J. The effectiveness of body weight-
Clinical use of the Odstock dropped foot stimulator: its effect on the speed and supported gait training and floor walking in patients with chronic stroke. Arch
effort of walking. Arch Phys Med Rehabil 1999;80:1577–83. Phys Med Rehabil 2005;86:1557–64.
[16] Kottink AI, Tenniglo MJ, de Vries WH, Hermens HJ, Buurke JH. Effects of an [42] Nolan KJ, Yarossi M, McLaughlin P. Changes in center of pressure displacement
implantable two-channel peroneal nerve stimulator versus conventional with the use of a foot drop stimulator in individuals with stroke. Clin Biomech
walking device on spatiotemporal parameters and kinematics of hemiparetic (Bristol Avon) 2015;30:755–61.
gait. J Rehabil Med 2012;44:51–7. [43] Ring H, Treger I, Gruendlinger L, Hausdorff JM. Neuroprosthesis for footdrop
[17] Bae YH, Ko YJ, Chang WH, Lee JH, Lee KB, Park YJ, et al. Effects of Robot-assisted compared with an ankle-foot orthosis: effects on postural control during
Gait Training Combined with Functional Electrical Stimulation on Recovery of walking. J Stroke Cerebrovasc Dis 2009;18:41–7.
Locomotor Mobility in Chronic Stroke Patients: A Randomized Controlled [44] Schmid A, Duncan PW, Studenski S, Lai SM, Richards L, Perera S, et al.
Trial. J Phys Ther Sci 2014;26:1949–53. Improvements in speed-based gait classifications are meaningful. Stroke
[18] Hwang DY, Lee HJ, Lee GC, Lee SM. Treadmill training with tilt sensor 2007;38:2096–100.
functional electrical stimulation for improving balance, gait, and muscle [45] Studenski S, Perera S, Patel K, Rosano C, Faulkner K, Inzitari M, et al. Gait speed
architecture of tibialis anterior of survivors with chronic stroke: A randomized and survival in older adults. JAMA 2011;305:50–8.
controlled trial. Technol Health Care 2015;23:443–52. [46] Bohannon RW, Glenney SS. Minimal clinically important difference for change
[19] Dunning K, O’Dell MW, Kluding P, McBride K. Peroneal Stimulation for Foot Drop in comfortable gait speed of adults with pathology: a systematic review. J Eval
After Stroke: A Systematic Review. Am J Phys Med Rehabil 2015;94:649–64. Clin Pract 2014;20:295–300.
[20] Sharif F, Ghulam S, Malik AN, Saeed Q. Effectiveness of Functional Electrical [47] Fulk GD, Ludwig M, Dunning K, Golden S, Boyne P, West T. Estimating clinically
Stimulation (FES) versus Conventional Electrical Stimulation in Gait Rehabili- important change in gait speed in people with stroke undergoing outpatient
tation of Patients with Stroke. J Coll Physicians Surg Pak 2017;27:703–6. rehabilitation. J Neurol Phys Ther 2011;35:82–9.
[21] Park SJ, Wang JS. The immediate effect of FES and TENS on gait parameters in [48] Gok H, Kucukdeveci A, Altinkaynak H, Yavuzer G, Ergin S. Effects of ankle-foot
patients after stroke. J Phys Ther Sci 2017;29:2212–4. orthoses on hemiparetic gait. Clin Rehabil 2003;17:137–9.
[22] Kottink AI, Hermens HJ, Nene AV, Tenniglo MJ, van der Aa HE, Buschman HP, [49] Tyson SF, Sadeghi-Demneh E, Nester CJ. A systematic review and meta-
et al. A randomized controlled trial of an implantable 2-channel peroneal analysis of the effect of an ankle-foot orthosis on gait biomechanics after
nerve stimulator on walking speed and activity in poststroke hemiplegia. Arch stroke. Clin Rehabil 2013;27:879–91.
Phys Med Rehabil 2007;88:971–8. [50] Alam M, Choudhury IA, Bin Mamat A. Mechanism and design analysis of
[23] Moher D, Liberati A, Tetzlaff J, Altman DG, Group P. Preferred reporting items articulated ankle foot orthoses for drop-foot. ScientificWorld J
for systematic reviews and meta-analyses: the PRISMA statement. Int J Surg 2014;2014:867869.
2010;8:336–41. [51] Ng MM, Hill KD, Batchelor F, Burton E. Factors Predicting Falls and Mobility
[24] Maher CG, Sherrington C, Herbert RD, Moseley AM, Elkins M. Reliability of the Outcomes in Patients With Stroke Returning Home After Rehabilitation Who
PEDro scale for rating quality of randomized controlled trials. Phys Ther Are at Risk of Falling. Arch Phys Med Rehabil 2017;98:2433–41.
2003;83:713–21. [52] Batchelor FA, Mackintosh SF, Said CM, Hill KD. Falls after stroke. Int J Stroke
[25] Olivo SA, Macedo LG, Gadotti IC, Fuentes J, Stanton T, Magee DJ. Scales to assess 2012;7:482–90.
the quality of randomized controlled trials: a systematic review. Phys Ther [53] Berg KO, Wood-Dauphinee SL, Williams JI, Maki B. Measuring balance in the
2008;88:156–75. elderly: validation of an instrument. Can J Public Health 1992;83(Suppl 2):S7–11.
[26] Balshem H, Helfand M, Schunemann HJ, Oxman AD, Kunz R, Brozek J, et al. [54] Hiengkaew V, Jitaree K, Chaiyawat P. Minimal detectable changes of the Berg
GRADE guidelines: 3. Rating the quality of evidence. J Clin Epidemiol Balance Scale, Fugl-Meyer Assessment Scale, Timed ‘‘Up & Go’’ Test, gait
2011;64:401–6. speeds, and 2-minute walk test in individuals with chronic stroke with
[27] Higgins J, Green S. Cochrane handbook for systematic reviews of interventions, different degrees of ankle plantarflexor tone. Arch Phys Med Rehabil
5 ed., 2011. 2012;93:1201–8.