ORIGINAL ARTICLE

A randomized controlled trial of self-regulated modified constraint-

induced movement therapy in sub-acute stroke patients
K. P. Y. Liua, K. Balderib, T. L. F. Leungc, A. S. Y. Yuec, N. C. W. Lamd, J. T. Y. Cheungc, S. S. M. Fonge, C.
M. W. Sumc, M. Bissettf, R. Ryea and V. C. T. Mokg

a
Western Sydney University, Penrith, NSW; bConcord Hopsital, Concord, NSW, Australia; cHong Kong Hospital Authority, Hong Kong,
China; dRecovre, Sydney, NSW, Australia; eThe University of Hong Kong, Hong Kong, China; fGriffith University, Gold Coast, Qld,
Australia; and gThe Chinese University of Hong Kong, Hong Kong, China

EUROPEAN JOURNAL OF NEUROLOGY

Keywords:
functional recovery,
modified constraint-
induced movement
therapy, randomized
controlled trial,
self-regulation, stroke
Received 27 October 2015
Accepted 4 April 2016
European Journal of
Neurology 2016, 0: 1–10
doi:10.1111/ene.13037
Background and purpose: Emerging research suggests the use of self-regulation
(SR) for improving functional regain in patients post stroke. SR is proposed
to produce an added effect to effective modified constraint-induced movement
therapy (mCIMT). This study aimed to examine the effect of a self-regulated
mCIMT programme (SR-mCIMT) for functional regain in patients with sub-
acute stroke.
Methods: Eighty-six patients completed the trial: SR-mCIMT, n = 29;
mCIMT, n = 31; or conventional functional rehabilitation, n = 26. All inter-
ventions were 2-week therapist-guided training. Outcome measurements, taken
by a blinded assessor, examined arm function [Action Research Arm Test
(ARAT), Fugl-Meyer Assessment (FMA)], daily task performance [Lawton
Instrumental Activities of Daily Living Scale (Lawton IADL)] and self-per-
ceived arm use in functional tasks [Motor Activity Log (MAL)].
Results: Significant differences were found with the SR-mCIMT outperform-
ing the other groups after the intervention (ARAT, P = 0.006; FMA, Lawton
IADL and MAL, all Ps < 0.001). In terms of the carry-over effect, the
SR-mCIMT group outperformed in the hand and coordination subscales of
ARAT and FMA (P = 0.012–0.013) and the self-perceived quality of arm use
(P = 0.002).
Conclusion: A combination of SR and mCIMT could produce an added effect
in functional regain in patients post stroke.

through trial sequential analysis also showed that

Introduction
A recent systematic review demonstrated that con-
straint-induced movement therapy (CIMT) produces a
limited improvement in arm motor function and no
significant improvement for use of the arm in func-
tional tasks than conventional treatment in people
with stroke [1]. Transfer of training from the
improved arm motor function to reducing its effect on
disability was not observed in the studies included in
the review. Another review investigating the evidence
Correspondence: K. P. Y. Liu, Western Sydney University, Locked
Bag 1797, Penrith, 2751 NSW, Australia (tel.: +61 2 4620 3432;
fax: +61 2 4620 3792;
e-mail: karen.liu@westernsydney.edu.au).
results on arm motor function were inconclusive [2].
Constraint-induced movement therapy can be con-
sidered a multifaceted intervention. The impact on
arm impairment and motor function may be due not
solely to the constraint but also to the type and
amount of exercise [1]. Typically, CIMT programmes
are therapist-directed, whereby patients follow thera-
pists’ guidance when they practise CIMT. Promoting
transfer of training is important to ensure the treat-
ment effect is applicable to real life [3]. Kwakkel et al.
[4] commented that behavioural strategies which pro-
mote transfer of learning are mostly absent in CIMT
protocols. Taub et al. [3] and Takebayashi et al. [5]
adopted a behavioural analysis approach that pro-
moted transfer of skills learnt in CIMT and found

© 2016 EAN 1
2 K. P. Y. LIU ET AL.
increased treatment effect on arm function and self-
perceived amount of arm use. Similarly, our recent
review [6] also suggested that learning and generaliza-
tion in CIMT could be enhanced if an active learning
component, to increase patient-directed self-awareness
through the use of self-regulation (SR), was added.
Self-regulation, which stems from social cognitive
theory, describes reflective learning and problem solv-
ing by bringing problems and solutions to a conscious
level [7]. Based on the concept of SR, an effective SR
programme in improving daily task performance for
people with stroke was developed [8,9]. The SR strat-
egy aimed to give patients greater responsibility in
their own rehabilitation with a focus on improving
self-awareness and facilitating acknowledgement of
individuals’ functional problems, and thus promoting
learning and transfer of training.
This study aimed to investigate the effect of com-
bining SR and CIMT for enhancing functional arm
regain, daily task performance and self-perceived arm
use in functional tasks in patients with sub-acute
stroke. The modified CIMT (mCIMT) was adopted.
Methods
Participants and randomization
Patients receiving inpatient rehabilitation post stroke
were recruited from Shatin Hospital and Tuen Mun
Hospital in Hong Kong. Patients were eligible if they
(i) had experienced a first ischaemic type stroke with
lesion in the primary motor cortex as confirmed by a
computed tomography scan, (ii) had a stroke onset of
less than 3 months, (iii) were aged above 60 and (iv)
had at least 10° of active metacarpophalangeal joint
and interphalangeal joint extension, 20° of active wrist
extension. Patients were excluded if they (i) had exces-
sive spasticity in the affected limb with a score of 2 or
more on the Modified Ashworth Scale, (ii) had exces-
sive pain in the affected limb with a score 4 or more
using a Visual Analog Scale, (iii) had cognitive
impairment with a score below 19 on the Mini-Mental
State Examination (MMSE) and (iv) had been diag-
nosed with depression according to Diagnostic and
Statistical Manual of Mental Disorders IV criteria.
Eligible patients were randomized by drawing lots
into one of the three intervention groups. The random-
ization was conducted by a blinded researcher who was
not involved in the interventions or study measures.
Intervention
The interventions were 10 1-h therapist-guided indi-
vidual training sessions (daily on weekdays, total
2 weeks) administered by three occupational therapists
who had over 3 years of experience working in stroke
rehabilitation one for each group, SR-mCIMT,
mCIMT and the control intervention.
Patients learnt a total of 10 daily tasks (five per
week) [10,11]. The tasks were graded over the 2-week
period and included brushing teeth, dressing (both
upper and lower body), folding laundry, putting
clothes on a hanger in week 1; and using a telephone,
preparing a cup of tea, sweeping the floor, washing
towels, washing dishes in week 2. All groups were
trained to practise the assigned tasks and received the
same amount of task practice. All tasks were learnt in
the clinic room of the hospital.
The control intervention used conventional func-
tional rehabilitation which involved the therapist
demonstrating the adapted task performance followed
by the patient’s supervised practice using both arms
[12]. Patients in the mCIMT group entered into the
protocol as described by Page and colleagues [13].
Their non-hemiplegic arm was restrained in a mitt for
4 h every day (including the 1-h therapist-guided
training) monitored by the nursing staff in the hospi-
tal. As per the control group, the therapist provided
demonstration on the adapted task performance but
with one arm (the side of the patient’s hemiplegic
arm), and patients practised the tasks with the unre-
strained hemiplegic arm under the therapist’s supervi-
sion. Performance feedback and strategies for
improvement were given by the therapists in both the
control and mCIMT groups. In the SR-mCIMT inter-
vention, patients entered into the same mCIMT proto-
col as in the mCIMT group. Instead of the
demonstration-and-practice protocol, patients were
taught using the SR strategy adopted from Liu et al.
[8,9]. This involved patients self-reflecting on their
abilities and deficits in performing the tasks, identify-
ing problems and solutions in achieving the most
independence in the tasks, and practising the adapted
tasks. Feedback on their self-reflected abilities and
self-identified problems and solutions was provided
with the assistance of a video playback of their task
performance. The therapist provided guidance
throughout the process.
Besides the 1-h intervention, each patient received 1
h of daily physical therapy addressing mobilization,
strengthening and walking. These therapists were
blinded to the group allocation.
Study measures
All patients were assessed using the MMSE, the
Barthel Index (BI) and the study measures before they
entered into the intervention (T0). They were assessed
© 2016 EAN

on the study measures after the intervention (T1) and
at 1 month after the intervention to assess the carry-
over effect (T2). Assessments for T0 and T1 were
completed in the inpatient clinic, a different room
from where training was provided. At T2, all patients
were assessed in the outpatient clinic. All study mea-
sures were conducted by a blinded assessor who was
not involved in the randomization or the delivery of
the interventions. Both the assessor and patients were
asked not to talk with each other about their treat-
ment in order to preserve the blinding.
The primary study measures were improvements in
arm function [the Action Research Arm Test
(ARAT)], the upper extremity motor subsection of the
Fugl-Meyer Assessment (FMA) and daily task perfor-
mance [the Lawton Instrumental Activities of Daily
Living Scale (Lawton IADL)]. The secondary study
measure examined patients’ self-perceived arm use in
functional tasks [the Motor Activity Log (MAL)].
The ARAT contains 19 items examining the proxi-
mal and distal arm function [14]. The upper extremity
motor subsection of the FMA involves the motor
function of the upper arm, wrist and hand, and coor-
dination [15]. The Lawton IADL examines the perfor-
mance of eight functional tasks such as housekeeping
and shopping [16]. The MAL is a patient-rated ques-
tionnaire where patients rate the Quality of Movement
(QOM) and Amount of Use (AOU) of the affected
arm when completing 28 tasks [17]. All of these mea-
sures have been identified as valid and reliable for use
in patients post stroke.
Statistical analyses
Using PASS 13 software, with reference to the results
in task performance as reflected by the Functional
Independence Measure motor score obtained from the
previous SR study [9], a sample of 24 in each group
would give a power of 80%, an alpha value of 0.05
and effect size of 0.50. An additional 20% was added
to account for possible attrition.
All analyses were conducted using the IBM Statisti-
cal Package for the Social Sciences, version 22. Differ-
ences between the three groups in demographic
variables were assessed using analysis of variance
(ANOVA) and the chi-squared test. Baseline measures of
the standardized tests were compared using the
Kruskal–Wallis test. The Shapiro–Wilk test was used
to evaluate the normality of the outcome valuables.
Since the normality could not be accepted, non-para-
metric tests were used. The per-protocol analysis
included all patients who had data available at T0, T1
and T2. Differences in the gains across the study mea-
sures, before and after intervention, between the three
© 2016 EAN
SELF-REGULATED CONSTRAINT-INDUCED THERAPY 3
groups were measured using the Kruskal–Wallis test.
The Mann–Whitney U test was used to further test
the differences, if any, between any of the two groups.
Rank analysis of variance tests with results at T1 as
the covariate were used to assess the carry-over effect
of the study measures (T2) for the three groups with
post hoc least significant difference tests to detect dif-
ferences between any two groups. Bonferroni correc-
tion was adopted with P ≤ 0.013 as significant for
multiple comparisons.
Ethical considerations
Ethics approval was obtained from the Hong Kong
Polytechnic University Human Ethics Committee, the
Joint Chinese University of Hong Kong – New Terri-
tories East Cluster Clinical Research Ethics Commit-
tee and the New Territories West Cluster Clinical and
Research Ethics Committee. Written informed consent
was obtained from all patients.
Trial registration
The study was registered with the Clinical Trial Regis-
tration (ClinicalTrials.gov Identifier NCT02480140).
Results
Participants
Of the 86 patients who completed the trial with all T0,
T1 and T2 data available (SR-mCIMT, n = 29; mCIMT,
n = 31; control, n = 26), no significant difference in
demographic variables was found between the three
groups or between any two groups (P = 0.218–0.982) at
T0 (CONSORT diagram, Fig. 1; Table 1). Patient BI
scores at T0 indicated that all groups required moderate
assistance in self-care tasks. No significant differences in
the study measures at baseline were found between the
three groups (Table 1) or between any two groups
(P = 0.063–0.949).
Gains after intervention
Significant differences in the total scores of the
ARAT, FMA, Lawton IADL and MAL were found
between the three groups (P = 0.006 to P < 0.001).
The differences in subscales of the study measures are
reported in Table 1.
For the differences between any of the two groups,
the SR-mCIMT group showed significantly better per-
formance in total score (P = 0.006) and ‘pinch’ subscale
(P < 0.001) than the control group in the ARAT
(Fig. 2). The mCIMT group had significantly better
4 K. P. Y. LIU ET AL.

Assessed for eligibility

(n = 110)

Excluded
- Not meeting inclusion
criteria (n = 20)
Enrolment

Randomized (n = 90)

Allocated to SR- Allocated to Allocated to control

mCIMT group (n = mCIMT group (n = group (n = 27)
31) 32)
Allocation

Discontinued Discontinued Discontinued

intervention (n = 2) intervention (n = 1) intervention (n = 1)
(Both had T0 data only: (Had T0 data only: (Had T0 data only:
One experienced fever Patient requested to be Patient had fever
and one had discharged from in- during week one of the
uncontrolled high blood patient rehabilitation at intervention and was
pressure during week week two of the consequently removed
Follow up

one of the intervention intervention) from the study)

and were consequently
dropped from the study)

Analysed (n = 26)
Analysis

Analysed (n = 29) Analysed (n = 31)

Figure 1 CONSORT diagram.

performance in the total score (P = 0.007) and ‘grip’
and ‘pinch’ subscales (P = 0.008 and 0.002 respectively)
than the control group. No significant difference was
detected between the SR-mCIMT and the mCIMT
groups. In the FMA, the SR-mCIMT group showed
significantly better performance in total score
(P < 0.001) and all subscales (P < 0.001 to P = 0.007)
than the control group (Fig. 3). The mCIMT group
had significantly better performance than the control
group in the total score (P = 0.002) but not in any of
the subscales. Comparing the SR-mCIMT and the
mCIMT groups, the SR-mCIMT group had signifi-
cantly better performance in the total score (P = 0.011)
and ‘upper limb coordination’ subscale (P < 0.001).
Significant differences were found in the Lawton
IADL between the SR-mCIMT and control groups
(P < 0.001), and the SR-mCIMT and mCIMT groups
(P = 0.001), but not between the control and mCIMT
groups (P = 0.361) (Fig. 4). In the MAL, significant
differences were found in both the AOU and QOM
between the SR-mCIMT and control groups and
between the control and mCIMT groups (all
P < 0.001) (Fig. 5). A significant difference was also
found in the AOU between the SR-mCIMT and
mCIMT groups (P < 0.001) but not in QOM between
these two groups (P = 0.699).
Carry-over effects
In the ARAT, a significant difference in ‘grip’ subscale
was found between the three groups (P = 0.002)
(Table 1). Post hoc tests revealed significant differ-
ences in the ‘pinch’ subscale between the SR-mCIMT
and mCIMT groups (P = 0.012), and in ‘grip’ and
‘gross movement’ subscales between the SR-mCIMT
and control groups (P = 0.001 and 0.012 respectively)
(Fig. 2). The SR-mCIMT group was found to have
better carry-over effects in these measures. In the
FMA, significant differences in ‘hand motor function’
and ‘upper limb coordination’ subscales were found
between the three groups (P = 0.008 and 0.009 respec-
tively) (Table 1). Post hoc tests revealed significant
differences in the ‘hand motor function’ and ‘upper
limb coordination’ subscales between the SR-mCIMT

© 2016 EAN
© 2016 EAN
Table 1 Baseline characteristics and results of study measures at baseline, post intervention and 1-month follow-up

Control group (n = 26) mCIMT group (n = 31) SR-mCIMT group (n = 29) P values

Gains Carrying-
after over
T0 T1 T2 T0 T1 T2 T0 T1 T2 At baseline intervention effect

Age (years), 67.87 (9.42) 65.07 (6.70) 66.80 (6.11) 0.713

mean (SD)a
Mena 13 (50%) 23 (74.2%) 14 (48.3%) 0.431
Left hemiplegiaa 17 (65.4%) 16 (51.6%) 16 (55.2%) 0.114
Days since stroke, 9.69 (1.00) 8.44 (0.62) 9.15 (1.26) 0.732
mean (SD)a
MMSE, mean (SD)a 26.83 (2.12) 26.19 (2.01) 25.88 (1.76) 0.975
BI, mean (SD)a 58.17 (26.93) 54.44 (15.65) 58.16 (13.84) 0.823
ARAT, mean (SD)
Grasp 10.62 (0–18) 13.88 (6–18) 14.00 (6–18) 11.29 (0–18) 17.81 (17–18) 17.94 (17–18) 1.10 (0–18) 16.00 (5–18) 16.72 (5–18) 0.847 0.459 0.743
Grip 6.31 (2–12) 7.49 (6–12) 7.44 (4–12) 4.57 (0–12) 10.68 (8–12) 10.81 (8–12) 7.28 (0–12) 10.76 (6–12) 11.34 (6–12) 0.962 0.030 0.002*
Pinch 9.92 (0–18) 11.88 (0–18) 11.54 (0–18) 7.10 (0–18) 13.03 (1–18) 12.65 (0–18) 7.34 (0–18) 16.39 (13–18) 17.10 (15–18) 0.738 <0.001* 0.048
Gross movement 7.77 (3–18) 8.00 (5–9) 7.31 (3–9) 7.65 (3–9) 8.16 (6–9) 8.00 (6–9) 5.83 (0–9) 7.93 (3–9) 8.28 (4–9) 0.611 0.089 0.039
Total 34.62 (7–66) 42.31 (18–57) 41.38 (18–57) 32.19 (4–57) 49.68 (32–57) 49.39 (32–57) 30.55 (0–55) 51.66 (32–57) 53.45 (35–57) 0.826 0.006* 0.097
FMA, mean (SD)
Upper arm motor 23.15 (17–29) 24.31 (19–29) 24.85 (19–30) 24.32 (15–28) 26.35 (20–30) 27.10 (19–30) 22.66 (7–29) 26.14 (14–33) 26.76 (15–33) 0.489 <0.001* 0.481
function
Wrist motor 5.46 (1–9) 6.15 (2–9) 5.54 (2–10) 5.39 (1–8) 6.58 (3–9) 6.65 (4–9) 4.21 (0–9) 6.03 (2–10) 6.55 (3–10) 0.201 0.040 0.019
function
Hand motor 9.62 (6–13) 10.54 (7–14) 11.92 (7–14) 9.68 (2–14) 11.45 (4–14) 11.71 (4–14) 9.21 (4–14) 11.66 (8–14) 12.48 (8–14) 0.563 0.002* 0.008*
function
Upper limb 2.00 (0–5) 2.58 (0–5) 2.81 (0–6) 2.77 (0–4) 4.00 (3–6) 3.68 (1–6) 2.41 (0–5) 4.93 (3–6) 5.31 (3–6) 0.732 <0.001* 0.009*
coordination
Total 40.23 (24–51) 43.58 (28–54) 45.12 (28–60) 42.16 (24–53) 48.39 (35–58) 49.13 (35–59) 38.48 (15–54) 48.76 (29–62) 51.10 (33–62) 0.461 <0.001* 0.424
IADL, mean (SD) 2.85 (0–8) 4.54 (0–6) 4.31 (2–7) 3.23 (0–7) 5.61 (0–7) 5.29 (0–8) 2.17 (0–7) 6.76 (1–8) 6.83 (1–8) 0.202 <0.001* 0.514
MAL, mean (SD)
AOU 30.88 (0–88) 36.38 (5–89) 47.69 (5–88) 31.14 (0–98) 51.83 (6–105) 74.41 (24–139) 22.39 (1–49) 80.31 (28–135) 106.46 (54–140) 0.605 <0.001* 0.019
QOM 42.42 (7–90) 57.58 (14–120) 63.19 (21–119) 46.97 (0–107) 89.14 (14–133) 100.69 (21–119) 46.64 (0–108) 93.18 (11–135) 110.68 (45–140) 0.831 <0.001* 0.002*

mCIMT, modified constraint-induced movement therapy; SR-mCIMT, self-regulated modified constraint-induced movement therapy; T0, baseline; T1, post intervention; T2, 1-month follow-up;
MMSE, Mini-Mental State Examination; BI, Barthel Index; ARAT, Action Research Arm Test; FMA, Fugl-Meyer Assessment; IADL, Lawton Instrumental Activities of Daily Living Scale;
MAL, Motor Activity Log; AOU, Amount of Use; QOM, Quality of Movement. aCharacteristics taken at baseline T0; *P ≤ 0.013.
SELF-REGULATED CONSTRAINT-INDUCED THERAPY
5
6 K. P. Y. LIU ET AL.

* * Gains after
Control and SR-mCIMT Control and SR-mCIMT intervention
* * *
Control and mCIMT Control and mCIMT Control and mCIMT

* *
Control and SR-mCIMT Control and SR-mCIMT Carrying-
* over effect
mCIMT and SR-mCIMT

50

40

30

20

10

0
Control

Control

Control

Control

Control
mCIMT
mCIMT

mCIMT
mCIMT

mCIMT
mCIMT

mCIMT
mCIMT

mCIMT
mCIMT

Grasp Grip Pinch Gross movement Total

T0 (Baseline)

T1 (Post-intervention)

T2 (One-month follow-up)
Figure 2 Between-group comparisons of the Action Research Arm Test (median scores).

and mCIMT groups (P = 0.013 and 0.012 respec- and control group (P = 0.004). As with the ARAT,
tively) and in the ‘wrist motor function’ and ‘upper the SR-mCIMT group was found to have better
limb coordination’ subscales between the SR-mCIMT carry-over effects in these measures.
and control groups (P = 0.006 and 0.014 respectively) For the MAL, a significant difference was found in
(Fig. 3). A significant difference was also found in the QOM between the three groups (P = 0.002) (Table 1).
‘hand motor function’ subscale between the mCIMT Post hoc tests showed a significantly better carrying-

© 2016 EAN
 SELF-REGULATED CONSTRAINT-INDUCED THERAPY 7

* * * * * Gains after
Control and SR-mCIMT Control and SR-mCIMT Control and SR-mCIMT Control and SR-mCIMT Control and SR-mCIMT intervention
* *
mCIMT and SR-mCIMT mCIMT and SR-mCIMT

*
Control and mCIMT
* * Carrying-
Control and SR-mCIMT Control and SR-mCIMT over effect
* *
mCIMT and SR-mCIMT mCIMT and SR-mCIMT

50

40

30

20

10

0
Control

SR-mCIMT

Control

SR-mCIMT

Control

SR-mCIMT

Control

SR-mCIMT

Control

SR-mCIMT
mCIMT

mCIMT

mCIMT

mCIMT

mCIMT

Upper arm Wrist Hand Coordination Total

motor function motor function motor function
T0 (Baseline)

T1 (Post-intervention)

T2 (One-month follow-up)
Figure 3 Between-group comparisons of the Fugl-Meyer Assessment (median scores).

© 2016 EAN
8 K. P. Y. LIU ET AL.
* Gains after
Control and SR-mCIMT
0
Control
T0 (Baseline)
T1 (Post-intervention)
T2 (One-month follow-up)
Figure 4 Between-group comparisons of the Lawton Instrumental Activities of Daily Living Scale (median scores).
over effect on AOU and QOM for the SR-mCIMT
group compared with the control group (P = 0.007
and P < 0.001 respectively) (Fig. 4). There was no sta-
tistical difference in the Lawton IADL between the
three groups or between any of the two groups
(Table 1 and Fig. 5).
Discussion
Consistent with previous studies, significant differences
were found in most of the subscales in the ARAT and
FMA between the SR-mCIMT or mCIMT with the
control groups [1]. The combination of SR with
mCIMT was found to enhance patients’ arm function
specifically in coordination, daily task performance
and self-perceived amount of arm use in functional
tasks after the intervention better than the mCIMT
intervention
*
mCIMT and SR-mCIMT
mCIMT SR-mCIMT
alone group. The effect of SR also extended to
1 month after the intervention for hand function,
pinch and upper limb coordination. This concurs with
our hypothesis on the benefits of SR in promoting
relearning and transfer of training in patients post
stroke [8,9] and providing an added effect to mCIMT
which further enhances daily task performance and
arm function [6].
Looking closely at arm function, the effect of SR
was more evident on coordination and finer use of
hand like the ‘pinch’ subscale in the ARAT and ‘hand
motor function’ subscale in the FMA. Fine motor
function and coordination require patients’ monitor-
ing and correction and involve cortical control [18].
The use of SR could possibly enhance patients’ moni-
toring by self-reflecting on the successful use of hand
skills and coordination.
© 2016 EAN
 SELF-REGULATED CONSTRAINT-INDUCED THERAPY 9

* *
Gains after
Control and SR-mCIMT Control and SR-mCIMT intervention
*
mCIMT and SR-mCIMT
* *
Control and mCIMT Control and mCIMT

* * Carrying-
over effect
Control and SR-mCIMT Control and SR-mCIMT

0
Control mCIMT SR-mCIMT Control mCIMT SR-mCIMT

Amount of use Quality of movement

T0 (Baseline)

T1 (Post-intervention)

T2 (One-month follow-up)
Figure 5 Between-group comparisons of the Motor Activity Log (median scores).

The value of adding SR to intervention was also
shown by both the objective performance-based and
subjective self-perceived arm use in task performance.
Significantly better performance in the Lawton IADL
was found between the SR-mCIMT and mCIMT
groups, and the SR-mCIMT and control groups. How-
ever, when comparing the mCIMT and the control
groups, the mCIMT patients did not outperform the
control group. mCIMT did not seem to provide the
same benefit as the SR-mCIMT. It is postulated that

© 2016 EAN
10 K. P. Y. LIU ET AL.
this benefit could be due to the enhanced transfer of
training provided by SR from arm function to daily
task performance that could possibly be missing in the
mCIMT group [3,5]. In the self-perceived arm use, the
SR-mCIMT group had a significantly better amount of
use of the affected arm in functional tasks than the
mCIMT group. The SR-mCIMT patients might have
engaged in activities by having more self-regulated use
of the affected arm in functional tasks. The SR-mCIMT
group showed significant carry-over differences in the
quality and amount of arm use compared with the con-
trol group. A carry-over effect was not found in the
mCIMT alone group. This further substantiates the
benefit of adding SR to the mCIMT intervention.
Self-regulation involves self-checking, self-reflection,
comparing and evaluating outcomes [6,8,9]. It is
hypothesized that SR provides a top-down approach
by giving control over patients’ own recovery and
mCIMT offers a bottom-up effect to enhance their
affected upper limb function. In combination, they
play an important role in promoting arm motor func-
tion and using the arm in functional tasks [6,8,9]. This
study was limited by the comparatively short follow-
up time. A longer follow-up time, say 6 months,
would be beneficial to show the clinical implication
for the person after discharge from the rehabilitation
programme. Further studies are recommended to
review the longer carry-over effect.
Acknowledgements
The authors would like to thank the staff and patients
in Shatin Hospital and Pok Oi Hospital, Hong Kong,
for their support. This study was supported by Hong
Kong Polytechnic University, Departmental Research
Grant, and Western Sydney University, School of
Science and Health Research Grant awarded to
K.P.Y. Liu.
Disclosure of conflicts of interest
The authors declare no financial or other conflict of
interest.
References
1. Corbetta D, Sirtori V, Castellini G, Moja L, Gatti R.
Constraint-induced movement therapy for upper extrem-
ities in people with stroke. Cochrane Database Syst Rev
2015; 10: Art. No.: CD004433.
2. Castellini G, Gianola S, Banzi R, et al. Constraint-
induced movement therapy: trial sequential analysis
applied to Cochrane collaboration systematic review
results. Trials 2014; 15: 512.
3. Taub E, Uswatte G, Mark VW, et al. Method for
enhancing real-world use of a more affected arm in
chronic stroke: transfer package of constraint-induced
movement therapy. Stroke 2013; 44: 1383–1388.
4. Kwakkel G, Veerbeek JM, van Wegen EEH, Wolf SL.
Constraint-induced movement therapy after stroke. Lan-
cet Neurol 2015; 14: 224–234.
5. Takebayashi T, Koyama R, Amanso S, et al. A 6-
month follow-up after constraint-induced movement
therapy with and without transfer package for patients
with hemiparesis after stroke: a pilot quasi-randomized
controlled trial. Clin Rehabil 2013; 27: 418–426.
6. Leung DPK, Liu KPY. Review of self-awareness and its
clinical application in stroke rehabilitation. Int J Rehabil
Res 2011; 34: 187–195.
7. Dinsmore DL, Alexander PA, Loughlin SM. Focusing
the conceptual lens on metacognition, self-regulation,
and self-regulated learning. Educ Psychol Rev 2008; 20:
391–409.
8. Liu KPY, Chan CCH, Lee TMC, Li LSW, Hui-Chan
CWY. Case reports on self-regulatory learning and gen-
eralization for people with brain injury. Brain Inj 2002;
16: 817–824.
9. Liu KPY, Chan CCH. A pilot randomized controlled
trial of self-regulation in promoting function in acute
post-stroke patients. Arch Phys Med Rehabil 2014; 95:
1262–1267.
10. Liu KPY, Chan CCH, Wong RSM, et al. A randomized
controlled trial of mental imagery augment generaliza-
tion of learning in acute poststroke patients. Stroke
2009; 40: 2222–2225.
11. Liu KPY, Chan CCH, Lee TMC, Hui-Chan CWY.
Mental imagery for promoting relearning for people
after stroke: a randomized controlled trial. Arch Phys
Med Rehabil 2004; 85: 1403–1408.
12. James AB. Restoring the role of independent person. In:
Radomski MV, Trombly Latham CA, eds. Occupational
Therapy for Physical Dysfunction, 6th edn. Philadelphia,
PA: Lippincott Williams & Wilkins, 2008: 774–816.
13. Page SJ, Sisto SA, Levine P, Johnston MV, Hughes M.
Modified constraint induced therapy: a randomized fea-
sibility and efficacy study. J Rehabil Res Dev 2001; 38:
583–590.
14. Lyle RC. A performance test for assessment of upper
limb function in physical rehabilitation treatment and
research. Int J Rehabil Res 1981; 4: 483–492.
15. Fugl-Meyer AR, Jaasko L, Leyman I, Olsson S, Steglind
S. The post-stroke hemiplegia patient. 1. A method for
evaluation of physical performance. Scand J Rehabil
Med 1975; 7: 13–31.
16. Lawton MP, Brody EM. Assessment of older people:
self-maintaining and instrumental activities of daily liv-
ing. Gerontologist 1969; 9: 179–186.
17. Uswatte G, Taub E, Morris D, Light K, Thompson PA.
The Motor Activity Log-28: assessing daily use of the
hemiparetic arm after stroke. Neurology 2006; 67: 1189–
1194.
18. Archambault PS, Ferrari-Toniolo S, Caminiti R, Batta-
glia-Mayer A. Visually-guided correction of hand reach-
ing movements: the neurophysiological bases in the
cerebral cortex. Vision Res 2015; 110: 244–256.
© 2016 EAN