intensive therapy following upper limb botulinum toxin a injection in young children with unilateral...

DEVELOPMENTAL MEDICINE & CHILD NEUROLOGY ORIGINAL ARTICLE

Intensive therapy following upper limb botulinum toxin Ainjection in young children with unilateral cerebral palsy:a randomized trial

BRIAN HOARE1,2 | CHRISTINE IMMS1,3,4 | ELMER VILLANUEVA5 | HYAM BARRY RAWICKI2 |THOMAS MATYAS6,7 | LEEANNE CAREY1,7

1 Department of Occupational Therapy, La Trobe University, Bundoora; 2 Victorian Paediatric Rehabilitation Service, Monash Children’s, Melbourne; 3 MurdochChildrens Research Institute, Melbourne; 4 National Centre for Clinical Outcomes Research and School of Allied and Public Health, Australian Catholic University,Melbourne; 5 Gippsland Medical School, Monash University, Churchill; 6 School of Psychological Sciences, La Trobe University, Bundoora; 7 Stroke Division, FloreyInstitute of Neuroscience and Mental Health, Heidelberg, Australia.

Correspondence to Dr Brian Hoare, Victorian Paediatric Rehabilitation Service, Monash Children’s, 246 Clayton Road, Clayton, Vic. 3168, Australia. E-mail: [email protected]

PUBLICATION DATA

Accepted for publication 9 October 2012.

Published online

ABBREVIATIONS

AHA Assisting Hand Assessment

BOT Bimanual occupational therapy

CIMT Constraint-induced movement therapy

COPM Canadian Occupational

Performance Measure

EMD Estimated mean difference

GAS Goal Attainment Scale

mCIMT Modified constraint-induced

movement therapy

PEDI Pediatric Evaluation of

Disability Inventory

QUEST Quality of Upper Extremity

Skills Test

AIM Botulinum toxin A (BoNT-A) combined with occupational therapy is effective in

improving upper limb outcomes in children with unilateral cerebral palsy (CP). It is now

essential to identify the most effective therapies following BoNT-A. Given the added burden

for children and families, the aim of this study was to explore whether modified constraint-

induced movement therapy (mCIMT) leads to sufficiently superior gains compared with

bimanual occupational therapy (BOT) in young children with unilateral CP following BoNT-A

injections.

METHOD In this randomized, controlled, evaluator-blinded trial, 34 children (20 males, 14

females; mean age 3y, SD 1y 4mo, range 18mo–6y) with unilateral CP were randomized

using concealed allocation to one of two 8-week interventions. The experimental group

(n=17) received BoNT-A and mCIMT. The comparison group (n=17) received BoNT-A and

BOT. Participants were recruited from a physical rehabilitation clinic and randomized

between August 2003 and May 2009. Primary outcome was measured using the Assisting

Hand Assessment at 3 months. Secondary outcomes were measured at 3 months and

6 months using the Quality of Upper Extremity Skills Test, the Pediatric Evaluation of Disability

Inventory, CanadianOccupational PerformanceMeasure, and theGoal Attainment Scale.

RESULTS There were no clinically important differences between groups at baseline.

Immediately following intervention, there was no evidence of a superior effect for BoNT-A

+ mCIMT as determined by the Assisting Hand Assessment (estimated mean difference

[EMD] 0.81, upper 95% confidence limit 3.6; p=0.32) or secondary outcomes. However, both

groups showed improvement over time (BoNT-A + mCIMT: EMD 2.7, 95% confidence interval

[CI] 0.7–5.2; BONT-A + BOT: EMD 4.7, 95% CI 2.1–8.6). Follow-up at 6 months also

demonstrated no superior effect for BoNT-A + mCIMT.

INTERPRETATION Following upper limb injection of BoNT-A, there was no evidence that

mCIMT, despite the significantly increased intensity of the home programme, produced a

superior effect across a range of outcomes compared with a structured programme of BOT in

young children with unilateral CP.

There is high-level evidence that injection of botulinum toxin A(BoNT-A), in combination with occupational therapy, improvesupper limb outcomes in children with cerebral palsy (CP)1 How-ever, studies evaluating BoNT-A as an adjunct to occupationaltherapy have primarily focused on the effect of BoNT-A and,thus, have provided limited information on post-injection thera-pies.2 Further investigation is essential to establish what specifictherapy, and atwhat intensity, will enhance functional outcomes.2

Injection of BoNT-A alone does not improve the abil-ity of a child with unilateral CP to use the injected limb.1

Post-injection therapy should, therefore, aim to create theopportunity, motivation, experience, and environment inwhich a child can learn how to use his or her impairedlimb.3 Constraint-induced movement therapy (CIMT) hasbeen proposed as a method of achieving this.2 By con-straining use of the dominant limb, CIMT allows theprovision of intensive practice of activities with theimpaired limb. Although highly intensive models ofCIMT have proved popular, a 2007 Cochrane reviewidentified a significant treatment effect only in a single

© The Authors. Developmental Medicine & Child Neurology © 2012 Mac Keith Press DOI: 10.1111/dmcn.12054 1

trial that adopted a less intensive, modified form ofCIMT.3,4

Bimanual occupational therapy (BOT) in children withunilateral CP is not new.5 Despite this, structured pro-grammes of BOT have rarely been described and evalu-ated. BOT focuses on improving a child’s occupationalperformance (the performance of self-care, productivity,and leisure activities) in the context of motivating, mean-ingful, and purposeful bimanual activities.6 In 2007, abimanual approach to training called hand–arm bimanualintensive therapy was evaluated.7 Despite similarities withBOT, there are specific differences between the therapies,8

particularly the intensity of treatment used in hand–armbimanual intensive therapy (6h per day for 2wks). Childrenparticipating in hand–arm bimanual intensive therapy dem-onstrated improved bimanual performance compared witha usual care comparison group.7 In 2008, a CIMT protocolwas compared with hand–arm bimanual intensive therapyin children aged 3 to 13 years.9 Outcomes from the Assist-ing Hand Assessment (AHA) demonstrated that improve-ments in bimanual performance of similar magnitude wereachieved in both groups.

The primary purpose of this study was to compare theeffects, following upper limb injection of BoNT-A, ofmodified constraint-induced movement therapy (mCIMT)and a clearly defined, conventional programme of BOT onbimanual upper limb performance in young children withunilateral CP. Because of the significantly increased com-mitment required to participate in mCIMT, we sought toexplore if mCIMT leads to sufficiently superior gains tojustify the added burden for children, families, and thera-pists. Our a priori hypothesis was that mCIMT would bemore effective than BOT in improving outcomes.

METHODSetting and participantsChildren were recruited (August 2003 until May 2009)from the Physical Rehabilitation Clinic of Monash Chil-dren’s, Southern Health, Victoria, Australia. The random-ized controlled trial (RCT) was approved by the ethicscommittees of Southern Health and La Trobe Universityand was registered with the Australian and New ZealandClinical Trials Registry (ACTRN12605000002684).

Children were eligible to participate in the study if theymet the following criteria: diagnosis of congenital spasticunilateral CP; age 18 months to 6 years; active movement ofthe impaired upper limb allowing reach forward to anelevated position and grasp of a cube from a table; ability tofollow simple one-stage commands; moderate levels ofincreased muscle tone (i.e. 1–2 on the modified AshworthScale)10 and spasticity (i.e. 1–2 on the modified TardieuScale);11 and no fixed contracture. The Manual Ability Clas-sification System12 was not available during the design phaseof this trial. All children were also required to be appropriatefor upper limb BoNT-A injection. Children who otherwisemet the inclusion criteria were excluded if they had receivedBoNT-A injections in the upper limb in the past 12 months

or had had prior upper limb surgery. Informed consent wasobtained before enrolment in the study.

Random assignmentFollowing consent and prior to baseline assessment, chil-dren were randomized into pairs matched by age (SD6mo). A computer-generated set of random numbers wasused to create an allocation sequence that was contained inindividual sealed, opaque envelopes for use by the chiefinvestigator. As children were recruited, the next envelopein the sequence was opened and the child assigned to thestated group. All randomization, sequence generation, andpreparation of group allocation materials were performedby a third party who had no direct contact with the clinicalaspects of the trial.

InterventionThe details of the intervention provided in this RCT havebeen previously reported8 and will be outlined brieflyhere.

Intervention received by both groupsEach child received injections of BoNT-A (Botox; AllerganAustralia P/L, Gordon, NSW, Australia). Injections wereperformed under general anaesthetic. Muscle localizationwas undertaken using Teflon-coated Botox needles (37mm,27 gauge) with neuromuscular electrical stimulation. Themaximum dose was 15U/kg (up to 400U) and dilution was100U/mL. Children with increased resistance to passivestretch (� 1 on the modified Ashworth Scale) or who dem-onstrated signs of muscle shortening were provided withan overnight custom-made thermoplastic resting/stretchingsplint.

One month after injection, children allocated to theexperimental group received mCIMT. Children allocatedto the comparison group received conventional BOT. Allchildren received individual, clinic-based treatment sessionsof approximately 60 minutes, twice weekly for 8 weeks,provided by the principal investigator (BH), who wasblinded to all outcome data. Home programmes for bothgroups were specifically designed for each child by thetreating therapist. On a daily basis, all caregivers com-pleted a logbook detailing the amount of home therapyundertaken. Logbooks were reviewed weekly by the treat-ing therapist.

Intervention group (modified constraint-induced movementtherapy)The mCIMT protocol was chosen based on the positiveoutcomes demonstrated by Eliasson et al.4 A Neoprene

What this paper adds• There was no evidence that mCIMT is superior to structured bimanual occu-

pational therapy.

• Sustained improvement can be achieved by unimanual or bimanual trainingfollowing BoNT-A injection.

• Eight weeks of twice-weekly upper limb intervention following BoNT-A injec-tion improves outcomes.

2 Developmental Medicine & Child Neurology 2012

(wetsuit material) glove was worn on the dominant, leastimpaired hand to restrict use of that limb. Children wereexpected to wear the glove for 3 hours per day, inclusiveof therapy time and the home programme, 7 days per weekfor the 8-week treatment period (total 168h). Glove usecould occur in sessions of no less than 30 minutes dura-tion. Unimanual tasks were selected to facilitate repetitivepractice of movement and skills of the impaired limb.Therapy was based on principles of motor learningtheory.13

Comparison group (bimanual occupational therapy)BOT targeted the development of specific hand skills andmotor planning abilities using repetitive practice of biman-ual activities. Knowledge of the AHA item difficulty hierar-chy14 served as an important guide for selecting specificactivities, but children were not trained to complete AHAtasks. Treatment incorporated principles of motor learningand cognitive-based motor intervention.15 Children andfamilies were encouraged to undertake a home programme,but no time requirements were specified. This was basedon current clinical practice and was designed to reflect dif-ferences in treatment intensity between protocols.

Data collection and instrumentsOutcomes were administered at baseline (1–2wks beforeinjection), and at 1 month, 3 months, and 6 months afterinjection by a senior occupational therapist blinded togroup assignment. Bimanual upper limb performance,assessed using the AHA, was selected as the primary mea-sure as the central aim of all upper limb motor-basedinterventions in children with hemiplegia is to improve useof the impaired hand as an assisting hand.

Primary outcome measureThe AHA (English version 4.4) is a standardized, crite-rion-referenced test of bimanual performance for childrenaged 18 months to 12 years with unilateral upper limbimpairment.14 It has strong evidence of validity and reli-ability.16 The AHA aims to measure how effectively a childuses the impaired hand in bimanual play activities using 22items. AHA raw scores were converted to interval-leveldata using Rasch analysis. Rescaled logit-based AHA unitsranging from 0 to 100 were obtained, with a higher num-ber indicating higher ability.17 The smallest detectable dif-ference for the AHA is 0.97 logits, which corresponds to 5AHA units.16 The AHA has demonstrated responsivenessto change following both CIMT and bimanual interven-tion.4,7,9 In this RCT, the AHA was administered by asenior occupational therapist blinded to group assignmentand scored by two of the AHA test developers, who wereblinded to group allocation and order of assessment.

Monitoring response to BoNT-A injectionThe modified Ashworth Scale10 and modified TardieuScale11 were used to monitor the effects of the BoNT-Ainjection. All modified Ashworth Scale assessments were

performed from maximal possible flexion to maximum pos-sible extension over the duration of about 1 second.10 Thevelocity used to determine the angle of catch (R1) for themodified Tardieu Scale was V3 (as fast as possible), asrecommended by Morris.18 Using a goniometer, the differ-ence between the R1 and the full passive range of move-ment (R2) was recorded. To improve reliability, measureswere jointly performed by the same senior occupationaltherapist and senior physiotherapist blinded to groupassignment.

Secondary outcome measuresThe Quality of Upper Extremity Skills Test (QUEST) is areliable and valid measure for evaluating the quality ofupper limb movement in children with CP.19 Thirty-fouritems of upper extremity function in four domains are eval-uated: dissociated movement, grasp, protective extension,and weight-bearing. Both the impaired and dominantupper limbs were tested. Data from individual QUESTdomains targeted by the intervention, and therefore likelyto change, were analysed. A standardized score rangingfrom 0 (low ability) to 100 (high ability) was obtained forboth the dissociated movement and grasp domains. Theseare reported separately as recommended by Thorleyet al.20 The QUEST was videotaped and scored by asses-sors blinded to group allocation and order of assessment.

The Pediatric Evaluation of Disability Inventory (PEDI)is a valid and reliable parent-report assessment of how achild functions in the context of daily life.21 Owing to thenature of interventions provided, only the self-care domain(functional skills and caregiver assistance) of the PEDI wasadministered. This is consistent with other trials evaluatingupper limb BoNT-A injections.22 Scaled scores (range 0–100) were used for data analysis. A change in score ofabout 11% on the total scale has been found to representclinically meaningful change.23

The Canadian Occupational Performance Measure(COPM)24 is designed to detect change in an individual’sperception of occupational performance in self-care abilities,productivity, and leisure activities. It has established validityand reliability.25 Owing to the age of the children in thisstudy, parental responses to the COPM were obtained,rather than the child’s.26 The five most important occupa-tional performance problem areas were selected and scoreswere obtained for performance and satisfaction, with perfor-mance ranging from 1 to 10. A change in summary scores (i.e. between initial and subsequent scores) of 2 or more pointsis considered clinically significant.25

The Goal Attainment Scale (GAS) is an individualized,criterion-referenced measure of treatment-inducedchange.27 Three of the most important goals identified byparents using the COPM were used for scaling using theGAS. At baseline, each goal was rated on a five-point scalefrom �2 (current level of performance) to +2 (muchgreater than expected outcome). Goals were not weightedand were, therefore, treated as being of equal importance.Achieved goals, those that changed from �2 (baseline level

Upper Limb Therapy Following BoNT-A Injection Brian Hoare et al. 3

of performance) to 0 (expected level of performance), weredeemed to represent clinically important change.

SafetyEach child was monitored prospectively by the chief inves-tigator. Following trial completion, all medical andresearch records were retrospectively audited. Adverseevents were recorded and classified according to whetherthey could be attributed to BoNT-A injection, generalanaesthesia, or therapy.

Sample sizeA priori sample size estimates were based on the projectedtreatment effect on the AHA at 3 months. The sample sizeof 17 children in the experimental group and 17 in thecomparison group was estimated to give an 80% probabil-ity of detecting an effect of 2.6 logits on the AHA, with atwo-tailed alpha of 5%, based on an anticipated standarddeviation of 3.14 logits.4

STATISTICAL ANALYSISAnalyses were on an intention-to-treat basis according toCONSORT guidelines28 and conducted using PASW Sta-tistics 18 software (SPSS Inc., Chicago, IL, USA). Descrip-tive statistics were calculated to identify potential baselinedifferences between the groups for all measures. The keyquestion was whether the gains produced by mCIMT weresufficiently superior to those achieved with BOT. Therefore,one-sided 95% confidence intervals (CIs) and one-tailed nullhypothesis tests were computed to compare gains immedi-ately following treatment and at the 6-month follow-up. Foreach treatment, two-sided 95% CIs were used to estimatemagnitude of gains immediately following treatment and atthe 6-month follow-up. Before analysis for the presence ofsignificant skew, heterogeneity of variance, and (whereappropriate) heterogeneity of covariance was conducted forall comparisons. Where significant violations were evident,data transformations were employed before the estimationof CIs. To improve precision of estimation and decrease riskof bias, the effects of initial AHA value, age, sex, and effectof BoNT-A alone (change to month 1 only) were investi-gated. All comparisons between groups were conducted withadjustment for the effects of these variables and withoutadjustments. We assessed the bivariate effects of age, inten-sity of treatment, and baseline AHA performance onoutcome using regression. For analysis of GAS data, the pro-portion of achieved goals in each group was calculated andassociations between group and goal attainment at 3 monthsand 6 months were assessed using v2 statistics.

RESULTSOf the 58 children screened, 35 were eligible to participatein the study (see Fig. 1). Thirty-five children consentedand were randomly assigned: 17 to the experimental group(BoNT-A + mCIMT) and 18 to the comparison group(BONT-A + BOT). Although we intended to analyseaccording to intention to treat, one participant withdrew

after randomization, but before baseline assessment.Therefore, analysis using the last observation carried for-ward principle was not possible. Data were obtained forthe remaining 34 participants. All outcomes are reportedfor each group at each follow-up period in Table SI(online supporting information).

Baseline comparisonsAt baseline, there was no evidence of clinically importantdifferences between groups for key demographic, impair-ment, and activity domain characteristics (Table I). Therewere 20 males and 14 females (mean age 3y, SD 1y 4mo,range 18mo–6y), and 18 children had right-sided hemiplegia.

Amount of interventionAll children from both groups received the allocated train-ing during the scheduled 16 clinic-based treatment ses-sions. Thus, every child included in this RCT received atotal of 16 hours of therapy-guided intervention over the8-week treatment period. Those in the mCIMT groupreceived this therapy with the glove on. Consistent withthe design of the intervention models, there was strongevidence of a difference between groups in the intensity ofhome programme (mean hours [95% CI]: BoNT-A + mCIMT 98.5 [81.9–115.1]; BONT-A + BOT 31.6,[15.4–47.9]; p� 0.001). Children in the BoNT-A + mCIMT group wore the restraint mitt (therapy ses-sions and home programme) for a mean 98.5 (SD 32)hours of the expected 168 hours.

Effect of BoNT-A in both groupsAs reported previously, the effects of BoNT-A alone werecontrolled for by analyses of covariance. There was no evi-dence of a difference between the two groups in the dose ofBoNT-A (BoNT-A + mCIMT: mean 5.7U/kg, 95% CI 4.5–6.8U/kg; BONT-A + BOT: mean 4.9U/kg, 95% CI 3.7–6.1U/kg; p=0.34), muscles injected, or reduction of spasticityor muscle tone (Tables SII–SIV, online supporting informa-tion). There was a reduction in elbow flexor spasticity at1 month in the BoNT-A + mCIMT group and at 3 monthsin the BONT-A + BOT group. This was not sustained at6 months in either group. Injection of BoNT-A reducedwrist flexor spasticity in the BONT-A + BOT group at 1month, 3 months, and 6 months. Changes in muscle toneusing the modified Ashworth Scale were consistent withthose found on the modified Tardieu Scale (Table SIV).

Primary outcome: Assisting Hand Assessment at 3 monthsWith or without adjustment for the effects of age, sex,baseline AHA values, and BoNT-A effects (Table II), therewas no evidence of superior effect for BoNT-A + mCIMTimmediately following intervention (adjusted estimatedmean difference [EMD] 0.81, upper 95% confidence limit3.6; p=0.32). Both groups, however, demonstratedsignificant gains in bimanual performance from baseline(BoNT-A + mCIMT: EMD 2.7, 95% CI 0.7–5.2;BONT-A + BOT: EMD 4.7, 95% CI 2.1–8.6; Fig. 2).


Secondary outcomesThere was no statistical evidence of the superiority forBoNT-A + mCIMT for any secondary outcome measureimmediately following intervention (Table II). In bothgroups there was evidence of statistically important gainson both domains of the PEDI and for performance andsatisfaction on the COPM. Despite improvements inbimanual performance, there was no evidence of

statistically important gains on the QUEST for eithergrasp (BoNT-A + mCIMT: EMD 6.1, 95% CI 0.0–12.3;BONT-A + BOT: EMD 5.1, 95% CI �1.0 to 11.3) ordissociated movement (BoNT-A + mCIMT: EMD 3.4,95% CI �4.3 to 11.0; BONT-A + BOT: EMD 3.3, 95%CI �4.3 to 11.0). Following either treatment, 66 out of102 or 65% of goals were scaled using the GAS at greaterthan or equal to their expected outcome (i.e. � 0; Fig. 3).

Assessed for eligibility(n=58)

Excluded (n=23)Not hemiplegic (n=3)

Allocated to intervention (mCIMT) (n=17) Allocated to control (BOT) (n=18)

Received allocated intervention (n=17)•

••

••••

•

•

•

•

•

•

•

•

•

•

•

Ataxic cerebral palsy (n=1)No active use of affected upper limb (n=3)

Did not receive allocated intervention(n=1)a

Randomized (n=35)

Enr

ollm

ent

Allo

catio

nF

ollo

w-u

pA

naly

sis

Unable to meet study commitments (n=7)Did not require BoNT-A injection (n=7)

Unable to follow one stage commands (n=2)

Received allocated intervention (n=17)

Lost to follow-up (n=0)

Analysed (n=17)

Excluded from analysis (n=0)

Analysed (n=17)

Excluded from analysis (n=1)

Discontinued intervention (n=0)

Lost to follow-up (n=0)

Discontinued intervention (n=0)

Did not receive allocated intervention(n=0)

Figure 1: Flow of participants through each stage of the randomized controlled trial. aAfter enrolment, one child withdrew before baseline assessment andtreatment because of family stress. BoNT-A, botulinum toxin type A; mCIMT, modified constraint–induced movement therapy; BOT, bimanual occupationaltherapy.

Table I: Baseline characteristics of groups

GroupSide ofhemiplegia (L/R) Sex (M/F)

MeanBoNT-Adose (U)(SD)

Mean age(mo) (SD)

Mean group measures at baseline (SD)

AHA QUEST PEDI

Logits UnitsDissociatedmovement Grasps

Functionalskills

Caregiverassistance

mCIMT(n=17)

11/6 11/6 77.9 (29.9) 36.1 (15.6) �2.6 (3.2) 40.4 (16.9) 70.1 (10.8) 52.6 (20.6) 31.8 (8.7) 37.0 (13.8)

BOT(n=17)

5/12 9/8 67.7 (30.6) 35.6 (16.4) �2.2 (3.0) 42.5 (15.6) 71.9 (11.7) 56.0 (21.2) 37.1 (10.7) 37.7 (14.2)

Total 16/18 20/14 72.8 (30.3) 35.8 (15.8) �2.4 (3.1) 44.0 (12.6) 71.0 (11.1) 54.3 (20.7) 34.4 (10.0) 37.3 (13.8)

L, left; R, right; M, male; F, female; BoNT-A, botulinum toxin A; U, units; AHA, Assisting Hand Assessment; QUEST, Quality of UpperExtremity Skills Test; PEDI, Pediatric Evaluation of Disability Inventory; mCIMT, modified constraint-induced movement therapy; BOT,bimanual occupational therapy.


There was no evidence of an interaction between groupand goal attainment (p=0.56).

Longer-term follow-upExcept for the QUEST grasp domain (EMD 5.8, upper95% confidence limit 11.6, p=0.05), there was no evidence ofstatistically significant superiority for BoNT-A + mCIMTfor any outcome measure at the 6-month follow-up(Table III). Consistent with outcomes immediatelyfollowing intervention, there was significant within groupgains for both groups at 6 months for the AHA, PEDI, andCOPM. There was no evidence of within-group gains onthe QUEST except for the grasp domain for theBoNT-A + mCIMT group (EMD 8.1, 95% CI 3.2–13.1).Goal achievement, measured using the GAS, was maintainedat 6 months in both groups (mCIMT: 34/51 [67%]; BOT:31/51 [61%]). There was no evidence of an interactionbetween group and goal attainment at 6 months (p=0.30).

Factors related to improvementInvestigation of individual variables using pairwise correla-tions indicated a small inverse correlation between changeon the AHA and age in both the BoNT-A + BOT group(r=�0.42) and the BoNT-A + mCIMT group (r=�0.30).This suggests that younger children improved more than

Table II: Differences within and between BoNT-A + mCIMT (n=17) and BoNT-A + BOT (n=17) and changes at 3 months (immediately following mCIMTor BOT) for both groups for primary and secondary outcomes

Measure Group

Unadjustedchangewithin groups,M0 toM3, EMD (95% CI)

Unadjusteddifferencebetween groups,M0 to M3, EMD(upper 95%confidence limit) p

Adjusted changewithin groups,M0 to M3,EMD (95% CI)

Adjusteddifferencebetweengroups, M0to M3, EMD(upper 95%confidencelimit) p

Primary outcomeAHA BoNT-A + BOT 4.7 (2.1–8.6)a �2.0 (1.1)a 0.13a 4.8 (2.5–7.1)a 0.8 (3.6)a 0.32a

BoNT-A + mCIMT 2.7 (0.7–5.2)a 5.6 (3.3–7.9)a

Secondary outcomesQUEST grasps BoNT-A + BOT 4.6 (�1.8�11.0) 2.1 (9.6) 0.32 5.1 (–1.0�11.3) 1.0 (8.3) 0.41

BoNT-A + mCIMT 6.7 (0.3–13.1) 6.1 (0.0–12.3)QUEST dissociatedmovement

BoNT-A + BOT 3.1 (�5.1�11.3) 0.5 (10.2) 0.47 3.3 (�4.3�11.0) 0.0 (9.1) 0.50BoNT-A + mCIMT 3.6 (�4.6–11.8) 3.4 (�4.3�11.0)

PEDI functional skills BoNT-A + BOT 8.0 (4.6�11.4) 1.5 (5.5) 0.26 7.3 (4.4–10.2) 3.0 (6.6) 0.08BoNT-A + mCIMT 9.6 (6.2–13.0) 10.3 (7.4–13.2)

PEDI caregiver assistance BoNT-A + BOT 9.1 (4.1–14.1) 0.3 (6.2) 0.46 9.0 (4.7–13.3) 0.6 (5.7) 0.42BoNT-A + mCIMT 9.4 (4.4–14.4) 9.6 (5.3–13.9)

COPM performance BoNT-A + BOT 3.1 (2.1–4.0) 0.2 (1.3) 0.38 3.0 (2.2–3.8) 0.3 (1.2) 0.30BoNT-A + mCIMT 3.3 (2.3–4.2) 3.3 (2.5–4.1)

COPM satisfaction BoNT-A + BOT 3.1 (2.0–4.2) 0.0 (1.3) 0.48 3.0 (2.1–3.9) 0.3 (1.6) 0.33BoNT-A + mCIMT 3.1 (2.0–4.2) 3.3 (2.4–4.1)

aAs a result of the presence of skew in the Assisting Hand Assessment (AHA) gains, a logarithmic transformation was applied to improvedistribution symmetry before analysis. These results are based on that analysis. Non-parametric comparisons and Hodges–Lehmann confi-dence intervals were also used to verify these estimates and provided very similar results to those tabled. Column 4: the unadjustedchange from baseline to 3 months (immediately post botulinum toxin type A [BoNT-A] injection and modified constraint-induced move-ment therapy [mCIMT] or bimanual occupational therapy [BOT]) for the mCIMT and BOT groups. Column 5: the unadjusted differencebetween groups in change from baseline to 3 months. A positive value indicates a greater increase from baseline to 3 months for themCIMT group for all measures. Column 9: The difference between groups in change from baseline to 3 months adjusted for age, sex,baseline AHA values, and BoNT-A effects. A positive value indicates a greater increase from baseline to 3 months for the mCIMT group forall measures. All p-values reported are one-tailed. M0, baseline assessment period; M3, immediate post-treatment assessment period;EMD, estimated mean difference; CI, confidence interval; QUEST, Quality of Upper Extremity Skills Training; PEDI, Pediatric Evaluation ofDisability Inventory; COPM, Canadian Occupational Performance Measure.

0

0

–0.5

–1

–1.5

–2.5

–3.5

–2

–3

–41 2 3 4 5 6

Month

mCIMT

BOT

Ass

istin

g H

and

Ass

essm

ent l

ogits

Figure 2: Results of the Assisting Hand Assessment (AHA) in logits for botu-linum toxin A (BoNT-A) and modified constraint-induced movement therapy(mCIMT) and BoNT-A and bimanual occupational therapy (BOT) at baseline,1 month, 3 months, and 6 months. Circles and squares indicate group meansfor mCIMT and BOT respectively. Error bars represent the 95% confidenceintervals.


No change frombaseline (–2)

< Expected outcome(–1)

> Expected outcome(+1)

Much > expectedoutcome (+2)

Num

ber

of c

hild

ren

Expected outcome(0)

No change from baseline (–2)

< Expected outcome (–1)

> Expected outcome (+1)

Much > expected outcome (+2)

Expected outcome (0)

35

30

25

20

15

10

10

6

1

1

0

0

5

33 31 7

7

79 10

9

9

9

9

8

7

12

14

12

12

11

13 14

13

13

11

5

3

mCIMT1mo

mCIMT3mo

mCIMT6mo

BOT1mo

BOT3mo

BOT6mo

Figure 3: Results of goal attainment scaling showing a table of raw data from each group at each time point. Each child had three Goal AttainmentScale (GAS) goals scaled at each time point (total 51 goals per group). The graph above each table column provides frequency counts in each GASscale category (�2, �1, 0, +1, +2) for the group presented in the column below. mCIMT, modified constraint–induced movement therapy; BOT, bimanualoccupational therapy.

Table III: Differences within and between BoNT-A + mCIMT (n=17) and BONT-A + BOT (n=17) and changes at 6-month follow-up for both groups forsecondary outcomes

Measure Group

Unadjusted changewithin groups, M0to M6, EMD (95% CI)

Unadjusteddifferencebetweengroups, M0 toM6, EMD (upperlimit 95% CI) p

Adjusted changewithin groups,M0 to M6, EMD(95% CI)

Adjusteddifferencebetweengroups,M0 to M6,EMD (upperconfidencelimit 95% CI) p

AHA BoNT-A + BOT 5.1 (2.3–9.3)a �0.8 (1.4)a 0.27a 6.0 (3.7�8.4)a �0.6 (2.3)a 0.36a

BoNT-A + CIMT 3.7 (1.4–7.2)a 5.5 (3.1�7.8)a

QUEST grasps BoNT-A + BOT 1.9 (�4.4�8.1) 6.7 (14.1) 0.07 2.3 (�2.6�7.3) 5.8 (11.6) 0.05BoNT-A + CIMT 8.6 (2.3–14.9) 8.1 (3.2–13.1)

QUEST dissociatedmovement

BoNT-A + BOT 3.8 (�3.2�10.8) �6.2 (2.0) 0.11 4.0 (�2.4�10.4) �6.6 (0.9) 0.07BoNT-A + CIMT �2.4 (�9.5�4.6) �2.6 (�9.1�3.8)

PEDI functional skills BoNT-A + BOT 12.2 (8.4–16.0) �1.8 (2.6) 0.24 11.4 (7.8–15.0) �0.2 (4.1) 0.47BoNT-A + CIMT 10.4 (6.6–14.1) 11.2 (7.6–14.7)

PEDI caregiverassistance

BoNT-A + BOT 12.6 (5.4–19.8) �2.7 (5.8) 0.30 12.1 (5.6–18.7) �1.8 (6.0) 0.35BoNT-A + CIMT 9.9 (2.7–17.2) 10.4 (3.8–16.9)

COPM performance BoNT-A + BOT 3.3 (2.4–4.1) �0.1 (0.9) 0.42 3.2 (2.4–3.9) 0.1 (1.0) 0.45BoNT-A + CIMT 3.1 (2.3–4.0) 3.2 (2.5–4.0)

COPM satisfaction BoNT-A + BOT 3.4 (2.3–4.5) �0.3 (1.0) 0.35 3.2 (2.4–4.1) 0.1 (1.1) 0.45BoNT-A + CIMT 3.1 (2.1–4.2) 3.3 (2.5–4.2)

aAs a result of the presence of skew in the Assisting Hand Assessment (AHA) gains, a logarithmic transformation was applied to improvedistribution symmetry before analysis. These results are based on that analysis. Non-parametric comparisons and Hodges–Lehmann confi-dence intervals were also used to verify these estimates and provided very similar results to those tabled. Column 4: the unadjustedchange from baseline to 6 months (immediately post botulinum toxin type A [BoNT-A] injection and modified constraint-induced move-ment therapy [mCIMT] or bimanual occupational therapy [BOT]) for the mCIMT and BOT groups. Column 5: the unadjusted differencebetween groups in change from baseline to 3 months. A positive value indicates a greater increase from baseline to 6 months for themCIMT group for all measures. Column 9: the difference between groups in change from baseline to 6 months adjusted for age, sex, base-line AHA values, and BoNT-A effects. A positive value indicates a greater increase from baseline to 3 months for the mCIMT group for allmeasures. All p values reported are one-tailed. M0, baseline assessment period; M6, post-treatment assessment at 6-month follow-up;EMD, estimated mean difference; CI, confidence interval; QUEST, Quality of Upper Extremity Skills Training; PEDI, Pediatric Evaluation ofDisability Inventory; COPM, Canadian Occupational Performance Measure.


older children regardless of intervention type. Using log-book data, there was an apparently paradoxical inverse cor-relation of small magnitude of change with increasingintensity of home practice for both groups (BoNT-A + BOT, r=�0.36; BoNT-A + mCIMT group, r=�0.13).In both groups there was also an inverse correlation forchange on the AHA and baseline performance (r=�0.39).For the BoNT-A + BOT group, the association (r=�0.62)was large, suggesting that children with lower baselineabilities improved more than those with higher baselineabilities. The association was small for the BoNT-A + mCIMT group (r=�0.17).

SafetyThere were nine adverse events following injection ofBoNT-A. Six children experienced excessive grip weaknessthat resolved by 4 weeks post injection. One child reportedsoreness at the injection site that resolved a few days afterinjection and another reported vomiting post injection.This was deemed by a medical specialist to be related tothe general anaesthesia. One child was admitted 1 weekafter the injection with blockage of an existing ventriculo-peritoneal shunt. This was considered to be unrelated toBoNT-A injection or general anaesthesia. No adverseevent was reported from mCIMT or BOT.

DISCUSSIONThis RCT demonstrated no evidence of a superior effectfor mCIMT on almost all outcomes when compared withBOT. This was despite adopting statistical methods toprovide the best opportunity to demonstrate superiority.The primary hypothesis that mCIMT would be superior asa result of the increased intensity of practice was not sup-ported. Considering the clinical importance of outcomesfor the AHA (range 0–100 units), which demonstratedupper 95% confidence limits for difference between groupsof 3.6 units (immediately following intervention) and 2.3units (6mo follow-up), the probability for mCIMT superi-ority of a useful size when allowing for statistical error islow. This is also consistent with the upper 95% confidencelimits for secondary measures including PEDI functionalskills (month 3 [M3]=6.6, month 6 [M6]=4.1), PEDI care-giver assistance (M3=6.6, M6=4.1), COPM performance(M3=1.2, M6=1.0), and COPM satisfaction (M3=1.6,M6=1.1). For the QUEST scales, both interventionsresulted in marked individual differences in the observedchanges. These large individual differences are manifestedas wide (imprecise) CIs for the mean changes associatedwith each treatment and estimates of the extent of CIMTsuperiority. Therefore, the only outcomes in this RCT,where it is not possible to dismiss the superiority ofmCIMT, include the QUEST grasp domain (M3=8.3,M6=11.6) and the QUEST dissociated movement domainimmediately following intervention (M3=9.1). The resultsof this trial are inconclusive for these outcomes. Takinginto consideration the increased burden for children andfamilies undertaking mCIMT, and in the absence of clear

superiority, these findings suggest that either of these clini-cally feasible models of intervention could be offered tochildren and families after injection of BoNT-A.

Despite statistical evidence of improvements at a grouplevel, even after controlling for important characteristicssuch as age, baseline AHA performance, and sex, there waslarge individual variability in response to both interven-tions. Fifty-six per cent of children (65% in the BoNT-A + mCIMT group; 47% in the BONT-A + BOT group)did not achieve change equal to or greater than the small-est detectable difference of 5 AHA units. The reason forthis variability remains unclear and requires further investi-gation. In contrast, four children who received BOT(Table SV: ID 16, 20, 21) or mCIMT (Table SV: ID 11)improved by more than 17 AHA units. All were female,aged less than 26 months, and had high baseline scores forthe dissociated movement subtest of the QUEST (i.e.� 62.5/100) but low baseline AHA scores (� 33 AHAunits). These children spontaneously used their impairedlimb in bimanual tasks only after verbal prompting or notat all, demonstrating what has been described as develop-ment disregard.29 That is, despite a relatively mild motorimpairment, the children did not spontaneously use theirimpaired limb in bimanual tasks.

For children in the BoNT-A + mCIMT group it wasexpected that 168 hours of mitt use would be achieved (3hper day for 8wks). However, the mean time achieved was59% of this (98.5h, SD 32h, range 22–143h). Many fami-lies reported difficulties due to illness, sleep times, andusual family, work, and social demands. Despite the varia-tion, there was no evidence that increased intensity of mittuse improved bimanual performance outcomes (r=�0.13).The lower actual mitt use in this study is consistent withother studies in which a removable restraint was used inyoung children and treatment was provided by teachers orparents in a home/preschool setting. In two studies byEliasson et al.,4,30 and a recent study by Wallen et al.,31

families achieved 53%, 58%, and 60% respectively, of anexpected total of 112 hours (2h/d). These findings suggestthat, when using a removable restraint, families have diffi-culty achieving, on average, more than 1 to 1.5 hours ofdaily restraint over a 2-month period. Despite this, clini-cally important improvements were observed in all studies,with small to large effect sizes. In comparison with othertrials of mCIMT or CIMT,4,9,30–35 an EMD from baselineof 5.6 adjusted AHA units for mCIMT in this RCT sug-gests that outcomes are within an equivalent range (mean4.5 AHA units; range 2.24–7.10).

The provision of therapy in this RCT (undertaken byBH) led to important clinical insights. The much higherintensity of the home programme required for mCIMTplaced increased demands on children and their families.Modified CIMT, however, was much easier to implementin therapy sessions. Restraint replaced the need for repeti-tive prompts to use the impaired hand. However, the useof restraint limited the choice and nature of activities prac-ticed. Restraint facilitated spontaneous use of the involved


limb for repetitive unimanual practice of grasp, hold, andrelease activities. This experience highlights the key differ-ence between interventions. Despite known impairments inmotor planning36 and bimanual coordination,37 mCIMTdid not allow the therapist to teach children how to usethe two hands together. In contrast, bimanual interventionallowed the use of cognitive–motor training principlesto teach a child how to use the hands to perform comple-mentary actions. Supported by a lack of superiority ofmCIMT, the essential ingredient for improving bimanualperformance appears to be related not to the intensity oftherapy, but to the ingredients of therapy. Therapistsshould perhaps consider each approach as complementaryrather than opposing options.

One limitation of this trial was the knowledge of groupassignment by children, families, and intervention thera-pist. Double blinding was not possible owing to the overtnature of mCIMT, and this may have led to parental biasduring administration of parent-reported outcomes such asthe PEDI, COPM, and GAS. To minimize the potentialimpact of this, the study hypotheses were not revealed tochildren or families. We also did not ask blinded assessorsto reveal whether they remained blinded to allocation ornominate their perception of each child’s allocation. Thelack of a true comparison group who may have receivedBoNT-A alone but no movement-based therapy was also alimitation; however, this issue is difficult to overcome, par-ticularly in the light of recent evidence supporting the useof occupational therapy in addition to BoNT-A in thispopulation. Single-centre trials such as this RCT are slowto recruit, and timely inclusion of larger samples is diffi-cult, thus affecting the sample size.

CONCLUSIONSFollowing injection of BoNT-A, there was no evidence of supe-riority for mCIMT compared with a structured programme ofBOT on bimanual upper limb performance, functional skills,occupational performance, or goal attainment outcomes inyoung children with unilateral CP. In the absence of superiority,clinicians should consider the specific goals for intervention andchoose the most developmentally appropriate, family-friendly,and convenient of these approaches.

ACKNOWLEDGEMENTS

The authors wish to express their sincere gratitude to the children

and their families who participated in this trial. They also sincerely

thank Trisnawati Tanumihardjo, Melanie Toy-Laing, and Cate

Clancy for administering the assessments and Lena Krumlinde-

Sundholm, Marie Holmefur, and Sue Greaves for scoring all the

videos. We also thank Natasha Lannin for her assistance and on-

going support with statistics. Finally, the authors acknowledge Al-

lergan Australia, Southern Health, and La Trobe University for

supporting this research. The study was funded by grants from La

Trobe University, Southern Health, and Allergan Australia Pty

Ltd.

DISCLOSURES

Allergan Australia provided partial support by providing the

BoNT-A (Botox) used in the study, by payment of research assis-

tants for blinded administration and scoring of assessments, and

video-editing services. The authors have no pecuniary interest in

Allergan.

BH is an occupational therapist and has received sponsorship

from Allergan Australia to attend and teach at conferences and

meetings but has no personal financial interest in Botox or any

related product.

CI is co-investigator of an RCT investigating the effect of

repeat injections of BoNT-A and occupational therapy in the

upper limbs of children with unilateral CP that has received sup-

port from Allergan Australia. In 2008, CI received a grant from

Allergan Australia to present results of this trial at the American

Academy of Cerebral Palsy and Developmental Medicine in

Atlanta, but has no personal financial interest in Botox or any

related product.

HBR has received sponsorship from Allergan Australia to

attend and teach at conferences and meetings but has no personal

financial interest in Botox or any related product.

SUPPORTING INFORMATION

The following additional material may be found online.

Table SI: Results of follow-up assessments for modified con-

straint-induced therapy and bimanual occupational therapy at

baseline, 1 month, 3 months, and 6 months

Table SII: Summary of botulinum toxin A injections for both

groups

Table SIII: Effects of botulinum toxin A injection on muscle

spasticity

Table SIV: Effects of botulinum toxin A injection on muscle

tone

Table SV: Baseline characteristics of individual participants

REFERENCES

1. Hoare BJ, Wallen MA, Imms C, Villanueva E, Rawicki

HB, Carey L. Botulinum toxin A as an adjunct to treat-

ment in the management of the upper limb in children

with spastic cerebral palsy (UPDATE). Cochrane Data-

base Syst Rev 2010; 1: CD003469.

2. Hoare BJ, Imms C. Upper-limb injections of botu-

linum toxin-A in children with cerebral palsy: a critical

review of the literature and clinical implications for

occupational therapists. Am J Occup Ther 2004; 58:

389–97.

3. Hoare BJ, Wasiak J, Imms C, Carey L. Constraint-

induced movement therapy in the treatment of the

upper limb in children with hemiplegic cerebral palsy.

Cochrane Database Syst Rev 2007; 2: CD004149.

4. 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: 266–75.

5. Eliasson AC. Bimanual training for children with unilat-

eral CP – is this something new? Dev Med Child Neurol

2007; 49: 806.

6. Law ML, Cooper BA, Strong S, Stewart D, Rigby P,

Letts L. Theoretical contexts for practice of occupa-

tional therapy. In: Christiansen CH, Baum CM, Editors.


Occupational Therapy: Enabling Function and Well-

Being. Thorofare, NJ: Slack, 1997: 72–104.

7. Gordon AM, Schneider JA, Chinnan A, Charles JR. Effi-

cacy of a hand-arm bimanual intensive therapy (HABIT)

in children with hemiplegic cerebral palsy: a randomized

control trial. Dev Med Child Neurol 2007; 49: 830–8.

8. Hoare BJ, Imms C, Rawicki HB, Carey L. Modified con-

straint-induced movement therapy or conventional occu-

pational therapy following injection of botulinum toxin-

A to improve bimanual performance in children with

hemiplegic cerebral palsy: a randomised controlled trial

methods paper. Biomed Cent Neurol 2010; 10: 58–78.

9. Gordon AM, Chinnan A, Gill S, Petra E, Hung YC,

Charles J. Both constraint-induced movement therapy

and bimanual training lead to improved performance of

upper extremity function in children with hemiplegia.

Dev Med Child Neurol 2008; 50: 957–8.

10. Bohannon RW, Smith MB. Interrater reliability of a

modified Ashworth scale of muscle spasticity. Phys Ther

1987; 67: 206–7.

11. Boyd RN, Graham HK. Objective measurement of clini-

cal findings in the use of botulinum toxin type A for the

management of children with cerebral palsy. Eur J Neu-

rol 1999; 6(Suppl. 4): S23–5.

12. Eliasson AC, Krumlinde-Sundholm L, R€osblad B, et al.

The Manual Ability Classification System (MACS) for

children with cerebral palsy: scale development and evi-

dence of validity and reliability. Dev Med Child Neurol

2006; 48: 549–54.

13. Smith RA, Wrisberg CA. Motor Learning and Perfor-

mance: A Situation-Based Learning Approach, 4th edn.

Champaign, Illinois: Human Kinetics, 2008.

14. Krumlinde-Sundholm L, Eliasson AC. Development of

the Assisting Hand Assessment: a Rasch-built measure

intended for children with unilateral upper limb impair-

ments. Scand J Occup Ther 2003; 10: 16–26.

15. Eliasson AC. Improving the use of hands in daily activi-

ties: aspects of the treatment of children with cerebral

palsy. Phys Occup Ther Pediatr 2005; 25: 37–60.

16. Holmefur M, Aarts P, Hoare BJ, Krumlinde-Sundholm

L. Test-retest and alternate forms reliability of the

assisting hand assessment. J Rehabil Med 2009; 41: 886–

91.

17. Krumlinde-Sundholm L. Reporting outcomes of the

assisting hand assessment: what scale should be used?


18. Morris S. Ashworth and Tardieu Scales: their clinical

relevance for measuring spasticity in adult and paediatric

neurological populations. Phys Ther Rev 2002; 7: 53–62.

19. DeMatteo C, Law M, Russell DJ, Pollock N, Rosen-

baum P, Walter S. The reliability and validity of the

Quality of Upper Extremity Skills Test. Phys Occup Ther

Pediatr 1993; 13: 1–18.

20. Thorley M, Lannin N, Cusick A, Novak I, Boyd R.

Construct validity of the Quality of Upper Extremity

Skills Test for children with cerebral palsy. Dev Med

Child Neurol 2012; 54: 1037–43.

21. Haley SM, Coster WJ, Ludlow LH, Haltiwanger J, An-

drellos P. Pediatric Evaluation of Disability Inventory:

Development, Standardization and Administration Man-

ual. Boston: PEDI Research Group, 1992

22. Russo RN, Crotty M, Miller MD, Murchland S, Flett P,

Haan E. Upper limb botulinum toxin A injection and

occupational therapy in children with hemiplegic cere-

bral palsy identified from a population register: a single-

blind randomised controlled trial. Pediatrics 2007; 119:

1149–58.

23. Iyer LV, Haley S, M , Watkins MP, Dumas HM. Estab-

lishing minimal clinically important differences for

scores on the Pediatric Evaluation of Disability Inven-

tory for Inpatient Rehabilitation. Phys Ther 2003; 83:

888–98.

24. Law M, Baptiste S, Carswell A, McColl MA, Polatajko

H, Pollock N. Canadian Occupational Performance

Measure. Ottawa: CAOT Publications, 1998.

25. Carswell A, McColl MA, Baptiste S, Law M, Polatajko

HJ, Pollock N. The Canadian Occupational Perfor-

mance Measure: a research and clinical literature review.

Can J Occup Ther 2004; 71: 210–22.

26. Cusick A, McIntyre S, Novak I, Lannin N, Lowe K. A

comparison of Goal Attainment Scaling and the Cana-

dian Occupational Performance Measure for pediatric

rehabilitation research. Pediatr Rehabil 2006; 9: 149–57.

27. Kiresuk TJ, Smith JE, Cardillo JE. Goal Attainment

Scaling: Applications, Theory and Measurement. Hills-

dale, NJ: Lawrence Erlbaum Associates, 1994.

28. Schulz KF, Altman DG, Moher D. CONSORT 2010

Statement: updated guidelines for reporting parallel

group randomised trials. BMC Med 2010; 8: 18.

29. DeLuca S. Intensive Movement Therapy With Casting

for Children With Hemiparetic Cerebral Palsy: A

Randomised Controlled Trial [Dissertation]. Birming-

ham: University of Alabama at Birmingham, 2002.

30. Eliasson A-C, Shaw K, Berg E, Krumlinde Sundholm L.

An ecological approach of Constraint Induced Move-

ment Therapy for 2–3-year-old children: a randomized

control trial. Res Dev Disabil 2011; 32: 2820–8.

31. Wallen M, Ziviani J, Naylor O, Evans R, Novak I, Her-

bert RD. Modified constraint-induced therapy for chil-

dren with hemiplegic cerebral palsy: a randomized trial.


32. Aarts P, Jongerius P, Geerdink Y, Van Limbeek J, Ge-

urts A. Effectiveness of modified constraint-induced

movement therapy in children with unilateral spastic

cerebral palsy: a randomized controlled trial. Neuroreha-

bil Neural Repair 2010; 24: 509–18.

33. Al-Oraibi S, Eliasson A-C. Implementation of con-

straint-induced movement therapy for young children

with unilateral cerebral palsy in Jordan: a home-based

model. Disabil Rehabil 2011; 33: 2006–12.

34. Gordon AM, Hung Y-C, Brandao M, et al. Bimanual

training and constraint-induced movement therapy in

children with hemiplegic cerebral palsy: a randomized

trial. Neurorehabil Neural Repair 2011; 25: 692–702.

35. Sakzewski L, Ziviani J, Abbott DF, Macdonell RAL,

Jackson GD, Boyd RN. Randomized trial of constraint-

induced movement therapy and bimanual training on

activity outcomes for children with congenital hemiple-

gia. Dev Med Child Neurol 2011; 53: 313–20.

36. Craje C, van ElkM, Beeren M, van Schie HT, Bekkering

H, Steenbergen B. Compromised motor planning and

motor imagery in right hemiparetic cerebral palsy. Res

Dev Disabil 2010; 31: 1313–22.

37. Gordon A, Steenbergen B. Bimanual coordination in

children with cerebral palsy. In: Eliasson A, Burtner PA,

editors. Improving Hand Function in Children With

Cerebral Palsy: Theory, Evidence, and Intervention.

London: Mac Keith Press, 2008: 160–75.


intensive therapy following upper limb botulinum toxin a injection in young children with unilateral...

Documents