Secondary Logo

Journal Logo

Modified Constraint-Induced Movement Therapy as a Home-Based Intervention for Children With Cerebral Palsy

Psychouli, Pavlina OT, MSc, PhD; Kennedy, Colin R. MD, FRCP, FRCPC

doi: 10.1097/PEP.0000000000000227

Purpose: This study was designed to investigate the benefit to upper limb function of a home-based version of pediatric constraint-induced movement therapy, which was delivered across 2 months.

Methods: Nine children (mean age: 6 years, 9 months) with hemiplegic cerebral palsy participated in this A1-B-C-A2 design, where A1 and A2 were nonintervention phases. In phases B and C, participants wore a splint on the unaffected hand. In phase C, motivating feedback through a computer game was added.

Results: The Melbourne Assessment of Unilateral Upper Limb Function and the Quality of Upper Extremity Skills Test scores were significantly higher at the end of phases B (P = .037 and P = .006, respectively) and C (P = .001 and P = .001, respectively). Melbourne scores remained higher at the end of phase A2 (P = .001).

Conclusions: A nonintensive form of home-based constraint-induced movement therapy was found to be effective. Improvements were larger after the second month of intervention.

The benefit to upper limb function is examined from a home-based version of pediatric constraint-induced movement therapy delivered in 2 months.

School of Occupational Therapy (Dr Psychouli), Technological Educational Institute of Athens, Greece; School of Medicine (Dr Kennedy), University of Southampton, Southampton, United Kingdom.

Correspondence: Pavlina Psychouli, OT, MSc, PhD, 45 Skamvonidon Str, 11853 Athens, Greece (

Grant Support: This study was supported by grants from the Greek State Scholarships Foundation (IKY) and the British Federation of Women Graduates.

This article is based on the research work completed by Dr Pavlina Psychouli as part of her PhD degree at the University of Southampton in the UK.

The authors declare no conflicts of interest.

Back to Top | Article Outline


Constraint-induced movement therapy (CIMT), in which use of a neurologically impaired limb is encouraged by restraint of the contralateral unimpaired limb, grew out of research with nonhuman primates in which somatosensory deafferentation of the dorsal root of the spinal nerve innervating one of the upper extremities had been performed. “Learned non-use,” a behavioral phenomenon linked to the clinical picture of hemiplegia in these primates, has been hypothesized to occur also in humans and is an important concept underpinning the rationale for the use of CIMT.

Constraint-induced movement therapy is a therapeutic intervention involving restraint of the uninvolved upper limb and intensive practice with the affected limb. The amount of practice and restraint time may not be appropriate for children and their families, who may have difficulty adhering fully to such a regimen. In a randomized controlled study1 of the use of a bivalved cast with 6 hours of exercise daily for a 3-week period, improvements were reported in the Quality of Upper Extremity Skills Test (QUEST). The authors encouraged future researchers to investigate whether a more child-friendly regimen with reduced treatment hours could also prove effective. Thus, researchers have attempted to develop modified, less intensive protocols of CIMT by reducing either the amount of practice or the daily restraint time or both.2–7

Home-based programs of CIMT may increase the adherence of children and their caregivers to the therapeutic program and may enhance family and child participation in activities that use the child's own toys and take place in their natural environment. Chen and colleagues8 reported better motor control changes induced by a home-based CIMT program compared with a dose-matched clinic-based traditional therapy, which included neurodevelopmental treatment techniques and unilateral and bilateral activity-oriented training. A randomized controlled trial9 investigated the same modified regimen of CIMT in 2 groups. The only differences arose from the randomization to either the home-based therapeutic environment, where the children had the opportunity to engage in real-life conditions using their own toys, or the clinic-based environment. The findings revealed greater improvements in the home-based group. Wallen et al3 did not find statistically significant differences between a modified type of constraint-induced therapy and intensive occupational therapy. The authors however provided very little information on the content of the therapeutic program, the activities used, and whether these were tailored to each child's needs. A similar design was used by Lin et al,4 who reported that modified CIMT was more beneficial than dose-matched classic occupational therapy. The CIMT program in this study was based on principles of shaping and relied on repetitive practice of activities of daily living skills when children were not in therapy sessions. Similarly, Al Oraibi and Eliasson9 reported superior outcomes compared with neurodevelopmental therapy not matched for intensity of the intervention with a modified home-based program of CIMT adhering to basic principles, such as individualization of treatment, positive feedback, parental education, and basic education of therapists providing the CIMT regimen. Research on home-based CIMT in children with cerebral palsy (CP) remains sparse and the topic needs further investigation.11

The limitations to the practicability of clinic-based intensive programs of CIMT in children with hemiplegia led us to develop a low-intensity home-based CIMT protocol.12 The aim of the present study was to examine the feasibility and the effectiveness of this protocol and to test whether a second intervention period would yield additional improvements.

Back to Top | Article Outline


Design and Setting

This was an uncontrolled clinical trial, An A1-B-C-A2 design was used, which is considered the most appropriate design for participants with motor disorders in childhood.13 The A1 and A2 periods were the baseline and follow-up, respectively, and the B and C periods were intervention phases. The screening process and all the assessments took place at the same clinical facility. The therapist responsible for the assessments was not involved in the treatment. The study received approval from the ethics committee.

Back to Top | Article Outline

Study Sample

Participants were recruited from 3 research sites via their pediatricians based at child development centers. All children continued with their routine “standard of care” therapy.

Inclusion criteria were (1) age 5 to 11 years; (2) diagnosis of congenital spastic hemiplegic CP; (3) performance above the 10th percentile on the Raven's Colored Progressive Matrices, a measure of nonverbal cognitive abilities that is relatively independent of motor skills;14 (4) commonly able to use the affected upper extremity for gross function in bilateral tasks, as determined by observation and from discussion with the child's occupational therapist and pediatrician; and (5) willing to take part in the study including both informed child assent and informed parental consent. Exclusion criteria were (1) fixed contractures that would limit the movement of the affected upper extremity by more than 1/4 of the normal range; and/or (2) behavioral problems that might interfere with the child's ability to comply with the protocol, as determined by the child's therapist.

Back to Top | Article Outline


The intervention took place at children's homes. Parents were given specific instructions to engage their children daily in some of the activities selected from the list they were given. Children were encouraged to wear the splint for as long as they were cooperative. The researcher and parents had frequent communication (once or twice within a week) to discuss any problems or concerns.

The 4 phases of the study (A1-B-C-A2) each had a 4-week duration. During both phases B and C, parents were instructed to apply a custom-made splint, covering the child's less affected hand and arm just below the elbow12 (Figure 1) at home for 2 hours a day, which could be divided into shorter sessions to increase adherence and practicality. While wearing the constraint, children were required to participate every day in some of the activities from the list, ensuring that daily living activities such as dressing were included, as well as play activities.

Fig. 1

Fig. 1

Children participated in functional activities during the intervention phases B and C. The activities were decided upon after consulting individually with the parents of each child to ensure that they would be interesting, motivating, and appropriate for each child. Parents listed all the everyday activities that their child would normally take part in, including dressing up, cleaning teeth, helping mother to prepare meals, etc, as well as turning pages to read a book, playing games with siblings, playing ball, and other activities in which the child preferred to engage during free time. Out of this list of activities, the researcher and parents decided upon those that could be practiced using only 1 hand. This process ended up with a list for each child that contained both everyday and play activities.

In phase C, constraint wearing and activity participation remained the same as in phase B but time playing a personal computer (PC) game was added. This game resembled “Pac-Man” and required unilateral manipulation of a joystick, movement of which was recorded as a measure of activity, while wearing the splint. The game lasted 20 minutes, which was chosen as an appropriate extra duration of exercising and feedback at the end of the day. At the end of the game, a colored bar was displayed on the screen, providing feedback by showing the child how much he/she had moved the affected hand, along with motivational cues to encourage the child to “keep trying.” This may be characterized as “augmented feedback”15 combined with extra practice for the upper limb.

Back to Top | Article Outline

Outcome Measures

Parents recorded on a daily log the total amount of time the splint was worn and the activities in which the children participated. The Melbourne Assessment of Unilateral Upper Limb Function (Melbourne) and the Quality of Upper Extremity Skills Test (QUEST) were used as the main measures of the effects of the interventions on the functional use of both upper limbs, especially the affected one. Both Melbourne and QUEST assessments were applied at the start and finish of the baseline period (A1) and at the end of phases B and C and the follow-up period (A2). There were thus 5 sets of scores, separated by 4 intervals of 1 month, referred to as T1, T2, T3, T4, and T5. Tests were administered and recorded on video for subsequent scoring by an assistant occupational therapist, blinded to the phase of the study to which the video applied. A detailed description of the design and timing of the assessments can be found in Table 1.



The Melbourne has been designed for children with CP, validated for ages 5 to 15 years, and provides a standardized means of scoring the ability to perform unilateral functional tasks with the affected limb. Sixteen items involving reach, grasp, release, and manipulation are scored on 3-, 4-, and 5-point scales and summed to provide a total score, which is converted to a percentage score.16 The QUEST has been specifically developed and validated for children with CP and is suitable for children 18 months or older.17 It includes items specifically related to hand function but also assesses movement of the adjacent joints. The test consists of 4 domains: dissociated movement, grasp, weight bearing, and protective extension scored on a yes/no scale and summed to provide a percentage score. The QUEST was applied according to the test manual to assess the function of both upper limbs.

Back to Top | Article Outline

Statistical Analysis

The mean score of assessments T1 and T2 was taken as the baseline score with which scores at assessments T3 and T4 (phase B and phase C, respectively) were compared. Persistence of benefit, 4 weeks after cessation of the intervention, was measured at assessment T5. Repeated measures analysis of variance (ANOVA) was used to test for the statistical significance of changes in Melbourne and QUEST scores between assessments at a level of P < 0.05 for 2-sided testing. Friedman and Wilcoxon's nonparametric tests were performed in addition to the parametric tests. Change in the PC game score over time was assessed for each participant during phase C. Content analysis was used for the daily logs.

Back to Top | Article Outline


Nine children (6 boys and 3 girls) with a mean age of 6 years and 9 months (range: 5 years, 1 month to 11 years) were included in the study. The frequency of their routine care therapy sessions and their baseline Melbourne and QUEST scores are shown in Table 2.



Analysis of the daily logs revealed that the splint was worn for 39 hours and 32 minutes on average over phase B, whereas during phase C the time increased slightly to reach 40 hours and 28 minutes. Only 1 child wore the splint for all 30 days during either phase. The other 8 children wore the splint over a range of 8 to 29 days. Children participated in similar activities during both phases B and C, except for the PC game, which was only available in phase C. In both phases B and C, the activities performed most commonly were brushing teeth/hair, eating finger food, getting dressed, and playing with toys or computer games. The game was played in phase C by 8 of the 9 children, the exception being child 5 who did not have access to a computer. During phase C, all the children gradually increased their scores on the PC game except for child 4, who used the game on only 9 days, fewer than any other participant.

All scores showed an approximately normal distribution in scatterplots, justifying the use of parametric repeated measures ANOVA tests. The corresponding (Friedman and Wilcoxon's) nonparametric tests showed statistically significant changes in all instances for which significance was found on ANOVA tests. Scores on the QUEST and the Melbourne Assessment were similar at the beginning and the end of the baseline period (Table 2), suggesting stable function for the group. Compared with baseline scores, the group mean scores on both the QUEST and Melbourne Assessment increased significantly after phases B (constraint + practice) and C (constraint + practice + motivating feedback). Changes in the group mean scores remained statistically significant at the end of the follow-up period (T5) only for the Melbourne Assessment (Table 3). Every participant individually had an increase in their individual outcome scores over the intervention periods.



On the Melbourne Assessment, 7 of the 9 participants reached the accepted threshold18 for a clinically important (12%) change in score between baseline and T4 and/or T5. Child 2 and child 4 were the most severely impaired, and child 3 the least impaired as determined by observation supported by the baseline scores of the Melbourne (42, 48, and 92, respectively) and QUEST (57, 53, and 89 respectively). These 3 children made small or moderate improvements.

Back to Top | Article Outline


Every individual participant receiving CIMT in this small study showed improvement on scores obtained on objective measures of upper limb function. The group mean increase in score was statistically significant and was also a clinically important change in most participants. An increase in the benefit to function during the second month of intervention was most clearly evident in the Melbourne Assessment, the primary outcome measure, and in 7 of the 9 participants remained a clinically important change 1 month after completing the intervention.

The QUEST data set was affected by the fact that 1 child did not cooperate for the final measurement, thus reducing the power to detect differences on that measure. The QUEST, which is used to assess the whole arm and range of movement, did not show persistent benefit 1 month after cessation of the intervention in contrast to the Melbourne Assessment, which is more focused on hand function and showed retention of improvements at the follow-up. The majority of the everyday activities and games that the children practiced at home predominantly required hand function, which is probably why, despite temporary improvements in large movements, a persistent effect was noticed only on hand skills.

All but 1 participant showed an increase in the scores (representing the amount of hand movement) obtained at the end of each game session. The computer game was the only additional activity during the second month of intervention (phase C), but the study does not allow us to distinguish between the effect on these scores resulting from a learning effect of continued CIMT over time and that resulting from an increase in motivation and feedback. A controlled study would be needed to test the potential increase of motivation when feedback games are included in a CIMT program.

Although consistency of the benefit of each intervention period (phase B and phase C) is apparent from the fact that it was observed in every participant, variability between participants with respect to functional outcome was observed, as reported in other studies.7,19 Our findings on the functional improvement of children with severe or minor impairments support the observation of Gordon et al20 that patients who are low or very high functioning do not benefit from CIMT as much as those functioning at an intermediate level. Other studies7,19 have, however, found that response to treatment was better in children with an initially lower ability. Eliasson et al7 suggested that these differences might be related to the outcome measures used and explained that the Assisting Hand Assessment, which was used in their study, measures a large range of skills, even very low abilities, making it possible to detect changes other than those related to grasping. Eliasson et al21 reported significant individual variation in improvements with no clear relationship to the hours of training, age, or children's mastery of behavior. Thus, some variation in the findings in published reports remains, highlighting the need for large multicenter trials to investigate which individual characteristics at baseline predict the greatest benefit from CIMT.

The fact that the hours of use were less than 60% of those specified in the protocol and that only 1 child wore the splint for all 30 days in any month of intervention highlights the difficulties in adherence to a home-based intervention with limited therapist participation. Similarly, large variability in time spent wearing the constraint was observed in the study of Klingels et al:19 an 80% compliance level of 40 hours was reported in only a small number of participants. Similar difficulties have been reported by other researchers who tested home-based CIMT protocols.7,10 All participants completed the program of Hsin et al,11 but the intervention was carried out by a therapist at children's homes, thus extra attention and support was given. In other studies,11,22 compliance was reported to be close to the goal as suggested by the 3.5 hours of daily constraint use. However, in these studies children only participated in practice sessions at 2 days per week, and therefore, to know whether the constraint time had been combined with functional use of the hemiplegic hand for the remaining days is not possible.

The Melbourne Assessment was also used by Gordon and her associates6 to assess the results of 2-hour daily therapy for 4 weeks. The lack of statistically significant improvements in that report could be due to a type II error attributable to the small sample size (n = 6) or due to the low functional status of the participants, as evident from the group mean Melbourne score of 52 at baseline. The functional gains found in the present study were larger than those observed in 3 other studies in which only moderate changes were found.5,19,23 The interventions tested were different in that the first of those studies5 used a group type of therapy for children, with activities not individually tailored for each child and none of the studies followed a home-based protocol. Klingels et al19 reported on 2 groups of which 1 received constraint without specific training. The other group followed a practice regimen that was distributed over 10 weeks, while training was taking place by a therapist over three 45-min weekly sessions. In the study of Sakzewski et al,23 variability of the functional status of the sample was large, which might account for the small changes in the Melbourne Assessment. Similar improvements in the QUEST score to those observed in the present study were reported by Choudhary et al.2 DeLuca et al1 used a cast as the constraint on the better functioning arm and an intensive regimen of 6 hours of CIMT daily for 21 days. Improvements on the QUEST were moderate compared with the present study, suggesting that a less intensive and more prolonged application of home-based CIMT comprising approximately 1.5 hours of daily constraint and training (that was the actual restraint time, according to the results) over 1 or 2 months might be as effective as an intensive intervention, confirming the findings of other researchers.7,4,11,22

Strengths of the present study include use of constraints that had been shown in a previous study to be the most acceptable and effective12 and the use of objective measures by an assessor blinded to whether the participant was in the baseline or intervention periods. The precision of the estimated treatment effect can be increased when the child acts as his/her own control.24 A nonintervention comparison period was provided by each participant and represented by the difference between the scores at the commencement and completion of the 4-week baseline period during which no statistically significant alteration in the group mean score was found. The addition of an extra month of intervention revealed that improvements may be larger when applying a low-intensity program over a period of 2 months. A limitation of the study was the small number of participants (n = 9), whereas the generalizability of the findings is limited by the exclusion of those with potentially significant comorbidities: a larger study would be needed to determine whether similar benefits are apparent in a broader range of participants. An adequately powered, pragmatic randomized controlled trial would be an appropriate next step to test the effectiveness of the suggested protocol and the potential added benefit that the inclusion of feedback might offer.

Back to Top | Article Outline


Our observations suggest that an intervention applied in a natural environment, combined with custom-made activities to increase motivation, may be effective in facilitation of the motor learning achieved with CIMT. An intervention of 1 to 2 hours of daily constraint use and practice may be more effective if applied in a 2-month therapy program rather than 1 month.

Back to Top | Article Outline


1. DeLuca SC, Echols K, Law CR, Ramey SL. Intensive pediatric constraint-induced therapy for children with cerebral palsy: randomized, controlled, crossover trial. J Child Neurol. 2006;21:931–938.
2. Choudhary A, Gulati S, Kabra M, et al. Efficacy of modified constraint induced movement therapy in improving upper limb function in children with hemiplegic cerebral palsy: a randomized controlled trial. Brain Dev. 2013;35:870–876.
3. Wallen M, Ziviani J, Herbert R, Evans R, Novak I. Modified constraint-induced therapy for children with hemiplegic cerebral palsy: a feasibility study. Dev Neurorehab. 2011;11:124–133.
4. Lin KC, Wang TN, Wu CY, et al. Effects of home-based constraint-induced therapy versus dose-matched control intervention on functional outcomes and caregiver well-being in children with cerebral palsy. Res Dev Disabil. 2011;32:1483–1491.
5. Aarts PB, Jongerius PH, Geerdink YA, van Limbeek J, Geurts AC. Effectiveness of modified constraint-induced movement therapy in children with unilateral spastic cerebral palsy: a randomized controlled trial. Neurorehabil Neural Repair. 2010;24:509–518.
6. Gordon A, Conelly A, Neville B, et al. Modified constraint-induced movement therapy after childhood stroke. Dev Med Child Neurol. 2007;49:23–27.
7. Eliasson AC, Krumlinde-Sundholm L, Shaw K, Wang C. Effects of constraint-induced movement therapy in young children with hemiplegic cerebral palsy: an adapted model. Dev Med Child Neurol. 2005;47:266–275.
8. Chen HC, Chen CL, Kang LJ, Wu CY, Chen FC, Hong WH. Improvement of upper extremity motor control and function after home-based constraint induced therapy in children with unilateral cerebral palsy: immediate and long-term effects. Arch Phys Med Rehabil. 2014;95:1423–1432.
9. Rostami HR, Malamiri RA. Effect of treatment environment on modified constraint-induced movement therapy results in children with spastic hemiplegic cerebral palsy: a randomized controlled trial. Disabil Rehabil. 2012;34:40–44.
10. Al-Oraibi S, Eliasson AC. Implementation of constraint-induced movement therapy for young children with unilateral cerebral palsy in Jordan: a home-based model. Disabil Rehabil. 2011;33:2006–2012.
11. Hsin YJ, Chen FC, Lin KC, Kang LJ, Chen CL, Chen CY. Efficacy of constraint-induced therapy on functional performance and health-related quality of life for children with cerebral palsy: a randomized controlled trial. J Child Neurol. 2012;27:992–999.
12. Psychouli P, Burridge J, Kennedy C. Forced use as a home-based intervention in children with congenital hemiplegic cerebral palsy: choosing the appropriate constraint. Disabil Rehabil Assist Technol. 2010;5:25–33.
13. Portney LG, Watkins MP. Foundations of Clinical Research: Applications to Practice. 3rd ed. New Jersey: Pearson & Prentice Hall; 2009.
14. Raven J, Raven JC, Court JH. Coloured Progressive Matrices: Introducing the Parallel Version. Oxford: Oxford Psychologists Press; 1998.
15. Wulf G, Schmidt RA. Feedback-induced variability and the learning of generalized motor programs. J Mot Behav. 1994;26:348–361.
16. Randall M, Carlin JB, Chondros P, Reddihough DS. Reliability of the Melbourne Assessment of Unilateral Upper Limb Function. Dev Med Child Neurol. 2011;43:761–767.
17. DeMatteo C, Law M, Russell D, Pollock N, Rosenbaum P, Walter S. The reliability and validity of the Quality of Upper Extremity Skills Test. Phys Occup Ther Pediatr. 1993;13:1–18.
18. Randall M, Johnson L, Reddihough D. The Melbourne Assessment of Unilateral Upper Limb Function: Test Administration Manual. Melbourne, Australia: Royal Children's Hospital.
19. Klingels K, Feys H, Molenaers G, et al. Randomized trial of modified constraint-induced movement therapy with and without an intensive therapy program in children with unilateral cerebral palsy. Neurorehabil Neural Repair. 2013;27:799–807.
20. Gordon AM, Charles J, Wolf SL. Methods of constraint-induced movement therapy for children with hemiplegic cerebral palsy: development of a child-friendly intervention for improving upper-extremity function. Arch Phys Med Rehabil. 2005;86:837–844.
21. Eliasson AC, Shaw K, Berg E, Krumlinde-Sundholm L. An ecological approach of Constraint Induced Movement Therapy for 2–3-year-old children: a randomized control trial. Res Dev Disabil. 2011;32:2820–2828.
22. Chen CL, Kang LJ, Hong WH, Chen FC, Chen HC, Wu CY. Effect of therapist-based constraint-induced therapy at home on motor control, motor performance and daily function in children with cerebral palsy: a randomized controlled study. Clin Rehabil. 2013;27:236–245.
23. Sakzewski L, Ziviani J, Abbott DF, Macdonell RA, Jackson GD, Boyd RN. Equivalent retention of gains at 1 year after training with constraint-induced or bimanual therapy in children with unilateral cerebral palsy. Neurorehabil Neural Repair. 2011;25:664–671.
24. Fleiss JL. Analysis of data from multiclinic trials. Control Clin Trials. 1986;7:267–275.

cerebral palsy; child; hemiplegia; home care; motor skills; physical therapy methods; upper limb function

Copyright © 2016 Academy of Pediatric Physical Therapy of the American Physical Therapy Association