DOI: 10.1542/peds.2008-3335 2009;123;e1111-e1122; originally published online May 18, 2009; Pediatrics

Leanne Sakzewski, Jenny Ziviani and Roslyn Boyd Upper-Limb Dysfunction in Children With Congenital Hemiplegia

Systematic Review and Meta-analysis of Therapeutic Management of

http://www.pediatrics.org/cgi/content/full/123/6/e1111located on the World Wide Web at:

The online version of this article, along with updated information and services, is

rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275. Grove Village, Illinois, 60007. Copyright © 2009 by the American Academy of Pediatrics. All and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elkpublication, it has been published continuously since 1948. PEDIATRICS is owned, published, PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

at University of Queensland Library on June 2, 2009 www.pediatrics.orgDownloaded from

REVIEW ARTICLES

Systematic Review and Meta-analysis of TherapeuticManagement of Upper-Limb Dysfunction in ChildrenWith Congenital HemiplegiaLeanne Sakzewski, BOccThya,b, Jenny Ziviani, PhD, MEd, BA, BAppScOTc, Roslyn Boyd, PhD, MSc(Physiotherapy)a,b

aQueensland Cerebral Palsy and Rehabilitation Research Centre, School of Medicine, and cSchool of Health and Rehabilitation Sciences, University of Queensland,Brisbane, Australia; bBrain Research Institute, Melbourne, Victoria, Australia

The authors have indicated they have no financial relationships relevant to this article to disclose.

ABSTRACT

CONTEXT.Rehabilitation for children with congenital hemiplegia to improve function inthe impaired upper limb and enhance participation may be time-consuming andcostly.

OBJECTIVES. To systematically review the efficacy of nonsurgical upper-limb therapeuticinterventions for children with congenital hemiplegia.

METHODS. The Cochrane Central Register of Controlled Trials, Medline, CINAHL (Cu-mulative Index to Nursing and Allied Health Literature), AMED (Allied and Com-plementary Medicine), Embase, PsycINFO, and Web of Science were searched up toJuly 2008. Data sources were randomized or quasi-randomized trials and systematicreviews.

RESULTS. Twelve studies and 7 systematic reviews met our criteria. Trials had strongmethodologic quality (Physiotherapy Evidence Database [PEDro] scale � 5), andsystematic reviews rated strongly (AMSTAR [Assessment of Multiple SystematicReviews] score � 6). Four interventions were identified: intramuscular botulinumtoxin A combined with upper-limb training; constraint-induced movement therapy;hand-arm bimanual intensive training; and neurodevelopmental therapy. Data werepooled for upper-limb, self-care, and individualized outcomes. There were small-to-medium treatment effects favoring intramuscular botulinum toxin A and occupa-tional therapy, neurodevelopmental therapy and casting, constraint-induced move-ment therapy, and hand-arm bimanual intensive training on upper-limb outcomes.There were large treatment effects favoring intramuscular botulinum toxin A andupper-limb training for individualized outcomes. No studies reported participationoutcomes.

CONCLUSIONS.No one treatment approach seems to be superior; however, injections ofbotulinum toxin A provide a supplementary benefit to a variety of upper-limb–training approaches. Additional research is needed to justify more-intensive ap-proaches such as constraint-induced movement therapy and hand-arm bimanualintensive training. Pediatrics 2009;123:e1111–e1122

CEREBRAL PALSY (CP) is the most common physical disability in childhood, occur-ring in 1 in 500 children, and 36% of children with CP have hemiplegia.1

Typically, the upper limb (UL) is more involved than the lower, with impairments ofspasticity, sensation, and reduced strength. Effective use of the arm and hand toreach, grasp, release, and manipulate objects is often compromised. Children withhemiplegia usually have the intellectual capacity to attend regular school; however, impaired arm function restrictstheir participation in educational, leisure, and later vocational roles.

UL rehabilitation to address UL dysfunction uses a wide range of intervention techniques and can be bothtime-consuming and costly. In a 2001 systematic review, Boyd et al2 described a suite of interventions targeting ULdeficits in children with CP. These interventions included behavioral and environmental treatments, physicalinterventions, the use of electrophysical and pharmacologic agents, and surgical management. They concluded thatthere was limited evidence for occupational therapy (OT) and casting with small treatment effect; however, there wasgrowing evidence for injections of botulinum toxin A (BoNT-A) combined with UL training to improving functionin children with CP.

www.pediatrics.org/cgi/doi/10.1542/peds.2008-3335

doi:10.1542/peds.2008-3335

KeyWordscerebral palsy, systematic review, meta-analysis, upper-limb rehabilitation

AbbreviationsCP—cerebral palsyUL—upper limbOT—occupational therapyBoNT-A—botulinum toxin ACIMT—constraint-induced movementtherapyRCT—randomized, controlled trialHABIT—hand-arm bimanual intensivetrainingNDT—neurodevelopmental treatmentPEDro—Physiotherapy Evidence DatabaseAMSTAR—Assessment of MultipleSystematic ReviewsCI—confidence intervalES—effect sizeSMD—standardized mean differenceQUEST—Quality of Upper Extremity SkillsTestPFMS—Peabody Fine Motor ScalesCOPM—Canadian OccupationalPerformance MeasureGAS—Goal Attainment ScalePEDI—Pediatric Evaluation of DisabilityInventoryBOTMP—Bruininks-Oseretsky Test ofMotor Proficiency

Accepted for publication Jan 30, 2009

Address correspondence to Leanne Sakzewski,BOccThy, Brain Research Institute,Neurosciences Building, Austin Health, PO Box5444, Heidelberg West, Victoria 3081,Australia. E-mail: s3144184@student.uq.edu.au

PEDIATRICS (ISSN Numbers: Print, 0031-4005;Online, 1098-4275). Copyright © 2009 by theAmerican Academy of Pediatrics

PEDIATRICS Volume 123, Number 6, June 2009 e1111 at University of Queensland Library on June 2, 2009 www.pediatrics.orgDownloaded from

Another review found a significant treatment effectfor constraint-induced movement therapy (CIMT) inonly 1 trial based on an assessment of bimanual use.3 Apositive trend favoring CIMT was shown in 2 otherstudies, and it was recommended that the approachreceive further investigation.3 Another systematic re-view that investigated UL BoNT-A injections for childrenwith CP was inconclusive.4 Together these reviews high-lighted the need to evaluate UL function by using objec-tive, reliable, and valid measures in future studies. Insubsequent years, results of other randomized, con-trolled trials (RCTs) have been published. In addition,new interventions such as hand-arm bimanual intensivetraining (HABIT)5 have emerged.

Considering the broad range of interventions, relativecost, and time commitment, it is timely to evaluate theefficacy of all therapeutic UL interventions. Of particularinterest is their impact on UL activity performance andhow this might translate into the broader context ofparticipation in home, school, and community life. Theaim of this systematic review was to determine the effi-cacy of UL interventions for children with congenitalhemiplegia (aged 0–18 years) on activity and participa-tion outcomes compared with no intervention or stan-dard care.

METHODS

Search StrategyArticles were identified through comprehensive searchesof computerized bibliographic databases Medline (1950to July 2008), CINAHL (Cumulative Index to Nursingand Allied Health Literature) (1982 to July 2008), Em-base (1988 to July 2008), Cochrane Central Register ofControlled Trials (CENTRAL, The Cochrane Library[r],Issue 3, 2008), AMED (Allied and Complementary Med-icine) (1985 to July 2008), PsycINFO (1967 to July2008), and Web of Science (1945 to July 2008).

The search used exploded Medical Subject Headings(MeSH) terms and text words:

1. “cerebral palsy” or “hemiplegia”

2. and “child” or “adolescent” or “infant”

3. and “physical therapy” or “occupational therapy” or“movement therapy” or “neurodevelopmental ther-apy” or “splinting” or “casting” or “botulinum toxin”or “functional electrical stimulation” or “strengthtraining” or “conductive education” or “virtual real-ity” or “constraint induced movement therapy”

4. and “upper limb” or “arm” or “hand”

5. and “randomized, controlled trial” or “random sam-pling” or “double blind method” or “single blindmethod” or “placebo” or “systematic review.”

Selection CriteriaTo be included in this systematic review, studies had tomeet the following criteria:

1. systematic reviews either with or without meta-analyses and randomized or quasi-randomized,controlled trials;

2. participants included children aged 0 to 18 yearswith congenital hemiplegia with spasticity affectingUL function; and

3. interventions to improve UL function and partici-pation were nonsurgical and included:

3.1. OT, physiotherapy, neurodevelopmental treatment(NDT), movement training, motor learning, con-ductive education, CIMT, strength training, and vir-tual-reality training, and/or;

3.2. splinting (soft or rigid), casting, neuromuscularelectrical stimulation; and/or

3.3. pharmacologic agents including BoNT-A.

Included studies required 50% of the intervention con-tent to focus on the UL.

4. Outcomes included UL activity and/or participationoutcomes.

Studies were excluded if:

1. trials were nonrandomized or had no control group(control group was defined as a group of childrenwho received standard care or no intervention; stan-dard care was defined as OT, physiotherapy, or ULrehabilitation provided at preexisting levels beforethe trial);

2. participants of the study did not include a subset ofchildren with hemiplegia;

3. outcomes reported were impairment only; or

4. studies were not published in English because of alack of translation services.

Eligibility for study inclusion was assessed indepen-dently by 2 reviewers (Ms Sakzewski and Dr Boyd). Theinitial search yield was reviewed by Ms Sakzewski on thebasis of title and abstract. The full text of the articles thatadhered to the inclusion criteria was then assessed byboth reviewers.

Data ExtractionDetails of study methodology and population were sum-marized for all trials (Table 1) and systematic reviews(Table 2). Methodologic quality of the included trialswas evaluated by using the PEDro (Physiotherapy Evi-dence Database) scale (Table 3).6,7 The scale has 11 cri-teria, each scored as 0 or 1, resulting in a possible total of11 points. Interrater reliability of the PEDro scale isadequate.8 Two reviewers (Ms Sakzewski and Dr Boyd)independently assessed methodologic quality. Studieswith scores of �5 were regarded as having adequateinternal validity and, therefore, were included in themeta-analysis.8

Methodologic quality of the included systematic re-views was assessed by using the Assessment of MultipleSystematic Reviews (AMSTAR) (Table 4).9 A score of 1was recorded for each criterion present, with a totalpossible score of 11. The AMSTAR has good face andcontent validity.9

e1112 SAKZEWSKI et al at University of Queensland Library on June 2, 2009 www.pediatrics.orgDownloaded from

Data SynthesisRevMan 5.0 (Cochrane Collaboration, Oxford, England)was used for data management and analysis. Outcomeswith continuous measurement scales were summarized foreach study by using means, effect sizes (ESs), and 95%confidence intervals (CIs) for treatment and controlgroups. ESs of 0.2 were considered small, 0.4 to 0.6 mod-erate, and 0.8 large.10 For meta-analysis, for comparisonbetween treatment and control groups for continuousmeasurement scales, standardized mean differences(SMDs) with 95% CIs were calculated. A fixed-effectmodel was used to pool treatment effects when trials usedsimilar outcome measures on similar populations.11 Heter-

ogeneity was evaluated by using the I2 statistic, with largerpercentage scores representing greater heterogeneity.12

RESULTS

Description of StudiesThe search strategy yielded 177 unique references.Twenty RCTs and 8 systematic reviews of therapeutic ULinterventions met inclusion criteria and were examinedby 2 independent reviewers.2–5,13–36 Seven trials and 1systematic review were excluded13,31–36; reasons for theexclusion are listed in Fig 1. Two articles reported thesame study.22,24

TABLE 2 Characteristics andMethods of Included Systematic Reviews

Review Intervention Population Included Studies Meta-analysis

Clinical Inference

Boyd et al2 Any UL intervention Spastic CP RCTs (n � 4), non-RCTs (n � 44) � Best evidence for OT and casting, smalltreatment effects; growing evidencefor BoNT-A to improve UL function

Brown andBurns26

NDT Neurological diagnosis(0–18 y)

RCTs (n � 17) Inconclusive; 9 studies reported nobenefit, 6 reported benefit of NDT;significant group variability acrossstudies and use of differentoutcome measures

Wasiak et al4 BoNT-A CP (0–19 y) RCTs (n � 2) Insufficient evidence to support use ofBoNT-A

Steultjenset al27

OT CP (�19 y) Efficacy studies (RCTs, pre-postor time series)

Inconclusive, reflecting variability ofinterventions, outcome measures,and poor methodologic quality ofincluded studies

Autti-Ramoet al28

UL and lower-limbcasting/orthoses

CP Systematic reviews (n � 5) Inconclusive evidence to support ULcasting and splinting

Reeuwijket al29

BoNT-A CP All Insufficient evidence for effect of Botoxon UL function

Hoare et al3 CIMT, forced-use therapy Hemiplegia (0–19 y) RCTs, controlled clinical trials(n � 3)

Significant treatment effect in singletrial (Eliasson et al33) and positivetrend in 2 others; concluded furtherevidence is required and CIMTshould remain in clinical trials

TABLE 1 Study Characteristics andMethods of RCTs

Study Design Diagnosis Age Treatment n Control n

Law et al (a)14 RCT CP 18 mo to 8 y Intense NDT and casting 19 Intensive NDT 18Law et al (b)14 RCT CP 18 mo to 8 y Regular NDT and casting 17 Regular NDT 18Law et al (a)15 RCT-crossover CP 18 mo to 4 y Intense NDT and casting first 26 Regular OT 24Law et al (b)16 RCT- crossover CP 18 mo to 4 y Intense NDT and casting second 26 Regular OT 24Corry et al16 DB RCT Hemi CP � 2CPa 4 to 19 y BoNT-A 7 Control 7Fehlings et al17 SB RCT Hemi CP 2 to 10 y BoNT-A and OT 14 OT 15Speth et al18 Matched-pairs RCT Hemi CP 4 to 16 y BoNT-A and OT/PT 10 OT/PT 10Lowe et al19 RCT Hemi CP 2 to 8 y BoNT-A and OT 21 OT 21Wallen et al (a)20 SB RCT CP 2 to 14 y BoNT-A 19 Control 15Wallen et al (b)20 SB RCT CP 2 to 14 y BoNT-A and OT 20 OT 17Russo et al21 SB RCT Hemi CP 3 to 16 y BoNT-A and OT 21 OT 22Taub et al22 RCT CP 7 mo to 8 y CIMT 9 Regular therapy 9Sung et al23 RCT Hemi CP �8 y FU and regular therapy 18 Regular therapy 13Deluca et al24 RCT crossover CP 7 mo to 8 y CIMT 9 Control 9Charles et al25 SB RCT Hemi CP 4 to 8 y CIMT 11 Control 11Gordon et al5 SB RCT Hemi CP 3 to 15 y HABIT 10 Regular therapy 10

Hemi indicates hemiplegia; DB, double-blind; SB, single-blind; PT, physiotherapy; FU, forced use.a 2CP indicates two other subtypes of CP other than hemiplegia. (1 triplegia, 1 quadriplex.)

PEDIATRICS Volume 123, Number 6, June 2009 e1113 at University of Queensland Library on June 2, 2009 www.pediatrics.orgDownloaded from

Study ParticipantsSeven trials specifically targeted 222 children with con-genital hemiplegia.5,17–19,21,25 Five studies had a broadersample of 176 children with CP that included a subset ofchildren with hemiplegia.14,15,20,22,24 One study consistedpredominantly of children with hemiplegia and 2 sub-jects with different distributions of CP.16 The age range ofchildren differed across trials, ranging from 18 months to16 years.

Types of InterventionFour broad categories of intervention were identified:NDT,14,15 CIMT22,24,25 and forced-use therapy,23 HABIT,5

and intramuscular BoNT-A injections to the UL with orwithout therapy.16–21 Details of the intervention pro-grams are summarized in Table 5.

Five identified systematic reviews included some tri-als identified in this current review. Two systematic re-views evaluated BoNT-A injections to the UL4,29. The firstincluded 2 RCTs16,17 and the second reported on 3RCTs.16–18 A systematic review of CIMT3 identified 1 RCTand 1 quasi-RCT of 49 subjects.23,24 A systematic reviewevaluating OT28 identified 2 RCTs of NDT14,15 included inthis review. Further review of systematic reviews29 iden-tified the same 2 RCTs evaluating UL casting14,15 but didnot find any RCTs of UL splinting. Our review has iden-tified another 5 RCTs across UL interventions.

Study DesignAll included studies were RCTs that compared an inter-vention to a control group that received either no inter-

FIGURE 1Results of search strategy of UL systematic review. a Two articles reported the samestudy.22,24

TABLE 3 Methodologic Quality Assessment of Included RCTs: PEDro Scale

Study 1 2 3 4 5 6 7 8 9 10 11 Total

Law et al14 (1991) 1 1 0 1 0 0 1 1 1 1 1 8Law et al15 (1997) 1 1 0 1 0 0 1 1 1 1 1 8Corry et al16 (1997) 0 1 1 1 1 1 1 1 1 1 1 10Fehlings et al17 (2000) 1 1 0 1 0 0 0 1 1 1 1 7Speth et al18 (2005) 1 1 1 0 0 0 1 1 1 1 1 8Lowe et al19 (2006) 1 1 1 1 0 0 1 1 1 1 1 9Wallen et al20 (2007) 1 1 1 0 0 0 1 1 1 1 1 8Russo et al21 (2007) 1 1 1 1 0 1 1 1 1 1 1 10Taub et al22 (2004) 1 1 1 1 0 0 1 1 1 1 1 9Sung et al23 (2005) 1 1 0 1 0 0 0 1 1 1 1 7Deluca et al24 (2006) 1 1 0 1 0 0 1 1 1 1 0 7Charles et al25 (2006) 1 1 0 0 0 0 1 0 0 1 1 5Gordon et al5 (2007) 1 1 1 0 0 0 1 1 1 1 1 8

Scaleof itemscore: 0, absent; 1, present. ThePEDro scale criteria are (1) specifiedeligibility criteria; (2) randomallocation; (3) concealedallocation; (4) similarity atbaselineonkeymeasures;(5) subject blinding; (6) therapist blinding; (7) assessor blinding; (8)�85% follow-up of at least 1 key outcome; (9) intention-to-treat analysis; (10) between-group statistical comparisonfor at least 1 key outcome; and (11) point estimates and measures of variability provided for at least 1 key outcome.

TABLE 4 Methodologic Quality of Included Systematic Reviews: AMSTAR

Review 1 2 3 4 5 6 7 8 9 10 11 Total

Boyd et al2 (2001) 1 0 1 1 0 1 1 1 1 0 0 7Brown and Burns26 (2001) 1 1 1 1 0 1 1 1 0 1 0 8Wasiak et al4 (2004) 1 1 1 1 1 0 1 1 0 0 1 8Steultjens et al27 (2004) 1 1 1 0 1 1 1 1 0 0 0 7Autti-Ramo et al28 (2006) 1 1 1 0 1 1 1 1 0 0 0 7Reeuwijk et al29 (2006) 1 1 0 0 1 1 1 1 0 0 0 6Hoare et al3 (2007) 1 1 1 1 1 1 1 1 1 0 1 10

Scale of item score: 0, absent; 1, present. The AMSTAR criteria are (1) a priori design; (2) duplicate study selection and data extraction; (3) comprehensive literature search; (4) inclusivepublication status; (5) included/excluded studies provided; (6) characteristics of included studies provided; (7) quality assessment of studies; (8) study quality used appropriately informulating conclusions; (9) appropriate methods used to combine studies; (10) publication bias assessed; and (11) conflict of interest stated.

e1114 SAKZEWSKI et al at University of Queensland Library on June 2, 2009 www.pediatrics.orgDownloaded from

TABL

E5

Structurean

dCo

nten

tofU

L-Interven

tion

Prog

rams

Stud

yCo

nten

tofInterventionProg

ram

Duration,

mo

Freq

uency,No.of

Sessions

Intensity

Hom

eProg

ram

Conten

tofC

ontrol

Group

Duration,

mo

Freq

uency,

No.of

Sessions

Intensity

Hom

eProg

ram

Lawetal(a)14

IntensiveNDT

62/wk

45min

30min/d

IntensiveNDTon

ly6

2/wk

4530

min/d

Bivalved

cast

7/wk

240min/d

Lawetal(b)14

Regu

larN

DT

61/wk(m

aximum

)to

1/mo(m

inimum

)45

min

15min

Regu

larN

DTon

ly6

1/wk(m

aximum

)to

1/mo(m

inimum

)45

15min

Bivalved

cast

240min/d

3/wk

3/wk

Lawetal(a)15

NDT

42min/w

k45

min

30min/d

Regu

larO

T-functio

nal

skills

41/wk(m

aximum

)to

1/mo(m

inimum

)45

Previous

therapy

Bivalved

cast

2/wk

240min/d

Lawetal(b)15

NDTandcastingsecond

4a/a

a/a

a/a

Regu

larO

T(a/a)

4a/a

a/a

a/a

Corryetal

16Bo

toxinjections

(n�

10):dilutio

n,1

mL/100U;m

aximum

dose,4–7

U/kg;totaldose,90–250

U;

Dyspo

rt(n

�4):dilutio

n,2.5mL/

500U;dose,8–9U/kg;totaldose,

160–400U

NR

NR

NR

NR

Placebo(saline)

NR

NR

NR

NR

Fehlings

etal

17

Botoxinjections:dose,2–6U/kg;OT:

streng

th,ADLs

60.5/wk

NR

NR

OT(asp

erinterventio

ngrou

p)

NR

0.5/wk

NR

NR

Spethetal

18Bo

toxinjections:dilutio

n,5U/0.1

mL;maximum

dose,400

U/kg;

OT:streng

th,FT,orthoses

63/wkeach

OTand

PTOT:30

min;

PT:30min

NR

OT(asp

erinterventio

ngrou

p)

63/wk

OT:30

min;

PT:30

min

NR

Loweetal

19Bo

toxinjections:dilutio

n,100U/0.5

mL;maximum

dose,�

8U/kg;

totaldose,82–220

U;O

T:go

als,

splintin

g,streng

th,FT

6OTno

treported

NR

Repo

rted

with

out

detail

OT(asp

erinterventio

ngrou

p)

6NR

NR

Repo

rted

with

outd

etail

Wallenetal

(a)20

Botoxinjections:dose,0.5–2U/kg;

maximum

dose,12U/kg;regu

lar

therapy

6Maintainedprevious

level

Previous

level

NR

Controlgroup

6Maintainedprevious

level

NR

Wallenetal

(b)20

Botoxinjections:dose,0.5–2U/kg;

maximum

dose,12U/kg;OT:

stretching

,castin

g,motor

training

,FT

31/wk

60min

NR

OTas

describ

edfor

interventio

ngrou

p3

1/wk

60NR

Russoetal

21Bo

toxinjections:dilutio

n,100U/m

L;maximum

dose,�

12U/kg;total

dose,300

U;O

T:ULWB,ballskills,

streng

th,FT

11/wk

60min

NR

OTas

describ

edfor

interventio

ngrou

p1

1/wk

60NR

Taub

etal

22

andDeluca

etal

24

Long

armcast,bivalved,andworn

durin

gCIMT;CIMT:reach,grasp,

manipulate,FT

0.7

7/wk

360min

NR

Maintainedprevious

levels

0.7

1session/wk

(minimum

)to4

sessions/w

k(m

aximum

)

132(m

ean)

NR

Sung

etal

23Shortarm

castapplied;OT:stretch,

reach,grasp,manipulate,FT

1.5

2/wk

0.5min

NR

OTas

perintervention

grou

p1.5

2/wk

30NR

Charlese

tal25

Slingwornon

unimpairedarm;

grou

psof2–4;CIMT-movem

ent

training

,play,FT

0.3

5/wk

360min

1/d2/dafter

treatm

entfor1

mo

Maintainedprevious

levels

NR

NR

NR

NR

Gordo

netal

5HAB

IT:group

sof4

children;

bimanualFMandmanipulative

GMactivities

0.3

5/wk

360min

1h/d;2h/dafter

treatm

entfor1

mo

Maintainedprevious

levels

NR

NR

NR

NR

a/aindicatesasabo

ve;NR,no

treported;PT,physiotherapy;units/kg,un

itsofBo

NT-Aperkgofbo

dyweigh

t;AD

Ls,activitiesofdaily

living;WB,weigh

tbearin

g;FT,fun

ctionaltraining.Bo

toxwasprovided

byAllergan

PLC(Irvine,CA);Dyspo

rtwasprovided

byIpsenLimited(Sloug

h,UnitedKing

dom).

PEDIATRICS Volume 123, Number 6, June 2009 e1115 at University of Queensland Library on June 2, 2009 www.pediatrics.orgDownloaded from

vention or standard care throughout the duration of thetrial. Two of the studies were crossover RCTs wherebythe subjects received the comparative or control inter-vention after a period of time.15,24

Types of OutcomeMeasuresA range of standardized and nonstandardized assess-ments was used across the studies. These could begrouped into assessments of UL function, self-care, andindividualized outcomes. No studies reported outcomesmeasuring participation.

Adverse EventsMinor adverse events included nausea and vomiting20,21

and transient weakness after BoNT-A injections.16,17,21

These events were short-acting and reversible. Minorskin irritations as a result of casting were reported in 3cases of CIMT, along with behavioral problems that pre-existed the intervention.24 In general, there was highretention within studies; however, Wallen et al20 andLaw et al14 reported 6 of 80 and 6 of 79 cases, respec-tively, that did not proceed because of dissatisfactionwith group allocation.

Qualitative AssessmentAll included trials had moderate-to-high quality, scoring�5 on the PEDro scale (Table 1). Three studies did nothave adequate data for meta-analysis.16,18,24 One studyreported median scores,18 1 reported median scores onnonstandardized assessments,16 and 1 reported summary

TABLE 6 Summary of Results of Studies That Reported on UL Outcomes

Study Outcome Timing,wk

N Treatment, Mean(�SD)

N Control, Mean(�SD)

SMD (95% CI) P

Law et al (a)14 PFMS 24 19 35.4 (13.9) 18 28.1 (18.4) 0.43 (�0.21 to 1.09) .03QUEST 66.8 (23) 18 47.9 (26.8) 0.74 (0.08 to 1.41)

Law et al (b)14 PFMS 24 17 33.7 (20.1) 18 30.8 (21.3) 0.13 (�0.53 to 0.80)QUEST 50.9 (25.7) 47.2 (28.9) 0.13 (�0.53 to 0.80)

Law et al (a)15 PFMS 16 26 21.8 (8.5) 24 20.9 (9.0) 0.10 (�0.45 to 0.66)QUEST 53.3 (22.9) 47.3 (27.7) 0.23 (�0.32 to 0.79)

Law et al (b)15 PFMS 40 26 24.7 (13.4) 24 24.9 (12.3) �0.02 (�0.57 to 0.54)QUEST 53.3 (25.1) 40 (24.4) 0.53 (�0.04 to 1.09)

Fehlings et al17 QUEST 4 14 32.54 (17.8) 15 27.6 (19.0) 0.26 (�0.47 to 0.99)12 28.5 (20.2) 30.4 (19.6)24 30.7 (18.8) 34.4 (24.4)

Lowe et al19 QUEST Post 21 43.9 (15.1) 21 36 (12.4) 0.55 (�0.07 to 1.17)12 46.2 (16) 37.1 (11.9)24 40.7 (14.7) 39.6(12.8)

Wallen et al (a)20 MUUL 12 13 63.69 (20.9) 9 61.4 (21.2) 0.18 (�0.67 to 1.03)24 64.26 (24.2) 58.7 (23.8)

QUEST 12 7 67.5 (17.4) 6 30.6 (35) 2.12 (0.76 to 3.48) .00024 62.1 (23.6) 30.6 (30.4)

Wallen et al(b)20 MUUL 12 11 57.4 (24.8) 11 63.5 (29.0) �0.22 (�1.06 to 0.62)24 58.0 (23.4) 64.8 (30.0)

QUEST 12 9 34.5 (38) 6 39.4 (20.6) �0.14 (�1.18, 0.89)24 28.3 (32.8) 7 36.7 (31.7)

Sung et al23 EDPT Post 18 7.6 (1.7) 13 7.0 (1.4) 0.38 (�0.34 to 1.10)Box and Block 10.5 (5.7) 8.5 (7.3) 0.30 (�0.41 to 1.02)

Taub et al22 and Deluca et al24 PMAL-amt Post 9 2.8 (1.1) 9 1.2 (0.8) 1.54 (0.49 to 2.60) .0043 2.6 (1.3) 1.9 (0.7)

PMAL-qual Post 2.7 (1.0) 1.8 (1.1) 0.73 (�0.23 to 1.68)3 2.6 (1.3) 1.8 (1.0)

EBS Post 21.5 (4.5) 15 (5.7) 1.22 (0.22 to 2.23) .02Charles et al25 Jebsen 1 11 278.5 (240.6) 11 301 (182.2) �0.10 (�0.94 to 0.73)

4 268.6 (238) 260.3 (153)24 272.5 (236.6) 297 (200)

BOTMP 1 7.2 (2.9) 5.2 (4.2) 0.53 (�0.32 to 1.38)4 7.6 (4.4) 5.5 (4.1)24 6.9 (3.7) 6.3 (5.1)

Gordon et al5 Jebsen Post 10 380.4 (182.9) 10 285.1 (230.1) 0.44 (�0.45 to 1.33)4 410.1 (155.7) 364.1 (151.3)

BOTMP Post 5.6 (3.6) 8.4 (5.2) �0.61 (�1.51 to 0.29)4 7.1 (4.7) 8.7 (5.6)

AHA Post 1.5 (1.8) 1.2 (2.1) 0.16 (�0.74 to 1.06)4 0.95 (1.7) 1.8 (2.0)

MUUL indicates Melbourne Assessment of Unilateral Upper Limb Function; EDPT, Erhardt Developmental Prehension Test; PMAL-amt, Pediatric Motor Activity Log, amount of use;PMAL-qual, Pediatric Motor Activity Log, quality of use; EBS, Emerging Behavior Scale; Jebsen, Jebsen Taylor Test of Hand Function; AHA, Assisting Hand Assessment; Post, immediatelyfollowing intervention.

e1116 SAKZEWSKI et al at University of Queensland Library on June 2, 2009 www.pediatrics.orgDownloaded from

statistics.24 Quality scores of the RCTs are presented inTable 3, and systematic reviews are shown in Table 4.

Primary Outcome: UL FunctionResults of studies reporting on UL outcomes are sum-marized in Table 6. All meta-analyses are summarized inFigs 2 and 3. Two outcome measures for UL function hadsufficient data to be pooled for meta-analysis in trials ofNDT and intramuscular BoNT-A. Children who receivedNDT and casting (n � 88) compared with a control groupthat received regular OT or regular NDT (n � 84) scoredan SMD of 0.4 (95% CI: 0.10 to 0.71; P � .009) on theQuality of Upper Extremity Skills Test (QUEST).14,15 Thesame groups were compared on the Peabody Fine MotorScales (PFMS) and achieved an SMD (fixed) of 0.15(95% CI: �0.15 to 0.45; P � .34).

Children who received BoNT-A injections and ULtraining (n � 51) compared with a control group thatreceived UL training (n � 48) achieved an SMD (fixed)of 0.41 (95% CI: 0.00 to 0.81; P � .05) on theQUEST.17,19,20

Achievement of Individualized GoalsResults of studies reporting on individualized outcomesare summarized in Table 7. Data pooled from 2 stud-ies19,20 that compared BoNT-A and OT (n � 60) andcontrol groups receiving OT or regular care (n � 53)received an SMD of 0.44 (95% CI: �0.11 to 0.99; P �.12) for Canadian Occupational Performance Measure(COPM) performance and an SMD of 0.48 (95% CI: 0.11to 0.86; P � .01) for COPM satisfaction. Goal Attainment

Scale (GAS) scores were pooled from 3 studies thatcompared BoNT-A and OT (n � 81) and control groupsthat received OT or regular care (n � 75) with an SMDof 0.95 (95% CI: 0.62 to 1.29; P � .0001).19–21

Self-care OutcomesResults of studies reporting self care outcomes are sum-marized in Table 8. Pediatric Evaluation of DisabilityInventory (PEDI) self-care scores were pooled from 2studies20,21 that compared BoNT-A and OT (n � 60) witha control group that received OT (n � 54) with an SMDof �0.03 (95% CI: �0.74 to 0.68; P � .94).

DISCUSSIONWith this systematic review we aimed to evaluate theefficacy of nonsurgical therapeutic UL interventionscompared with no treatment or standard care on activityand participation outcomes for children with congenitalhemiplegia. Thirteen RCTs and 7 systematic reviews metthe a priori inclusion criteria.

Four intervention approaches were identified: (1)NDT; (2) UL BoNT-A injections combined with training;(3) CIMT; and (4) HABIT. Two of the approachesbroadly target deficits of body structure and function(NDT and BoNT-A), and 2 focus on improving activitylimitations (CIMT and HABIT). NDT remains a com-monly used approach that aims to improve motor skillsby reducing underlying impairments and facilitation ofmore normal movement patterns.37,38 Meta-analyses ofNDT and casting yielded a medium treatment effect forquality of UL movement (QUEST) and a small effect to

TABLE 7 Summary of Results of Studies That Reported on Individualized Outcomes

Study Outcome Timing,wk

N Treatment, Mean(�SD)

N Control, Mean(�SD)

SMD (95% CI) P

Lowe et al19 COPM-perf Post 21 4.5 (0.9) 21 3.8 (1.4) 0.57 (�0.04 to 1.19)12 5.3 (1.4) 4.5 (1.4)24 5.9 (1.4) 5.1 (2.3)

COPM–sat Post 5.1 (1.4) 4.1 (1.8) 0.60 (�0.02 to 1.21) .0512 5.8 (1.4) 4.7 (1.8)24 6.2 (1.4) 5.4 (3)

GAS-family Post 36.1 (10.1) 27.1 (6.4) 1.02 (0.38 to 1.67) .0012 42 (10.1) 34.1 (9.2)24 46.8 (10.5) 40.1 (13.3)

GAS- ther Post 57.8 (13.8) 40.5 (11.9) 1.29 (0.63 to 1.96) .0012 61 (17.4) 46.8 (12.4)24 58.7 (15.6) 49.9 (12.4)

Wallen et al (a)20 COPM-perf 12 19 5.6 (1.4) 15 4.4 (1.3) 0.66 (�0.03 to 1.36) .0524 5.9 (1.8) 5.1 (1.6)

COPM–sat 12 6.5 (1.7) 5.4 (1.9) 0.11 (�0.57 to 0.79)24 6.8 (1.8) 6.3 (1.9)

GAS 12 42.3 (13.7) 32.9 (10.3) 0.79 (0.08 to 1.49) .0224 52.5 (13.3) 40.6 (12.0)

Wallen et al (b)20 COPM–perf 12 20 5.4 (2.1) 17 5.6 (1.8) �0.1 (�0.75 to 0.55)24 5.8 (2) 6.2 (1.8)

COPM–sat 12 6.6 (2.1) 6.1 (1.9) 0.24 (�0.41 to 0.89)24 6.6 (1.7) 6.9 (2.1)

GAS 12 51 (12.3) 42.2 (10.6) 0.98 (0.29 to 1.66) .0024 51.7 (13.3) 51.4 (11.1)

Russo et al21 GAS Post 21 44.6 (14.9) 22 31.6 (10.7) 0.97 (0.34 to 1.60) .00

perf indicates performance; sat, satisfaction; ther, therapist; Post, immediately following intervention.

PEDIATRICS Volume 123, Number 6, June 2009 e1117 at University of Queensland Library on June 2, 2009 www.pediatrics.orgDownloaded from

improve acquisition of fine motor skills (PFMS).14,15 TheQUEST is predominantly an impairment-based measurewith a small number of items that address activity per-formance; therefore, improvements in results may re-flect gains in range of motion and biomechanical align-ment as a result of casting. The 2 studies targeteddifferent age groups; therefore, varied results may sug-gest that older children could have a greater response tocasting and intensive NDT than younger children. Indi-vidualized goals (COPM) demonstrated the greatestchange with NDT,15 which could suggest that UL train-ing, regardless of type or intensity, directed towardachievement of goals that are meaningful to the childand family may result in significant changes in parentalperception of their child’s performance.

Another approach that targets body structure andfunction is BoNT-A injections. However, when used asan adjunct to UL training, activity-level change was ad-dressed. Six RCTs investigated UL BoNT-A injections.With the exception of the double-blind RCT study byCorry et al16 and a comparison by Wallen et al,20 theremaining studies combined BoNT-A injections with ULtraining. The therapy accompanying BoNT-A injectionsvaried across trials for duration (4 weeks to 6 months)and frequency (fortnightly to 3 times per week), giving arange in total dose of 4 to 12 hours. Data pooled from 3studies yielded only a small treatment effect for BoNT-Aand UL training compared with a control group to im-prove quality of movement (QUEST). This is not surpris-ing, considering the control groups received UL trainingthat was equivalent to that of the intervention groups in3 studies.17,19,20 These results suggest that UL trainingalone is effective in improving quality of UL movement,but BoNT-A provides a consistent supplementary effectto UL training.

The efficacy of BoNT-A in achieving individualizedoutcomes was moderately supported when using a val-idated tool, the COPM, and strongly supported with theGAS. A sizeable treatment effect favoring BoNT-A wasevident in 3 studies that reported GAS outcomes.19–21

Pooled data from 3 studies achieved a moderate treat-ment effect for BoNT-A (n � 60) compared with thecontrol group (n � 53) on COPM performance andsatisfaction.19,20

Two other interventions focus on either a unimanualor bimanual approach to UL training. With CIMT, avariety of constraints to the impaired limb (eg, plastercast, mitt, sling, and manual restraint) are used. Theaccompanying intensive training of the impaired limbinvolves either shaping of targeted UL movements25 orentire task practice, which is activity based.33 Alterna-tively, HABIT mirrors the intensity of CIMT but targetsintensive training of bimanual functional activities. BothCIMT and HABIT are underpinned by theories of motorlearning and neuroplasticity; however, one relies onfacilitating use-dependent reorganization due to inten-sive training of the impaired limb, and the other usesbimanual coordination and the unaffected hand tomodel motor control in the impaired hand.22,39

CIMT has generally been reported as a high-intensityintervention with a total dosage of intervention rangingfrom 60 to 126 hours. Alternatively, forced-use therapyinvolves immobilization of the unimpaired arm in a castfor 6 weeks accompanied by a total dose of 6 hours oftherapy.23 Preliminary results could suggest that CIMTnot only contributes to improved efficiency and qualityof UL movement but also increases spontaneous use ofthe involved limb in daily activity. Modified CIMT re-sulted in small treatment effect for improvements in ULmovement efficiency (Jebsen Taylor Test of Hand Func-tion) and moderate ESs to improve UL speed and dex-terity (Bruininks-Oseretsky Test of Motor Proficiency[BOTMP]).25 CIMT resulted in a large treatment effectfor the development of new UL motor skills and anincreased amount of use of the impaired limb.22 How-ever, these results should be viewed with caution, be-cause many of the outcome measures used (eg, PediatricMotor Activity Log, Emerging Behavior Scale, CaregiverFunctional Use Survey, and selected items of theBOTMP) have no reported reliability or validity in chil-dren with congenital hemiplegia.40 Furthermore, controlgroups in the studies continued with varying types ofstandard care. It is unclear whether improvement re-ported for the intervention group was product of CIMT orreflected a response to a higher intensity of intervention.

The 1 RCT that investigated HABIT showed a smalltreatment ES for bimanual performance when using avalidated tool, the Assisting Hand Assessment.5 How-

TABLE 8 Summary of Results of Studies Reporting on Self-care Outcomes

Study Outcome Timingwk

N Treatment Mean(�SD)

N Control Mean(�SD)

SMD (95% CI) P

Lowe et al19 PEDI self-care Post 21 53.1 (11.5) 21 44.2 (13.3) 0.69 (0.07 to 1.31) .0312 55.8 (11.5) 48.3 (11)24 40.7 (14.7) 51.1 (11.9)

Wallen et al (a)20 PEDI self-care 12 19 66.7 (12.7) 15 13 55.0 (18.2) 0.87 (0.16 to 1.57) .0124 63.2 (15.5) 58.8 (21.7)

Wallen et al (b)20 PEDI self-care 12 20 52.0 (14.5) 17 59.1 (17.7) �0.46 (�1.11 to 0.2)24 52.9 (16.3) 59.7 (17.2)

Russo et al21 PEDI self-care Post 21 54.8 (14.5) 22 59.7 (12.7) �0.35 (�0.95 to 0.26)AMPS-motor 0.5 (0.8) 0.72 (0.55) �0.32 (�0.92 to 0.29)AMPS-process 0.37 (0.9) 0.51 (0.66) �0.17 (�0.77 to 0.42)

Sung et al23 WeeFIM (self-care) Post 18 25.4 (5.8) 13 21.2 (8.7) 0.57 (�0.15 to 1.30)

AMPS indicates Assessment of Motor and Process Skills; WeeFIM, Functional Independence Measure for Children; Post, immediately following intervention.

e1118 SAKZEWSKI et al at University of Queensland Library on June 2, 2009 www.pediatrics.orgDownloaded from

ever, there were no improvements in unimanual speedand dexterity. For CIMT and HABIT, small samples andmajor baseline differences between groups are majorlimitations. Furthermore, it is unclear whether the con-straint is essential and what type and intensity of train-ing will best achieve improvements in activity limita-tions. Additional research is required to differentiate theeffects of an intensive unimanual approach (eg, CIMT)compared with a bimanual approach (eg, HABIT).

Acceptability and feasibility of any intervention is animportant consideration. A number of minor adverseevents were reported for BoNT-A that were short-actingand reversible. Very few adverse events were reportedfor CIMT or forced-use therapy. It could be reasonablyexpected that restraint of an unimpaired limb for suchextended time might lead to frustration and have behav-ioral consequences. Most studies reported that partici-pants adapted quickly to constraint with minimal diffi-culties encountered, with the exception of 1 child withbehavior difficulties.24

Adherence to intervention can be considered in termsof attendance at face-to-face sessions in addition to com-pletion of home programs. In general, attendance atNDT and casting sessions was relatively high and variedwith a mean of twenty-four 45-minute sessions of apossible 32 sessions (75%).15 In addition, home pro-grams were expected to be performed for 30 minutes/day over the 120-day intervention period. However,parents fully completed the 30-minute program, on av-erage, for 40 (40%) of the possible 120 days or partiallycompleted the program, on average, for 22 (18%) of 120days. Adherence to home programs in CIMT was 57%during the intervention period and 52% in the monthafter intervention,25 with similar rates reported forHABIT.26 Not all studies reported whether home pro-grams were provided or adhered to. Home programsprovided an additional dosage of intervention apart fromthe face-to-face sessions, and the relatively small rates ofcompletion may raise the question of whether an ade-quate dosage of intervention was received. It was sug-

gested recently that completion of home programs maydepend on a goal-directed approach or on the durationof intervention.41

A number of limitations exist with the current ev-idence for UL interventions. Small sample size in anumber of studies has contributed to baseline inequal-ity between intervention and control groups in 4 stud-ies.5,18,20,25 Interpretation of ESs for these studies, there-fore, needs to be considered with caution. Variableprotocols of treatment intensity and duration acrossstudies make comparison both between interventionsand across the differing models of intervention difficult.Inconsistent use of outcome measures across trials or useof assessments with no reported validity or reliability forchildren with CP is problematic. Evaluation using indi-vidualized outcome measures is useful, given the heter-ogenous nature of the population, and should be con-sidered in future UL intervention trials. Additional RCTsof adequate power are required particularly for CIMTand HABIT using reliable and valid instrumentation.

In this review we have highlighted that the strengthof any 1 approach over the others is not clearly evident.There are promising consistent effects of BoNT-A sup-plementing various UL training approaches and interest-ing effects for CIMT and HABIT. Pooling data from largerRCTs could give an indication of the best responders (eg,impact of age, sensory deficits, differences between uni-manual capacity and bimanual performance at baseline).Furthermore, the minimum dosage required to gain atreatment effect for any of the reviewed interventionsremains unclear. In addition, the type of training accom-panying BoNT-A needs to be investigated further. As yet,no intensive training approach has demonstrated suffi-cient additional treatment effect over and above-men-tioned supplementary injections of BoNT-A and ULtraining.

RESEARCH IMPLICATIONSThe relationship between the nature of the structuralbrain lesion or corticospinal (re)organization in CP and

FIGURE 2Meta-analyses of the effect of intensive NDT and casting versus NDT or regular OT. A, Results of UL quality of movement: QUEST; B, results of fine motor development: PFMS.

PEDIATRICS Volume 123, Number 6, June 2009 e1119 at University of Queensland Library on June 2, 2009 www.pediatrics.orgDownloaded from

the response to UL interventions for children with con-genital hemiplegia has only been reported in 1 prospec-tive study.42 Additional RCTs investigating this relation-ship are required. The characteristics of best respondersto the different interventions have not been reported todate.

CONCLUSIONSInterventions that address UL dysfunction in childrenwith hemiplegia are varied. There is growing evidencethat BoNT-A supplements UL, training particularly to-ward the achievement of individualized goals. Thereremain limited high-level studies of NDT. CIMT andHABIT are emerging interventions; however, additionalRCTs are required. Effective use of the UL affects on allaspects of daily life. Additional, suitably powered RCTsusing valid and reliable outcome measures are requiredto determine the efficacy of UL interventions. Investiga-tion of the minimum dosage of interventions to exact a

treatment effect is required to optimize the investmentof time and resources for service providers and for chil-dren with congenital hemiplegia and their families.

ACKNOWLEDGMENTS

This work was supported by National Health and Med-ical Research Council (NHMRC) Dora Lush postgraduatescholarship 384488 (to Ms Sakzewski), career develop-ment grant NHMRC 473860 (to Dr Boyd), and projectgrant (INCITE: A Randomised Trial of Novel Upper LimbRehabilitation in Children With Congenital Hemiplegia)NHMRC 368500.

We thank Andrew Gordon, Margaret Wallen, andIona Novak for sharing data for this review.

REFERENCES1. Stanley F, Blair E, Alberman E. Cerebral Palsies: Epidemiology and

Causal Pathways. London, United Kingdom: MacKeith Press; 2000

FIGURE 3Meta-analyses of the effect of BoNT-A and UL training versus UL training. A, Results of UL quality of movement: QUEST; B, results of individualized outcomes: COPM performance; C,results of individualized outcomes: COPM satisfaction; D, results of individualized outcomes: GAS; E, results of self-care skills: PEDI, self-care domain.

e1120 SAKZEWSKI et al at University of Queensland Library on June 2, 2009 www.pediatrics.orgDownloaded from

2. Boyd RN, Morris ME, Graham HK. Management of upper limbdysfunction in children with cerebral palsy: a systematic re-view. Eur J Neurol. 2001;8(suppl 5):150–166

3. Hoare B, Wasiak J, Imms C, Carey L. Constraint-inducedmovement therapy in the treatment of the upper limb inchildren with hemiplegic cerebral palsy. Cochrane Database SystRev. 2007;(2):CD004149

4. Wasiak J, Hoare B, Wallen M. Botulinum toxin A as an adjunctto treatment in the management of the upper limb in childrenwith spastic cerebral palsy. Cochrane Database Syst Rev. 2004;(4):CD003469

5. Gordon AM, Schneider JA, Chinnan A, Charles JR. Efficacy ofa hand-arm bimanual intensive therapy (HABIT) in childrenwith hemiplegic cerebral palsy: a randomized control trial. DevMed Child Neurol. 2007;49(11):830–838

6. Average A. The Delphi list: a criteria list for quality assessmentof randomized clinical trials for conducting systematic reviewsdeveloped by Delphi consensus. J Clin Epidemiol. 1998;51(12):1235–1241

7. Maher C, Sherrington C, Herbert R, Elkins M. Reliability of thePEDro scale for rating quality of randomised controlled trials.Phys Ther. 2003;83(8):713–721

8. Moseley A, Herbert R, Sherrington C, Maher C. Evidence forphysiotherapy practice: a survey of the Physiotherapy EvidenceDatabase (PEDro). Aust J Physiother. 2002;48(1):43–49

9. Shea BJ, Grimshaw JM, Wells GA, et al. Development ofAMSTAR: a measurement tool to assess the methodologicalquality of systematic reviews. BMC Med Res Methodol. 2007;7:10.Available at: www.biomedcentral.com/1471–2288/7/10. Ac-cessed March 7, 2008

10. Cohen J. Statistical Power Analysis for the Behavioural Sciences.New York, NY: Academic Press; 1977

11. Higgins J, Green S, eds. Cochrane handbook for systematicreviews of interventions: version 5.0.1. Available at: www.cochrane-handbook.org. Accessed September 3, 2008

12. Higgins J, Thompson S, Deeks J, Altman D. Measuring incon-sistency in meta-analyses. BMJ. 2003;327(7414):557–560

13. Reid D, Campbell K. The use of virtual reality with childrenwith cerebral palsy: a pilot randomized trial. Ther Recreation J.2006;40(2):255–268

14. Law M, Cadman D, Rosenbaum P, Walter S, Russell D, DeMat-teo C. Neurodevelopmental therapy and upper-extremity in-hibitive casting for children with cerebral palsy. Dev Med ChildNeurol. 1991;33(5):379–387

15. Law M, Russell D, Pollock N, Rosenbaum P, Walter S, King G.A comparison of intensive neurodevelopmental therapy pluscasting and a regular occupational therapy program for chil-dren with cerebral palsy. Dev Med Child Neurol. 1997;39(10):664–670

16. Corry IS, Cosgrove AP, Walsh EG, McClean D, Graham HK.Botulinum toxin A in the hemiplegic upper limb: a doubleblind trial. Dev Med Child Neurol. 1997;39(3):185–193

17. Fehlings D, Rang M, Glazier J, Steele C. An evaluation ofbotulinum-A toxin injections to improve upper extremityfunction in children with hemiplegic cerebral palsy. J Pediatr.2000;137(3):331–337

18. Speth LA, Leffers P, Janssen-Potten YJ, Vles JS. Botulinumtoxin A and upper limb functional skills in hemiparetic cerebralpalsy: a randomized trial in children receiving intensive ther-apy. Dev Med Child Neurol. 2005;47(7):468–473

19. Lowe K, Novak I, Cusick A. Low-dose/high-concentration lo-calized botulinum toxin A improves upper limb movement andfunction in children with hemiplegic cerebral palsy. Dev MedChild Neurol. 2006;48(3):170–175

20. Wallen M, O’Flaherty S, Waugh M. Functional outcomes ofintramuscular botulinum toxin type A and occupational ther-apy in the upper limbs of children with cerebral palsy: a ran-

domized controlled trial. Arch Phys Med Rehabil. 2007;88(1):1–10

21. Russo RN, Crotty M, Miller MD, Murchland S, Flett P, Haan E.Upper-limb botulinum toxin A injection and occupationaltherapy in children with hemiplegic cerebral palsy identifiedfrom a population register: a single-blind, randomized, con-trolled trial. Pediatrics. 2007;119(5). Available at: www.pediatrics.org/cgi/content/full/119/5/e1149

22. Taub E, Ramey S, DeLuca S, Echols K. Efficacy of constraint-induced movement therapy for children with cerebral palsywith asymmetric motor impairment. Pediatrics. 2004;113(2):305–312

23. Sung IY, Ryu JS, Pyun SB, Yoo SD, Song WH, Park MJ. Efficacyof forced-use therapy in hemiplegic cerebral palsy. Arch PhysMed Rehabil. 2005;86(11):2195–2198

24. Deluca SC, Echols K, Law CR, Ramey SL. Intensive pediatricconstraint-induced therapy for children with cerebral palsy:randomized, controlled, crossover trial. J Child Neurol. 2006;21(11):931–938

25. Charles J, Wolf S, Schneider J, Gordon A. Efficacy of a child-friendly form of constraint-induced movement therapy inhemiplegic cerebral palsy: a randomized control trial. Dev MedChild Neurol. 2006;48(8):635–642

26. Brown GT, Burns SA. The efficacy of neurodevelopmentaltreatment in paediatrics: a systematic review. Br J Occup Ther.2001;64(5):235–244

27. Steultjens EM, Dekker J, Bouter LM, van de Nes JC, LambregtsBL, van den Ende CH. Occupational therapy for children withcerebral palsy: a systematic review. Clin Rehabil. 2004;18(1):1–14

28. Autti-Ramo I, Suoranta J, Anttila H, Malmivaara A, Makela M.Effectiveness of upper and lower limb casting and orthoses inchildren with cerebral palsy: an overview of review articles.Am J Phys Med Rehabil. 2006;85(1):89–103

29. Reeuwijk A, van Schie PEM, Becher JG, Kwakkel G. Effects ofbotulinum toxin type A on upper limb function in childrenwith cerebral palsy: a systematic review. Clin Rehabil. 2006;20(5):375–387

30. McCubbin JA, Shasby GB. Effects of isokinetic exercise onadolescents with cerebral palsy. Adapt Phys Activ Q. 1985;2(1):56–64

31. Kawamura A, Campbell K, Lam-Damji S, Fehlings D. A ran-domized controlled trial comparing botulinum toxin A dosagein the upper extremity of children with spasticity. Dev Med ChildNeurol. 2007;49(5):331–337

32. Lowe K, Novak I, Cusick A. Repeat injection of botulinumtoxin A is safe and effective for upper limb movement andfunction in children with cerebral palsy. Dev Med Child Neurol.2007;49(11):823–829

33. Eliasson AC, Krumlinde-Sundholm L, Shaw K, Wang C. Effectsof constraint-induced movement therapy in young childrenwith hemiplegic cerebral palsy: an adapted model. Dev MedChild Neurol. 2005;47(4):266–275

34. Jannink MJ, van der Wilden GJ, Navis DW, Visser G, GussinkloJ, Ijzerman M. A low-cost video game applied for training ofupper extremity function in children with cerebral palsy: apilot study. Cyberpsychol Behav. 2008;11(1):27–32

35. Ozer K, Chesher SP, Scheker LR. Neuromuscular electricalstimulation and dynamic bracing for the management of up-per-extremity spasticity in children with cerebral palsy. DevMed Child Neurol. 2006;48(7):559–563

36. Dodd KJ, Taylor NF, Damiano DL. A systematic review of theeffectiveness of strength-training programs for people withcerebral palsy. Arch Phys Med Rehabil. 2002;83(8):1157–1164

37. Saleh M, Korner-Bitensky N, Snider L, et al. Actual vs. bestpractices for young children with cerebral palsy: a survey of

PEDIATRICS Volume 123, Number 6, June 2009 e1121 at University of Queensland Library on June 2, 2009 www.pediatrics.orgDownloaded from

paediatric occupational therapists and physiotherapists in Que-bec, Canada. Dev Neurorehabil. 2008;11(1):60–80

38. Bly L. A historical and current view of the basis of NDT. PediatrPhys Ther. 1991;3(3):131–135

39. Charles J, Gordon AM. Development of hand-arm bimanualintensive training (HABIT) for improving bimanual coordina-tion in children with hemiplegic cerebral palsy. Dev Med ChildNeurol. 2006;48(11):931–936

40. Gilmore R, Boyd R, Sakzewski L. A systematic review of upper

limb activity measures for 5–16 year old children with congen-ital hemiplegia. Dev Med Child Neourol. 2008; 50(Suppl 4):66

41. Novak I, Cusick A, Lowe K. A pilot study on the impact ofoccupational therapy home programming for young childrenwith cerebral palsy. Am J Occup Ther. 2007;61(4):463–468

42. Kuhnke N, Juenger H, Walther M, Berweck S, Mall V, Staudt M.Do patients with congenital hemiparesis and ipsilateral corticospi-nal projections respond differently to constraint-induced move-ment therapy? Dev Med Child Neurol. 2008;50(12):898–903

e1122 SAKZEWSKI et al at University of Queensland Library on June 2, 2009 www.pediatrics.orgDownloaded from

DOI: 10.1542/peds.2008-3335 2009;123;e1111-e1122; originally published online May 18, 2009; Pediatrics

Leanne Sakzewski, Jenny Ziviani and Roslyn Boyd Upper-Limb Dysfunction in Children With Congenital Hemiplegia

Systematic Review and Meta-analysis of Therapeutic Management of

& ServicesUpdated Information

http://www.pediatrics.org/cgi/content/full/123/6/e1111including high-resolution figures, can be found at:

References

http://www.pediatrics.org/cgi/content/full/123/6/e1111#BIBLat: This article cites 35 articles, 6 of which you can access for free

Subspecialty Collections

http://www.pediatrics.org/cgi/collection/musculoskeletal_system Musculoskeletal System

following collection(s): This article, along with others on similar topics, appears in the

Permissions & Licensing

http://www.pediatrics.org/misc/Permissions.shtmltables) or in its entirety can be found online at: Information about reproducing this article in parts (figures,

Reprints http://www.pediatrics.org/misc/reprints.shtml

Information about ordering reprints can be found online:

at University of Queensland Library on June 2, 2009 www.pediatrics.orgDownloaded from