Authors:

Heidi Anttila, MSc (Health Sci), PT Jutta Suoranta, MSc (Health Sci), PT Antti Malmivaara, PhD, MD Marjukka Ma kela , PhD, MD, MSc (Clin Epi) Ilona Autti-Ra mo , PhD, MD

Cerebral Palsy

Afliations:
From the Finnish Ofce for Health Technology Assessment, National Research and Development Centre for Welfare and Health, Helsinki, Finland (HA, JS, AM, MM, IAR); Tampere School of Public Health, University of Tampere, Tampere, Finland (JS); Department of General Practice, University of Copenhagen, Denmark (MM); The Social Insurance Institute, Helsinki, Finland (IAR); and Department of Child Neurology, Hospital for Children and Adolescents, University of Helsinki, Finland (IAR).

LITERATURE REVIEW

Effectiveness of Physiotherapy and Conductive Education Interventions in Children with Cerebral Palsy
A Focused Review
ABSTRACT
Anttila H, Suoranta J, Malmivaara A, Ma kela M, Autti-Ra mo I: Effectiveness of physiotherapy and conductive education interventions in children with cerebral palsy: a focused review. Am J Phys Med Rehabil 2008;87:478 501. We conducted a criteria-based appraisal of systematic reviews on the effectiveness of physiotherapy and conductive education interventions in children with cerebral palsy (CP). Computerized bibliographic databases were searched without language restriction up to August 2007. Reviews on trials and descriptive studies were included. Two reviewers independently identified, selected, and assessed the quality of the reviews using the criteria from the Overview Quality Assessment Questionnaire complemented with decision rules. Twenty-one reviews were included, six of which were of high methodological quality. Altogether, the reviews included 23 randomized controlled trials and 104 observational studies on children with CP. The high-quality reviews found some evidence supporting strength training, constraint-induced movement therapy, or hippotherapy, and insufficient evidence on comprehensive physiotherapy and occupational therapy interventions. Conclusions in the other reviews should be interpreted cautiously, although, because of the poor quality of the primary studies, most reviews drew no conclusions on the effectiveness of the reviewed interventions. Reviews on complex interventions in heterogeneous populations should use rigorous methods and report them adequately, closely following the Quality of Reporting of Meta-Analyses recommendations.
Key Words: Physiotherapy, Systematic Review, Cerebral Palsy, Quality Assessment, Clinical Applicability

Correspondence:
All correspondence and requests for reprints should be addressed to Heidi Anttila, Finnish Ofce for Health Technology Assessment, PO Box 220, FIN-00531 Helsinki, Finland.

Disclosures:
This study was funded by Finohta, a national government-funded organization for health technology assessment, and by a grant from the Academy of Finland. The authors have no nancial or personal conicts of interest. 0894-9115/08/8706-0478/0 American Journal of Physical Medicine & Rehabilitation Copyright 2008 by Lippincott Williams & Wilkins
DOI: 10.1097/PHM.0b013e318174ebed

478

Am. J. Phys. Med. Rehabil.

Vol. 87, No. 6

he principles of evidence-based practice are widely accepted among professionals.1,2 The strongest support for evidence-based decision making comes from updated, high-quality systematic reviews (SR). Such reviews identify the relevant studies, appraise their quality, and summarize the results, using sound scientic methodology.3,4 They can also help clinicians to nd relevant answers to clinical questions in a time-efcient and reliable way.3 Professionals treating children with CP often have limited time, skills, and resources to search for evidence and to interpret effectiveness studies.5 Cerebral palsy (CP) is an umbrella term for nonprogressive but often-changing motor impairment syndromes secondary to lesions or abnormalities of the brain arising in the early stages of development.6 Basic management of the motor disability in CP includes physiotherapy (PT) and a wide spectrum of other therapeutic interventions.7 Motor learning goals may also be incorporated into educational programs such as conductive education (CE) instead of separate rehabilitation interventions provided by different professionals.8 An appreciation of the quality of an SR is essential before deciding whether its conclusions should be followed. Such quality may mean the rigor of the review methods, or quality of reporting. Previous evaluations of SRs in many elds imply that readers should not accept them uncritically, and there is a need for improvement of the methodological quality and guidelines for reporting.9 Cochrane reviews are usually more rigorously conducted and reported than non-Cochrane reviews.9 12 There are at least 24 instruments to assess the quality of SRs.13 A rigorously developed and validated tool, the Overview Quality Assessment Questionnaire (OQAQ), has been constituted by Oxman and Guyatt.14,15 Hoving et al.16 have slightly modied this tool and applied it in rehabilitation research. The Quality of Reporting of Meta-Analyses statement describes the preferred way to present the abstract, introduction, methods, results, and discussion sections of a report of meta-analysis, including a ow diagram of the article identication and selection process.17 Balanced interpretation of the applicability and clinical relevance requires accurate information of the reviewed populations, interventions, comparison interventions, and outcomes.18 20 An essential feature of SRs is critical appraisal of the methodological quality of the included primary studies.3,21 Lack of adherence to dened quality criteria may explain the different results of studies on the same topic.22 Published SRs have heterogeneous approaches to assess methodologiJune 2008

cal quality, and this has been infrequently reported and incorporated into the analyses.10,23,24 In this study, we wanted to evaluate the methodological validity of SRs and their clinical usefulness when targeting a heterogeneous population and looking at variably applied interventions such as PT and CE in children with CP. The primary objective was to appraise the methodological quality of the reviews on the effectiveness of PT or CE interventions in children with CP, and to explore what needs to be done to enhance the quality of reviews. The secondary aims were to make conclusions about the effectiveness of the reviewed interventions, and to consider the included study designs, populations, interventions, outcome measures, and results of various PT interventions on children with CP to allow interpretation of possible evidence into clinical practice. Finally, our aim was to use all this information to make suggestions for future studies in this eld.

METHODS Locating and Selecting the Reviews

Only published SR articles were considered. To be included, these publications were required to have descriptions of the searched databases, search time period, and selection criteria for population and interventions. This review included interventions usually provided by physiotherapists and requiring therapeutic management7for instance, neurodevelopmental therapy (NDT), strength training, saddle riding, physical activity, swimming programs, functional therapy, and targeted training. In addition, interventions that in some countries or organizations may be provided either by physiotherapists or occupational therapists (upper-limb interventions) or special teachers (CE) were included. The main focus was to include reviews on therapeutic management without specialized equipment; thus, interventions of solely devices (electrical stimulation, biofeedback, orthotic, or other assistive devices) were excluded. The patients were children or adolescents (aged 3 mos to 20 yrs) with diagnosed CP. If the review had included other interventions or populations, it was included only if at least 80% of the included populations or interventions were similar to our criteria, or if the results of only the CP population and PT interventions were presented separately. Further, the review should report the results of the included studies. Reviews in Danish, English, Finnish, German, Norwegian, or Swedish were accepted. We searched Medline, CINAHL, the Cochrane Database of Systematic Reviews, Database of Abstracts of Reviews of Effects, American College of Physicians Journal Club, Health Technology AsEffectiveness of Interventions in CP

479

sessment database, and the Physiotherapy Evidence Database (http://www.pedro.fhs.usyd.edu.au/ index.html) without language restrictions back to the earliest time available and up until August 2007. An experienced information scientist planned the search strategies. High sensitivity search strategies for Medline and CINAHL databases developed by the University of York25 were employed and complemented with Medical Subject Headings or text words for populations and interventions. The search strategy for Medline is shown Table 1. From January 2003 to August 2007 the search results from Medline and CINAHL were limited to systematic reviews or review articles using the improved lters provided by these databases. The references of the identied review articles were checked by two reviewers (H.A., J.S.) to identify possible reviews. We also searched our personal les of studies and reviews on children with CP. Two reviewers (H.A. and R.K. or I.A.R.) independently screened the titles or abstracts identied in the initial search strategy for inclusion and exclusion criteria. When the title and abstract did not clearly indicate whether an article should be included, two reviewers (H.A., I.A.R.) evaluated the full article for inclusion criteria. The reviewers were not blinded to the names of authors and institutions, sources of funding or results of the review.

score of 18 (Appendix A). Reviews fullling all points, except the item of selection bias (as using two or more assessors for independently judging and selecting studies with predetermined criteria, and/or blinding reviewers to identifying features of study, or to treatment outcome), were regarded as being of high quality. Two evaluators (H.A., J.S.) separately assessed the quality of the included reviews. The discrepancies in evaluations were solved by discussion, and remaining disagreements were decided by a third reviewer (A.M.).

Analysis of the Reviewers Conclusions

The included reviews were classied according to the intervention types: (1) comprehensive PT approaches (e.g., neurodevelopmental or neurophysiological PT, home programs or Vojta), (2) strength training, (3) constraint-induced movement therapy (CIMT), (3) postural control, (4) soft tissue treatment, (6) hydrotherapy, (7) hippotherapy, 8) CE and (9) various (several of the above interventions in one review). For each group of interventions we considered and weighed up the conclusions according to the methodological quality of the SR. We also observed the number and type of included studies and their overlaps between the reviews to obtain a comprehensive overview of the research volume in this eld.

Data Extraction
One of two reviewers (either H.A. or J.S.) extracted the data. The included articles were allocated equally, and data from one review was extracted by both reviewers to ensure similarity. After data extraction the results were checked by the other reviewer. We tabulated the review focus, search strategies and inclusion criteria, data of the included populations, interventions, settings, outcome measures; number of studies and the study designs in each review; methods used in quality assessment and analyses; and the main results and conclusions, and reported adverse effects. For quantitative data we extracted the effect sizes of all outcome measures used.

RESULTS Article Identification and Selection

Figure 1 shows a ow chart of the literature searches and article selection. We found 21 SRs: four reviews on comprehensive PT,26 29 two on strength training,30,31 one on CIMT,32 one on postural control,33 one on soft tissue treatment,34 one on hydrotherapy,35 two on hippotherapy,36,37 four on CE,38 41 and ve reviews covering a wide range of various interventions.42 46

Methodological Quality
The methodological quality scores of the reviews are presented in Table 2. The search methods and inclusion criteria were at least partially described in all reviews, as these were our mandatory criteria for inclusion. Six reviews fullled all criteria other than blinding reviewers from author and outcome information.26,30,32,33,36,42 Twelve reviews26,29,30,3234,36 39,42,43 had dened quality-assessment criteria, and all but one37 used these. Many reviews had inadequacies in search and synthesis methods. The median quality score was 11 out of 18 points (range 317). A summary of the reviews focus and methods is given in Table 3. The methods of qualitative Am. J. Phys. Med. Rehabil.

Assessment of the Methodological Quality

The methodological quality of the included SRs was analyzed using a modied version16 of the method described and validated by Oxman et al.14,15 This checklist evaluates nine items covering search methods, selection of the articles, validity assessment and methods for synthesis, The modication, previously applied in the eld of rehabilitation, consists of the addition of decision rules to increase transparency of the assessment.16 Each item is scored from 0 to 2, with a maximum total

480

Anttila et al.

Vol. 87, No. 6

TABLE 1 High-sensitivity search strategy for identifying review articles in Medline, developed by the University of York25
1. Cerebral palsy/rh, th [rehabilitation, therapy] Cerebral palsy.mp. or cerebral palsy/ Exp physical therapy techniques/ (Physical therapy or physical therapies).ab,ti. Physiotherap$.ab,ti. Exp exercise therapy/ (Physical activity or physical activities).ab,ti. Exp physical therapy (specialty)/ Exp physical education and training/ Rehabilitation.mp. or REHABILITATION/ (Vojta or bobath or neurodevelop$ or NDT or Rood or Kabat or vibroacoust$).ab,ti. Early intervention (education)/ Conductive education.ab,ti. Conservative therap$.ab,ti. (Muscle strength$ or muscle training or motion or therapeutic exercise or excercise training or physical exercise or tness or aerobic training or kinetic chain).ab,ti. 16. Movement.mp. or EXERCISE MOVEMENT TECHNIQUES/or MOVEMENT/ 17. SWIMMING/or swimming.mp. or hydrotherapy.mp. 18. (Functional therapy or functional therapies).ab,ti. 19. (Self-care training or motor control or motor learning).ab,ti. 20. Occupational therapy.mp. or Occupational Therapy/ 21. (Constraint adj induced).mp. [mpti, ab, tx, kw, ct, ot, sh, hw] 22. Restraint, physical/ 23. (Forced adj2 treatment).mp. [mpti, ab, tx, kw, ct, ot, sh, hw] 24. (Psychomotor performance or sensation).mp. [mpti, ab, tx, kw, ct, ot, sh, hw] 25. Sensory integration.ab,ti. 26. (Sensory adj perceptual).mp. [mpti, ab, tx, kw, ct, ot, sh, hw] 27. Parentchild relations/or parents/or parent education.mp. 28. Physical stimulation.mp. or physical stimulation/ 29. (Posture or positioning).mp. [mpti, ab, tx, kw, ct, ot, sh, hw] 30. Facilitat$.ti,ab. 31. 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 or 19 or 20 or 21 or 22 or 23 or 24 or 25 or 26 or 27 or 28 or 29 or 30 32. 2 and 31 33. 1 or 32 34. Controlled.ab. 35. Design.ab. 36. Evidence.ab. 37. Extraction.ab. 38. Randomized controlled trials/ 39. Meta-analysis.pt. 40. Review.pt. 41. Sources.ab. 42. Studies.ab. 43. Or/3442 44. Letter.pt. 45. Comment.pt. 46. Editorial.pt. 47. Or/4446 48. 43 not 47 49. 33 and 48 Limitations (from January 2003 to August 2007): 50. Limit 49 to systematic reviews 51. Limit 49 to review articles 52. 51 and 50 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

June 2008

Effectiveness of Interventions in CP

481

FIGURE 1 Flow chart of the article selection process. categorization and synthesis varied considerably. Eight reviews classied the study results into outcome-related categories by dimensions of disability,27 ICIDH-2 (International Classication of Impairments, Disabilities and Handicaps),34,38,43 ICF (International Classication of Functioning, Disability and Health),30,35,36 or own classication.41 One review applied meta-analysis on randomized controlled trials (RCT).43 Effect sizes and condence intervals were available from three reviews.30,32,34 Nine reviews applied levels of evidence analysis. Four reviews26,31,35,43 used a method described by Sackett47,48 and one review36 modied this method2 to include Physiotherapy Evidence Database ratings. Three reviews27,33,34 applied American Academy for Cerebral Palsy and Developmental Medicine (AACPDM) methodology,49 and one review33 used the methodology of AACPDM and Sackett.47 One review42 applied the evidence synthesis method described by van Tulder.50 A full description of the characteristics of reviewed populations, interventions, outcome measures, and results of studies on children with CP is in Appendix B. The population in terms of age, type and severity of CP, the interventions, and the outcome measures are heterogeneous in all reviews and intervention groups. The included studies were conducted in various settings (clinic, home, school, or community). The settings are sufciently reported in only four reviews.30,38,39,42 Except for one review,32 the content of each intervention is described only with a short title. The number of different outcome measures reported varies from 6 to 30 per review. Two reviews do not report any outcomes.40,45

Conclusions on Effectiveness of the Interventions Included in the Reviews Comprehensive PT

One high-quality26 and three low-quality SRs on comprehensive PT approaches have evaluated 15 RCTs and 28 observational studies, of which 9 RCTs61 69 and 19 observational studies were on children with CP. Seven of the 9 RCTs (total number of children, n 309) and 5 of the 19 observational studies (n 493) are included in more than one review. The high-quality review concludes that the current research . . . does not clearly demonstrate the efcacy or inefcacy of NDT as a treatment approach.26(p242) Conclusions in the low-quality reviews are similar: The preponderance of the results . . . did not confer any advan2729

Characteristics of the Review Contents

The reviews were based altogether on 31 RCTs and 199 observational studies. Ten reviews included non-CP children,26,28,29,34,35,37,40,41,45,46 and four reviews also included interventions that were outside the scope of this review,33,42,43,45 which were excluded from the analyses. Twenty-three RCTs and 104 observational studies were on children with CP; of these, 13 RCTs and 29 observational studies were included in more than one review (Table 4).

482

Anttila et al.

Am. J. Phys. Med. Rehabil.

Vol. 87, No. 6

June 2008
Search Methods Selection Methods Validity Assessment Synthesis Search Comprehensiveness Inclusion Criteria Avoidance of Selection Bias Synthesis Methods Denition of the Validity Assessment Criteria Use of the Quality-Assessment Criteria Acceptability of the Synthesis Methods Conclusions Supported by Data Analysis Total Points (Max 18) 2 2 1 1 2 1 2 2 1 2 2 1 1 2 1 2 2 2 2 1 2 13 8 0 2 2 2 1 2 16 5 0 1 0 0 0 0 0 8 13 2 2 1 2 0 1 0 0 2 2 0 0 2 2 2 2 0 0 2 2 0 0 0 12 0 9 2 1 0 2 0 2 2 1 2 2 1 2 2 2 2 0 2 0 2 2 0 0 2 1 0 0 0 10 1 10 2 1 1 1 2 0 2 0 2 2 2 2 2 2 2 0 2 0 0 1 2 1 0 0 0 11 2 8 2 2 1 1 1 0 0 0 2 0 0 2 2 0 0 2 2 2 0 0 2 1 0 1 2 2 2 2 0 2 2 0 1 1 0 1 2 1 0 1 1 8 8 5 2 2 0 2 2 2 2 2 0 2 2 0 2 0 0 2 2 0 1 2 2 14 1 6 17 11 3 10 17 11 17 17 11 13 16 7 14 12 3 9 17 11 6 6 9

TABLE 2 Methodological assessment scores of the included 21 systematic reviews

First Author (Year)

Search Methods

2 2 1 1

2 2

2 2

Effectiveness of Interventions in CP

Comprehensive physiotherapy Brown26 (2001) Butler27 (2001) Parette28 (1991) Tirosh29 (1989) Strength training Dodd30 (2002) Darrah31 (1997) Constraint-induced movement therapy Hoare32 (2007) Postural control Harris33 (2005) Soft-tissue treatment Pin34 (2006) Hydrotherapy Getz35 (2006) Hippotherapy Snider36 (2007) Sterba37 (2007) Conductive education Darrah38 (2003) Ludwig39 (2000) Pedersen40 (2000) French41 (1992) Various interventions Steultjens42 (2004) Boyd43 (2001) Woolfson44 (1999) Hur45 (1995) Horn46 (1991) Yes Partially No

2 2 1 1

2 2 1 1 2 15 6 0

Scoring: 2, the criterion is fullled; 1, partially fullled or cannot tell; 0, not fullled or not reported; the decision rules are in Appendix A.

483

484
Objectives of the Review CCT (11) OD (6) 17 NDT: ? Designs Included* Quality Score Methods of Analyses (Search Period, Methodological Quality Assessment (QA), Categorization of the Results, Synthesis Method) Conclusions of Review* RCT (7) OD (14) Search until 1998 QA: 5 criteria, Quality Assessment of Randomised Clinical Trials by Jadad et al.51; 1 criterion, Concealment of Treatment Allocation scale by Schulz et al.52 Categorization by ndings/results (benets, statistical signicance) Levels-of-evidence analyses by Sackett48 Search (1956spring 2001) QA: No Categorization by outcomes (dimensions of disability) Levels-of-evidence analyses by Butler49 11 CCT (3) OD (10) RCT (7) CCT (2) 3 Search period (19601989) QA: No Categorization by the study designs Descriptive analyses Search (19731987) QA: 12 criteria, modied from Sackett53 Categorization by study results (benets/no benets) and methodology scores Descriptive analyses. Effect sizes and 95% CIs provided 10 Physiotherapy: efcacy 17 Strength training program: muscle strength , increased spasticity , mobility ?, function ?, participation ?, contextual factors ? 11 Progressive muscle strengthening: efcacy , muscle performance , functional abilities ? RCT (1) OD (6) Search (19661997) QA: 10 of 11 criteria on PEDro scale based on the Delphi list by Verhagen et al.54 Categorization by ICF Descriptive analyses Search (19661997) QA: No Levels-of-evidence analyses by Sackett47 RCT (2) CCT (1) Search until August 2006 QA: 4 criteria by Schulz et al.52 Studies analyzed separately Descriptive analyses: data entered into review manager 4.2. Effect sizes and 95% CIs were provided 17 Modied CIMT: CIMT: , Forced use:

TABLE 3 Characteristics of methods and conclusions of systematic reviews on physiotherapy and conductive education (n 21)

Anttila et al.

Review (Year)

Comprehensive physiotherapy Brown and Burns26 Efcacy on NDT in pediatric (2001) subjects (18 yrs) diagnosed with a neurological dysfunction

Butler and Darrah27 (2001)

Current state of evidence about NDT in children with CP

Parette et al.28 (1991)

NDT: dynamic ROM , abnormal motor responses ?, slowing or prevention of contractures ?, facilitation of normal motor development ?, functional motor activities ?, socialemotional development ?, language ?, cognitive development ?, home environments ?, parentchild interaction ?, parent satisfaction ? More intensive NDT: benet Therapeutic intervention: efcacy Intensive therapeutic intervention: efcacy ?

Tirosh and Rabino29 (1989)

Efcacy and intensity of therapeutic interventions (OT and PT) for infants and young children (5 yrs) with CP Efcacy and effectiveness of PT interventions in the rehabilitation of children with CP

Strength training Dodd et al.30 (2002)

Effects of strength training or RCT (1) progressive resistance exercise OD (9) program for adults or children with CP Reviews (4)

Am. J. Phys. Med. Rehabil.

Darrah et al.31 (1997)

Effects of progressive resistance muscle strengthening in children with a diagnosis of CP

Constraint-induced movement therapy Hoare et al.32 (2007) Effectiveness of CIMT, modied CIMT, or forced use in the treatment of affected upper limb in children (19 yrs) with hemiplegic CP

Vol. 87, No. 6

June 2008
Designs Included* Quality Score Methods of Analyses (Search Period, Methodological Quality Assessment (QA), Categorization of the Results, Synthesis Method) Conclusions of Review* OD (12) Search (19902004) QA: AACPDM Quality Assessment Scale by Butler49 Categorization by study designs (group/single subject) Levels-of-evidence analyses by Sackett47 for group designs; Butler49 for single-subject designs Search until April 2006 QA: PEDro scale54 Categorization by outcomes (ICIDH-2) Levels-of-evidence analyses by Butler49 Mean effect sizes and condence intervals calculated Search (1966January 2005) QA: No Categorization by outcomes (ICF) Levels-of-evidence synthesis by Sackett48 13 11 17 Externally generated movements: Postural perturbations: reactive balance Group NDT or practice: RCT (4) OD (3) Passive stretching: ROM , spasticity Sustained stretching vs. manual stretching: ROM , spasticity RCT (1) OD (10) Hydrotherapy: respiratory function , activity ?, participation ? RCT (3) OD (6) 17 RCT (0) CCT (3) OD (7) Search (18062005) QA: RCTs by PEDro scale54; other designs by NewcastleOttawa scale55 Categorization by interventions and outcomes (ICF)Levels-of-evidence synthesis by Sackett,2 modied to include PEDro scale54 Search (1981December 2005) QA: 16 criteria, Critical Review Form for Quantitative Studies by Law et al.56 Categorization by interventions Descriptive analyses Search (1966Fall 2001) QA: 7 criteria (inclusion/exclusion criteria, intervention, measures used, blinding, statistical evaluation, dropouts, controlling the variables and limiting bias) Categorization by outcomes (dimension of disability, ICIDH-2) Descriptive analyses 7 All intervention categories: gross motor function RCT (1) OD (14) 14 CE: ?

TABLE 3 Continued

Review (Year)

Objectives of the Review

Postural control Harris and Roxborough33 (2005)

Efcacy and effectiveness of postural control intervention strategies for children with CP

Soft-tissue treatment Pin et al.34 (2006)

Effectiveness of passive stretching by using ICIDH-2 in children with CP

Hydrotherapy Getz et al.35 (2006)

Effectiveness of aquatic interventions with regard to the ICF dimensions in children with neuromotor impairments

Hippotherapy Snider et al.36 (2007)

Effectiveness of hippotherapy and therapeutic horseback riding on impairments, activities, and participation in children with CP

Hippotherapy: muscle symmetry Therapeutic horseback riding or hippotherapy: activities , participation ?

Sterba37 (2007)

Effectiveness of horseback riding used as therapy to improve gross motor function in children with CP

Effectiveness of Interventions in CP

Conductive education Darrah et al.38 (2003)

Current state of evidence of CE programs in children with CP

485

486
Objectives of the Review RCT (1)CCT (4) OD (4) 12 CE: ? Designs Included* Quality Score Methods of Analyses (Search Period, Methodological Quality Assessment (QA), Categorization of the Results, Synthesis Method) Conclusions of Review* RCT (1) OD (8) RCT (1) OD (5) 9 3 CE: Search (19662000) QA: 6 criteria (sampling strategy, population, setting, intervention, statistical methods, and outcome measures) by University of Alberta,57 Jadad,58 and Lonigan et al.59 Categorization by participants (children and parents) Descriptive analyses Search period not available QA: No Categorization by effects (yes/no) Descriptive analyses Search period not available QA: No Categorization by outcomes Descriptive analyses CE: ? Search until June 2003 QA: 11 criteria by van Tulder et al.60 and 1 criterion by Wells et al.55. Categorization by interventions Levels-of-evidence analysis by van Tulder et al.50 17 RCT (7) CCT (1) OD (9) RCT (5, of which 2 were botulinum studies) OD (51) 11 RCT (2) OD (8) RCT (7) CCT (2) OD (28) RCT (2) OD (26) 6 All interventions: ? 6 Therapeutic interventions: ? Search (1966December 2000) QA: Only RCTs by PEDro scale54 Categorization by interventions, ICIDH-2 Meta-analysis of 3 studies with same outcome measure Levels-of-evidence analyses by Sackett48 Search period not available QA: No Categorization by intervention groups Descriptive analyses Search (19661994). QA: No Categorization by study designs Descriptive analyses Search (19831989) QA: No Categorization by interventions Descriptive analyses 9 NDT: ? Behavioral programming (how to train):

TABLE 3 Continued

Review (Year)

Anttila et al.

Ludwig et al.39 (2000)

Effectiveness of CE and intensive therapy on the overall learning and health status of children with CP or other motor disorders (or the perceptions of parents whose children had received CE)

Pedersen40 (2000)

Examination of CE programs based on CE principles in studies using control groups

French and Nommensen41 (1992)

Empirical legitimacy of CE programs outside Hungary

Various interventions Steultjens et al.42 (2004)

Efcacy of six OT intervention categories for children (19 yrs) with CP

Boyd et al.43 (2001)

Efcacy of different treatments for the management of upper-limb dysfunction in children with CP

Comprehensive OT: ? Training of sensorimotor function: ? Training of skills: ? Training of sensorimotor function vs. training of skills: ? Parental counseling: ? Advice/instruction on assistive devices: ? Provision of splints: ? All interventions: ?

Woolfson44 (1999)

Hur45 (1995)

Efcacy on educational programs for infants and preschool (5 yrs) children with CP Effect of PT interventions for children with CP

Am. J. Phys. Med. Rehabil.

Horn46 (1991)

Effectiveness of basic motor skills interventions (training of motor skills or components of skills) for children (10 yrs) with signicant motor decits attributable to neurological disorders

Vol. 87, No. 6

* As stated by the author; Improved outcome; indications for improvement; evidence for ineffectiveness; ? insufcient evidence. CP, cerebral palsy; PT, physiotherapy; OT, occupational therapy; NDT, neurodevelopmental therapy; CIMT, constraint-induced movement therapy; ROM, range of motion; CE, conductive education; RCT, randomized controlled trial; CCT, clinical controlled trial; OD, other design; ICIDH-2, International Classication of Impairments, Disabilities and Handicaps; PEDro scale, Physiotherapy Evidence Database scale; QA, quality-assessment methods.

Brown26

Butler27

Parette28

Tirosh29

Dodd30

Darrah31

Hoare32

Harris33 Pin34

Getz35

Snider36 Sterba37

Darrah38

Ludwig39

Pedersen40 French41

Steultjens42 Boyd43

Woolfson44

Hur45

First Author (Year) 1 1 1 1 1 1 1b 1b 1 1 1 1 1b 1 1 1b,c 1 1b 1 1 1b 1 1 1b 1b 1 1 1 4

b

Sample Size 1b

Horn46

June 2008
Systematic Reviews (First Author) No. of Reviews Including the Study 1 2 1 1 1 1 1 1 1b 1 1b,c 1b 1 1b 1 1 1 1 1 1 1b,c 1 1b 1b 1 4 4 1 1 2 1 4 1 7 1 3 3 5 1 2 6 6 5 1 1 2 0 3 1
c

TABLE 4 Included RCTs on children with CP and their overlaps in the 21 systematic reviews

Randomized Controlled Trials

31 14 15 18

58

34a

50 28 20 19 15 73 19 48

Effectiveness of Interventions in CP
7 1 1 1 8 1b 1b 1b 5

Sung71 (2005) Cherng78 (2004) Benda79 (2003) Deluca72 (2002), Taub73 (2004) McConachie82 (2000) Reddihough81 (1998) Law61 (1997) Steinbok62 (1997) Dorval77 (1996) MacKinnon80 (1995) ODwyer74 (1994) Law63 (1991) Richards75 (1991) Palmer64 (1990), Palmer65 (1988) Tremblay76 (1990) Hanzlik83 (1989) McGubbin70 (1985) Sommerfeld66 (1981) Talbot84 (1981) Sellick85 (1980) Scherzer67 (1976) Carlsen68 (1975) Wrigth69 (1973) Sum 3 2 1
Not classied as an RCT by the authors; only the Palmer et al.
65

21 20 30 29

59 20 22 12 47 702

1
publication was included in the review.

1 1 1 1 5

1 1 1 1 6

487

The trial included additional 32 nonrandomized children;

tage to NDT over the alternatives . . .27(p22); . . . only four studies used a rigorous design, and three of these concluded that no evidence exists for the efcacy of the intervention . . .29(p555); or the available literature offers some support for the efcacy of therapeutic interventions for infants and young children with cerebral palsy.28(p5)

Hippotherapy
One high-quality36 and one lower-quality37 review compare therapist-directed hippotherapy vs. recreational horseback riding therapy. These reviews include three RCTs78 80 and seven observational studies (n 100). Of these, two RCTs78,80 and six observational studies are included in both reviews. The results of Snider et al.36 indicate that hippotherapy has short-term positive effects on muscle symmetry in the trunk and hip and that therapeutic horseback riding is no more effective than other therapies for improving muscle tone. Observational studies have shown positive effects of both hippotherapy and therapeutic horseback riding on activities. The low-quality review37 states that clinicians and therapists can recommend hippotherapy as an efcacious, medically indicated therapy for gross motor rehabilitation of children with CP.

Strength Training
We found one high-quality30 and one lowerquality31 SR on strength training in children with CP, evaluating altogether 1 RCT and 11 observational studies (n 102). Four studies are included in both reviews, including the RCT.70 The conclusions are similar: strength training programs improve muscle strength in children and young adults with CP, with no adverse effects on spasticity.30,31

CIMT
One high-quality Cochrane review32 analyzed two RCTs7173 and one controlled clinical trial (CCT) (n 94). This SR found a signicant treatment effect [on bimanual performance] using modied CIMT in a single trial. A positive trend favoring CIMT and forced use was also demonstrated.32(p10)

CE
The effectiveness of CE has been evaluated in four reviews.38 41 These include 1 RCT81 and 21 observational studies (n 1264), with 7 of the observational studies being included in more than one review. The overall conclusions of these reviews are concordant: the number of studies was too small, and the quality was too low, to make conclusions about the effectiveness or ineffectiveness of CE.

Postural Control
From one high-quality review33 on interventions aiming to improve postural control, we included four observational studies on NDT, rocker platform, and massed practice (n 22). The review concludes with suggestive evidence for the effectiveness of interventions comprising externally generated movement on the development of postural control, promising evidence for postural perturbations improving reactive balance when a high number of repetitions is provided, and moderately strong evidence for the lack of group-level effects of 1 wk of NDT or practice.

Various Interventions
One high-quality42 and four low-quality reviews43 46 include different types of interventions from 13 RCTs61,63-70,81-85 and 47 observational studies. The reviewers conclusions unanimously pinpoint the paucity of evidence. According to Steultjens et al.,42 evidence for the efcacy of occupational therapy is insufcient in all intervention categories. Horn et al.46 have found no evidence of the effectiveness or ineffectiveness of NDT, sensory integration or naturalistic programming. No conclusions are made on treatment approaches for upper-limb dysfunction,43 on training and behavior-modication techniques in conjunction with PT,45 or on multidomain developmental and CE programs,44 because of the paucity of evidence and methodological limitations.

Soft-Tissue Treatment
One low-quality review34 evaluated three RCTs74 76 and two observational studies on passive stretching in children with CP (n 89). The conclusion is that the effectiveness of passive stretching remains weak, although some evidence indicates that sustained stretching is preferable to manual stretching in improving range of motion and reducing spasticity.

DISCUSSION Hydrotherapy
In one low-quality review on aquatic interventions,35 one RCT77 and four observational studies address children with CP (n 68). Getz et al.35 conclude that hydrotherapy might improve respiratory function in children with CP. We identied and critically analyzed 21 SRs on PT and CE interventions in children and adolescents with CP. Our analysis of the quality of evidence summaries and of the volume, characteristics, and effectiveness of primary studies in this eld provides insights into the current scientic Am. J. Phys. Med. Rehabil.

488

Anttila et al.

Vol. 87, No. 6

basis for clinical decision making and future research agendas.

Clinical Heterogeneity of the Reviewed Interventions

CP is a heterogeneous condition where the developmental potential and goals for rehabilitation vary with age.93 Many reviews include non-CP children, which may bias conclusions when results are not analyzed separately. Thus, it is difcult to determine which patient groups may benet from the studied interventions. An SR in this eld can be improved by focusing on clearly dened target groups. Many older reviews include a variety of incomparable interventions. Complex interventions with several, often vaguely dened, interacting components can complicate analyses, decreasing clinical applicability. Interventions may be insufciently described in the original studies,94 and interventions in different countries may actually not be comparable at all, despite similar names.39 Narrower inclusion criteria for interventions may allow better comparison across studies, as seen in the recent reviews.30 37 First steps toward international intervention categories in PT with adults have been taken,95 which may help future evidence syntheses. It is important to know whether all clinically relevant outcomes have been reported. Numerous noncomparable outcome measures were used in the studies, and the clinical relevance of many of them remains unclear. Without a consensus on measures to apply in CP94 or in rehabilitation in general,96 the combination of results across studies is problematic. All these factors may complicate reviews of complex interventions in this heterogeneous population, as recognized earlier.92 Clinicians and researchers would benet from a more precise description of the studies in terms of population, interventions, comparison interventions, and outcomes to increase the clinical applicability of reviews. Many problems were caused by insufcient reporting of the details of the reviewed studies, possibly because of poor reporting in original studies.97 We recommend using guidelines such as the Quality of Reporting of Meta-Analyses statement to increase the quality of the review report.17

Recommendations for SRs

SRs are based on critically appraised, highquality effectiveness research, usually RCTs. Six of the identied SRs were of high quality.26,30,32,33,36,42 All reviews included observational studies, possibly because of the limited number of RCTs available. The terminology in various observational study designs was very mixed. Some reviews did not even recognize RCTs among their included studies.26,28,37 Assessment of the included studies revealed that reviews on the same topics included somewhat different studies. No review had excluded studies on the basis of quality. The differences may be attributable to different search periods, search terms and databases, or somewhat different foci on inclusion criteria. We recommend that future reviews clearly dene what study designs are to be included. Twelve reviews had dened quality-assessment criteria, and all but one used these. However, most quality criteria only suit RCTs, not observational studies. In three reviews,27,33,34 quality was assessed by a tool49 that raises singlecase studies to the level of RCTs in the evidence hierarchy. Today, the AACPDM methodology has been updated86 to meet the criteria of evidencebased evaluation.87 The variety of quality-assessment tools reects the lack of consensus as to which components and what tools would best assess trial quality.10,88,89 Previous research on the role of nonrandomized studies and case series in SRs in other elds has been hampered by both the paucity and the poor quality of these studies.90,91 More research is needed on how the methodological features of observational studies affect outcomes in this eld. The qualitative synthesis methods were built on different combinations of different aspects across the reviews. The categorization of the results was made either by outcomes, interventions, study designs, study quality, or populations. These categories are then summarized, either descriptively by levels of evidence analyses, or by counting for the numbers of studies in different categories. Most of these hide important factors, such as the number of patients included and the real effect sizes. Only three reviews provide effect sizes together with the condence intervals. A common understanding on how to summarize ndings on individual studies in a qualitative synthesis is obviously needed, as found earlier on Cochrane reviews in PT and occupational therapy interventions.92 June 2008

Effectiveness and Clinical Applicability of the Reviewed Interventions

The six high-quality reviews allow conclusions on some of the interventions reviewed. Evidence of comprehensive PT approaches26 and occupational therapy interventions42 is insufcient. The four high-quality reviews on more focused interventions provided positive evidence on some outcomes: strength training on muscle strength,30 intensive upper-extremity training on bimanual Effectiveness of Interventions in CP

489

performance,32 hippotherapy on muscle symmetry and activities,36 and effectiveness of externally generated movements and postural perturbations to reactive balance.33 The four reviews posing targeted questions may be clinically easier to apply because they include a limited number and type of interventions and outcomes. For example, in the strength training review, the interventions, outcome measures, and patient inclusion criteria are fairly unambiguous.30 The positive evidence on effectiveness is based on only one RCT70 and several concordant observational studies. The evidence-grading system by the GRADE Working Group98 suggests upgrading for cohort studies, when two or more observational studies show a consistent association, with no plausible confounders. This was the case for the studies on strength training. Some low-quality reviews have made conclusions on indicative evidence of passive and sustained stretching on range of motion and spasticity,34 hydrotherapy on respiratory function,35 and hippotherapy and horseback riding therapy for gross motor performance.37 This evidence should be interpreted cautiously because of the methodological limitations. In the other reviews, the authors concordantly state that they can make no conclusion on the effectiveness or ineffectiveness of the reviewed interventions. These reviews thus provide no support on applying the studied interventions in clinical practice.

CONCLUSIONS
SRs of PT or CE interventions in children with CP require cautious interpretation of the ndings. On the basis of six high-quality reviews, conclusions on the effectiveness of some interventions on specic outcomes could be made. Otherwise, the effects remain unclear or unsupported by data. The low number of RCTs resulted in the inclusion of a large variety of observational studies in reviews. Well-conducted studies on current treatment options as well as new treatment approaches using valid outcomes are obviously needed. Because reviews on rehabilitation within such a heterogeneous population as CP are demanding to conduct, compliance with methodological guidelines on reporting, such as the Quality of Reporting of MetaAnalyses statement, is recommended.

ACKNOWLEDGMENTS
The authors thank professor Regina Kunz, PhD, MD, MSc (Epi), for reviewing the abstracts until June 2003; information scientists Riitta Grahn, MSc, and Jaana Isoja rvi, MSocS, for their support in the literature search; and Mark Phillips, BA, for his help in reviewing the language of the article.
REFERENCES
1. World Confederation of Physiotherapy: Evidence Based PractiseAn International Perspective. Report of an Expert Meeting of WCPT Member Organisations. 1315 October, 2001. London, The Chartered Society of Physiotherapy, 2002 2. Sackett DL, Straus SE, Richardson WS, Rosenberg W, Haynes RB: Evidence-Based Medicine: How to Practice and Teach EBM. Edinburgh, Churchill Livingstone, 2000 3. Oxman AD, Cook DJ, Guyatt G: Users guides to the medical literature. VI. How to use an overview. Evidence-Based Medicine Working Group. JAMA 1994;272:136771 4. Khan KS, Kunz R, Kleijnen J, Antes G: Systematic Reviews to Support Evidence-Based Medicine: How to Review and Apply Findings of Healthcare Research. London, Royal Society of Medicine Press, 2003 5. Maher CG, Sherrington C, Elkins M, Herbert RD, Moseley AM: Challenges for evidence-based physical therapy: accessing and interpreting high-quality evidence on therapy. Phys Ther 2004;84:64454 6. Mutch LW, Alberman E, Hagberg B, Kodama K, Velickovic MW: Cerebral palsy epidemiology. Dev Med Child Neurol 1992;17:1825 7. Siebes RC, Wijnroks L, Vermeer A: Qualitative analysis of therapeutic motor intervention programmes for children with cerebral palsy: an update. Dev Med Child Neurol 2002;44:593603 8. Kozma I: The basic principles and present practise of conductive education. Eur J Spec Needs Educ 1995;10: 11123 9. Moher D, Tetzlaff J, Tricco AC, Sampson M, Altman DG: Epidemiology and reporting characteristics of systematic reviews. PLoS Medicine. 2007;4:e78 10. Moher D, Cook DJ, Jadad AR, et al: Assessing the quality of reports of randomised trials: implications for the conduct of meta-analyses. Health Technol Assess 1999;3:193 11. Shea B, Moher D, Graham I, Pham BA, Tugwell P: A comparison of the quality of Cochrane reviews and sys-

Recommendations for Clinical Studies in CP

Given the considerable variation in the CP population, interventions, and outcome measures, future studies would gain from agreement on a CP denition.99 Careful compiling of the interventions components and detailed reporting would allow interpretation of the applicability to clinical practice. Inconclusive evidence on the more comprehensive treatment approaches calls for better understanding of the intervention components,94 and a phase-oriented approach may be useful.100 Evidence from studies on focused interventions, such as strength training and CIMT, could be used when developing complex interventions. Indeed, the many single components of PT interventions could be studied separately. Many treatment approaches have not been rigorously evaluated, and well-conducted studies are still needed. Given the number of outcome measures used, a consensus on outcomes is needed. Future studies should apply validated measures covering all ICF components and health-related quality of life.

490

Anttila et al.

Am. J. Phys. Med. Rehabil.

Vol. 87, No. 6

tematic reviews published in paper-based journals. Eval Health Prof 2002;25:11629 12. Boluyt N, van der Lee JH, Moyer VA, Brand PLP, Offringa M: State of the evidence on acute asthma management in children: a critical appraisal of systematic reviews. Pediatrics 2007;120:133443 13. Shea B, Dube C, Moher D: Assessing the quality of reports of systematic reviews: the QUOROM statement compared to other tools, in Egger M, Davey Smith G, Altman DG (eds): Systematic Reviews in Health Care: Meta-Analysis in Context, ed 2. London, BMJ Publishing Group, 2001, pp 12239 14. Oxman AD, Guatt GH: Validation of an index of the quality of review articles. J Clin Epidemiol 1991;44:12718 15. Oxman AD, Guatt GH, Dinger J, et al: Agreement among reviewers of review articles. J Clin Epidemiol 1991;44:918 16. Hoving JL, Gross AR, Gasner D, et al: A critical appraisal of review articles on the effectiveness of conservative treatment for neck pain. Spine 2001;26:196205 17. Moher D, Cook D, Eastwood S, Olkin I, Rennie D, Stroup D: Improving the quality of reports of meta-analyses of randomised controlled trials; the QUOROM statement. Quality of reporting meta-analyses. Lancet 1999;354: 1896900 18. Guatt G, Drummond R (eds): Users Guides to the Medical Literature. Essentials of Evidence-Based Clinical Practice. Chicago, American Medical Association, 2002 19. van Tulder M, Furlan A, Bombardier C, Bouter LM; Editorial Board of the Cochrane Collaboration Back Review Group: Updated method guidelines for systematic reviews in the Cochrane Collaboration Back Review Group. Spine 2003;28:12909 20. Malmivaara A, Koes BW, Bouter LM, van Tulder MW: Applicability and clinical relevance of results in randomized controlled trials: the Cochrane review on exercise therapy for low back pain as an example. Spine 2006;31: 14059 21. Ju ni P, Altman DG, Egger M: Assessing the quality of randomised controlled trials, in Egger M, Davey Smith G, Altman DG (eds): Systematic Reviews in Health Care: Meta-Analysis in Context, ed 2. London, BMJ Publishing Group, 2001, pp 87108 22. Egger M, Ju ni P, Bartlett C: How important are comprehensive literature searches and the assessment of trial quality in systematic reviews? Empirical study. Health Technol Assess 2003;7:176 23. Moher D, Jadad R, Nichol G, Fenman M, Tugwell P, Walsh S: Assessing the quality of reports of randomized trials; an annotated bibliography of scales and checklists. Control Clin Trials 1995;16:6273 24. Moja LP, Telaro E, DAmico R, et al: Assessment of methodological quality of primary studies by systematic reviews: results of the metaquality cross sectional study. BMJ 2005;330:1053 25. Centre for Reviews and Dissemination: Search strategies to identify reviews and meta-analyses in MEDLINE and CINAHL. Available at: http://www.york.ac.uk/inst/crd/ search.htm. Accessed February 19, 2008 26. Brown GT, Burns SA: The efcacy of neurodevelopmental treatments in children: a systematic review. Br J Occup Ther 2001;64:23544 27. Butler C, Darrah J: AACPDM evidence report: effects of neurodevelopmental treatment (NDT) for cerebral palsy. Dev Med Child Neurol 2001;43:77890 28. Parette HPJ, Hendricks MD, Rock SL: Efcacy of therapeutic intervention intensity with infants and young children with cerebral palsy. Infants Young Child 1991;4:119 29. Tirosh E, Rabino S: Physiotherapy for children with cerebral palsy. Evidence for its efcacy. Am J Dis Child 1989; 143:5525

30. Dodd KJ, Taylor NF, Damiano DL: A systematic review of the effectiveness of strength training programs for people with cerebral palsy. Arch Phys Med Rehabil 2002;83: 115764 31. Darrah J, Fan JS, Chen LC, Nunweiler J, Watkins B: Review of the effects of progressive resisted muscle strengthening in children with cerebral palsy: a clinical consensus exercise. Pediatr Phys Ther 1997;9:127 32. 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 33. Harris SR, Roxborough L: Efcacy and effectiveness of physical therapy in enhancing postural control in children with cerebral palsy. Neural Plast 2005;12:22943 34. Pin T, Dyke P, Chan M: The effectiveness of passive stretching in children with cerebral palsy. Dev Med Child Neurol 2006;48:85562 35. Getz M, Hutzler Y, Vermeer A: Effects of aquatic interventions in children with neuromotor impairments: a systematic review of the literature. Clin Rehabil 2006;20:92736 36. Snider L, Korner-Bitensky N, Kammann C, Warner S, Saleh M: Horseback riding as therapy for children with cerebral palsy: is there evidence of its effectiveness? Phys Occup Ther Pediatr 2007;27:523 37. Sterba JA: Does horseback riding therapy or therapistdirected hippotherapy rehabilitate children with cerebral palsy? Dev Med Child Neurol 2007;49:6873 38. Darrah J, Watkins B, Chen L, Bonin C: Effects of Conductive Education Intervention for Children with a Diagnosis of Cerebral Palsy: An AACPDM Evidence Report. Available at: http://www.aacpdm.org/resources/ConEdOut.pdf. Accessed March 10, 2003 39. Ludwig S, Leggett P, Harstall C: Conductive Education for Children with Cerebral Palsy. Edmonton, Alberta Heritage Foundation for Medical Research, 2000. HTA 22 40. Pedersen AV: Conductive educationa critical appraisal. Adv Physiother 2000;2:7582 41. French L, Nommensen A: Conductive education evaluated: future directions. Aust Occup Ther J 1992;39:1724 42. Steultjens EMJ, Dekker J, Bouter LM, van de Nes JCM, Lambregts BLM, van den Ende CHM: Occupational therapy for children with cerebral palsy: a systematic review. Clin Rehabil 2004;18:114 43. Boyd RN, Morris ME, Graham HK: Management of upper limb dysfunction in children with cerebral palsy: a systematic review. Eur J Neurol 2001;8:15066 44. Woolfson L: Educational interventions for infants and preschool children with cerebral palsy: methodological difculties and future directions in evaluation research. Eur J Spec Needs Educ 1999;14:24053 45. Hur JJ: Review of research on therapeutic interventions for children with cerebral palsy. Acta Neurol Scand 1995; 91:42332 46. Horn EM: Basic motor skills instruction for children with neuromotor delays: a critical review. J Spec Educ 1991;25: 16897 47. Sackett DL: Rules of evidence and clinical recommendations on the use of antithrombotic agents. Chest 1986;89: 2535 48. Sackett DL: Rules of evidence and clinical recommendations on the use of antithrombotic agents. Chest 1989;95: 24 49. Butler C: AACPDM Methodology for Developing Evidence Tables and Reviewing Treatment Outcome Research. 1998 1999. Available at: www.aacpdm.org. Accessed March 20, 2003 50. van Tulder M, Cherkin D, Berman B, Lao L, Koes B: Acupuncture in low back pain, in: The Cochrane Library. Oxford, Update Software, 2003

June 2008

Effectiveness of Interventions in CP

491

51. Jadad AR, Moore RA, Carroll D, et al: Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996;17:112 52. Schulz KF, Chalmers I, Hayes RJ, Altman DG: Empirical evidence of bias. Dimensions of methodological quality associated with estimates of treatment effects in controlled trials. JAMA 1995;273:40812 53. Sackett DL: How to read clinical journals: V: to distinguish useful from useless or even harmful therapy. CMAJ 1981; 124:115662 54. Verhagen AP, de Vet HC, de Bie RA, et al: The Delphi list: a criteria list for quality assessment of randomized clinical trials for conducting systematic reviews developed by Delphi consensus. J Clin Epidemiol 1998;51:123541 55. Wells G, Shea B, OConnell D, et al: The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. Available at: http://www.ohri.ca/ programs/clinical_epidemiology/oxford.htm. Accessed 24th October, 2007 56. Law M, Stewart D, Pollock N, Letts L, Bosch J, Westmorland M: Guidelines for Critical Review FormQuantitative Studies. Available at: http://www.canchild.ca/Portals/ 0/outcomes/pdf/quantguide.pdf. Accessed 24th October, 2007 57. University of Alberta: Evidence-Based Medicine Tool Kit. Available at: http://www.ebm.med.ualberta.ca. Accessed 24th October, 2007 58. Jadad AR: Randomised Controlled Trials: A Users Guide. London, BMJ Books, BMJ House, 1998 59. Lonigan C, Elbert J, Johnson S: Empirically supported psychosocial interventions for children: an overview. J Clin Child Psychol 1998;27:13845 60. van Tulder MW, Assendelft WJJ, Koes BW, Bouter LM; Editorial Board of the Cochrane Collaboration Back Review Group: Method guidelines for systematic reviews in the Cochrane Collaboration Back Review Group for Spinal Disorders. Spine 1997;22:232330 61. Law M, Russell D, Pollock N, Rosenbaum P, Walter S, King G: A comparison of intensive neurodevelopmental therapy plus casting and a regular occupational therapy program for children with cerebral palsy. Dev Med Child Neurol 1997;39:66470 62. Steinbok P, Reiner AM, Beauchamp R, Armstrong RW, Cochrane DD: A randomized clinical trial to compare selective posterior rhizotomy plus physiotherapy with physiotherapy alone in children with spastic diplegic cerebral palsy. Dev Med Child Neurol 1997;39:17884 63. Law M, Cadman D, Rosenbaum P, Walter S, Russell D, DeMatteo C: Neurodevelopmental therapy and upper-extremity inhibitive casting for children with cerebral palsy. Dev Med Child Neurol 1991;33:37987 64. Palmer FB, Shapiro BK, Allen MC, et al: Infant stimulation curriculum for infants with cerebral palsy: effects on infant temperament, parent-infant interaction, and home environment. Pediatrics 1990;85:4115 65. Palmer FB, Shapiro BK, Wachtel RC, et al: The effects of physical therapy on cerebral palsy. A controlled trial in infants with spastic diplegia. N Engl J Med 1988;318:8038 66. Sommerfeld D, Fraser BA, Hensinger BN, Beresford CV: Evaluation of physical therapy service for severely mentally impaired students with cerebral palsy. Phys Ther 1981;61:33844 67. Scherzer AL, Mike V, Ilson J: Physical therapy as a determinant of change in the cerebral palsied infant. Pediatrics 1976;58:4752 68. Carlsen PN: Comparison of two occupational therapy approaches for treating the young cerebral-palsied child. Am J Occup Ther 1975;29:26772 69. Wright T, Nicholson J: Physiotherapy for the spastic child: an evaluation. Dev Med Child Neurol 1973;15:14663

70. McCubbin J, Shasby G: Effects of isokinetic exercise on adolescents with cerebral palsy. Adapt Phys Act Q 1985;2: 6575 71. Sung I-Y, Ryu J-S, Pyun S-B, Yoo S-D, Song W-H, Park M-J: Efcacy of forced-use therapy in hemiplegic cerebral palsy. Arch Phys Med Rehabil 2005;86:21958 72. DeLuca S: Intensive Movement Therapy with Casting for Children with Hemiparetic Cerebral Palsy: A Randomised Controlled Trial [dissertation]. Birmingham, The University of Alabama, 2002 73. Taub E, Ramey SL, DeLuca S, Echols K: Efcacy of constraint-induced movement therapy for children with cerebral palsy with asymmetric motor impairment. Pediatrics 2004;113:30512 74. ODwyer N, Neilson P, Nash J: Reduction of spasticity in cerebral palsy using feedback of the tonic stretch reex: a controlled study. Dev Med Child Neurol 1994;36:77086 75. Richards CL, Malouin F, Dumas F: Effects of a single session of prolonged plantarexor stretch on muscle activations during gait in spastic cerebral palsy. Scand J Rehabil Med 1991;23:10311 76. Tremblay F, Malouin F, Richards CL, Dumas F: Effects of prolonged muscle stretch on reex and voluntary muscle activations in children with spastic cerebral palsy. Scand J Rehabil Med 1990;22:17180 77. Dorval G, Tetreault S, Caron C: Impact of aquatic programmes on adolescents with cerebral palsy. Occup Ther Int 1996;3:24161 78. Cherng R, Liao H, Leung HWC, Hwang A: The effectiveness of therapeutic horseback riding in children with spastic cerebral palsy. Adapt Phys Act Q 2004;21:10321 79. Benda W, McGibbon NH, Grant KL: Improvements in muscle symmetry in children with cerebral palsy after equine-assisted therapy (hippotherapy). J Altern Complement Med 2003;9:81725 80. MacKinnon J, Noh S, Lariviere J, MacPhail A, Allan DE, Laliberte D: A study of therapeutic effects of horseback riding for children with cerebral palsy. Phys Occup Ther Pediatr 1995;15:1734 81. Reddihough DS, King J, Coleman G, Catanese T: Efcacy of programmes based on conductive education for young children with cerebral palsy. Dev Med Child Neurol 1998; 40:76370 82. McConachie H, Huq S, Munir S, Ferdous S, Zaman S, Khan N: A randomized controlled trial of alternative modes of service provision to young children with cerebral palsy in Bangladesh. J Pediatr 2000;137:76976 83. Hanzlik J: The effect of intervention on the free-play experience for mothers and their infants with developmental dealy and cerebral palsy. Phys Occup Ther Pediatr 1989; 9:3351 84. Talbot M, Junkala J: The effect of auditorally augmented feedback on the eye-hand coordination of students with cerebral palsy. Am J Occup Ther 1981;35:5258 85. Sellick KJ, Over R: Effects of vestibular stimulation on motor development of cerebral-palsied children. Dev Med Child Neurol 1980;22:47683 86. ODonnell M, Darrah J, Adams R, Butler C, Roxborough L, Damiano D: AACPDM Methodology to Develop Systematic Reviews of Treatment Interventions (Revision 1.1): 2004 Version. Available at: http://www.aacpdm.org/resources/ systematicReviewsMethodology.pdf. Accessed November 28, 2006 87. Centre for Evidence-Based Medicine: Levels of evidence. Available at: http://www.cebm.net/levels_of_evidence.asp. Accessed November 28, 2006 88. Sutton AJ, Abrams KR, Jones DR, Sheldon TA, Song F: Systematic reviews of trials and other studies. Health Technol Assess. 1998;2 89. Ju ni P, Witcshi A, Bloch R, Egger M: The hazards of

492

Anttila et al.

Am. J. Phys. Med. Rehabil.

Vol. 87, No. 6

scoring the quality of clinical trials in meta-analysis. JAMA 1999;282:105460 90. MacLehose RR, Reeves BC, Harvey IM, Sheldon TA, Russell IT, Black AMS: A systematic review of comparisons of effect sizes derived from randomised and non-randomised studies. Health Technol Assess 2000;4:1145 91. Dalziel K, Round A, Stein K, Garside R, Castelnuovo E, Payne L: Do the ndings of case series studies vary significantly according to methodological characteristics? Health Technol Assess 2005;9:1146 92. van den Ende CHM, Steultjens EMJ, Bouter LM, Dekker J: Clinical heterogeneity was a common problem in Cochrane reviews of physiotherapy and occupational therapy. J Clin Epidemiol 2006;59:9149 93. Rosenbaum PL, Walter SD, Hanna SE, et al: Prognosis for gross motor function in cerebral palsy: creation of motor development curves. JAMA 2002;288:135763 94. Kunz R, Autti-Ra mo I, Anttila H, Malmivaara A, Ma kela M: A systematic review nds that methodological quality is better than its reputation but can be improved in physiotherapy trials in childhood cerebral palsy. J Clin Epidemiol 2006;59:123948 95. Finger M, Cieza A, Stoll J, Stucki G, Huber E: Identication of intervention categories for physical therapy, based on the International Classication of Functioning, Disability and Health: a Delphi exercise. Phys Ther 2006;86: 120320 96. Haigh R, Tennant A, Biering-Sorensen F, et al: The use of outcome measures in physical medicine and rehabilitation within Europe. J Rehabil Med 2001;33:2738 97. Anttila H, Malmivaara A, Kunz R, Autti-Ra mo I, Ma kela M: Quality of reporting randomized, controlled trials in cerebral palsy. Pediatrics 2006;117:222230 98. Atkins D, Best D, Briss PA, et al: Grading quality of evidence and strength of recommendations. BMJ 2004;328: 14904 99. The denition and classication of cerebral palsy. Dev Med Child Neurol 2007;49:144 100. Campbell M, Fitzpatrik R, Haines A, et al: Framework for design and evaluation of complex interventions to improve health. BMJ 2000;321:6946

a search of unpublished or nonindexed literature (e.g., manual searches or letters to primary authors) 1 point: Cannot tell; search strategy partially comprehensive (e.g., at least one of the strategies in the foregoing section were performed) 0 points: No; search not comprehensive or not described well enough to make a judgment

Selection Methods
3. Were the criteria used for deciding which studies to include in the review reported? 2 points: Yes; inclusion and exclusion criteria clearly dened 1 point: Partially; reference to inclusion and exclusion criteria can be found in the paper but are not dened clearly enough to duplicate 0 points: No; no criteria dened 4. Was bias in the selection of articles avoided? 2 points: Yes; key issues inuencing selection bias were covered. Two of three of the following bias avoidance strategies were used: two or more assessors independently judged study relevance and selection using predetermined criteria, reviewers were blinded to identifying features of study (i.e., journal title, author(s), funding source), and assessors were blinded to treatment outcome. 1 point: Cannot tell; if only one of the three strategies above were used 0 points: No; selection bias was not avoided or was not discussed

APPENDIX A
Quality-assessment criteria for review articles. A modied version by Hoving at al.16 of an index constituted by Oxman and Guatt14 and Oxman et al.15 Maximum total score is 18.

Validity Assessment
5. Were the criteria used for assessing the validity for the studies that were reviewed reported? 2 points: Yes; criteria dened explicitly 1 point: Partially; some discussion or reference to criteria but not sufciently described to duplicate 0 points: No; validity or methodological quality criteria not used or not described 6. Was the validity for each study cited assessed using appropriate criteria (either in selecting studies for inclusion or in analyzing the studies that are cited)? 2 points: Yes; the criteria used address the major factors inuencing bias (for example: population, intervention, outcomes, follow-up) 1 point: Partially; some discussion of methodological review strategy, but not Effectiveness of Interventions in CP

Search Methods
1. Were the search methods used to nd evidence (primary studies) on the primary question(s) stated? 2 points: Yes; includes description of databases searched, search strategy, and years reviewed. Described well enough to duplicate. 1 point: Partially; partial description of methods, but not sufcient to duplicate search 0 points: No; no description of search methods 2. Was the search for evidence reasonably comprehensive? 2 points: Yes; must include at least one computerized database search as well as June 2008

493

clearly described with predetermined criteria 0 points: No; criteria not used or not described

Synthesis
7. Were the methods used to combine the ndings for the relevant studies (to reach a conclusion) reported? 2 points: Yes; qualitative or quantitative methods are acceptable 1 point: Partially; partial description of methods to combine and tabulate; not sufcient to duplicate 0 points: Methods of combining studies not stated or described 8. Were ndings of the relevant studies combined appropriately relative to the primary question the review addresses? 2 points: Yes; combining of studies seems acceptable 1 point: Cannot tell; should be marked if in doubt

0 points: No; no attempt was made to combine ndings, and no statement was made regarding the inappropriateness of combining ndings; should be marked if a summary (general) estimate was given anywhere in the abstract, the discussion, or the summary section of the paper, and the method of deriving the estimate was not described, even if there is a statement regarding the limitations of combining the ndings of the studies reviewed 9. Were the conclusions made by author(s) supported by the data or analysis reported in the review? 2 points: Yes; data, not merely citations, were reported that support the main conclusions regarding the primary question(s) that the overview addresses 1 point: Partially 0 points: No; conclusions not supported or unclear

494

Anttila et al.

Am. J. Phys. Med. Rehabil.

Vol. 87, No. 6

APPENDIX B Study designs, populations, interventions, outcomes, and results on children with cerebral palsy in 21 systematic reviews
Studies on Children with Diagnosed CP Study Designs (n) Population Interventions, Length (L), and Settings (S) Outcome Measures Results

June 2008
10 5 studies with control groups, 5 studies without control groups (n 299) Diagnosed CP Age: 3 mos to 14 yrs NDT, facilitation and functional interventions: short-leg casting, sensory integration, Rood and proprioceptive neuromuscular facilitation and selective posterior rhizotomy Control groups: Regular OT sessions, infant stimulation program L: 7 days to 12 mos, sessions: from two visits to daily sessions S: Home or nr 21 7 RCT, 3 cohort studies, 3 before after case series, 4 multiple crossover trials, 4 case series (n 471) Children with CP (spastic, ataxic, athetoid, hypotonic, di-, hemi-, tri-, or quadriplegic; mild, moderate, or severe; mobile and nonmobile) Age: 115 yrs NDT individual therapy home program, intensive NDT individual therapy, Bobath method, facilitation group therapy, neurophysiologic individual therapy Control groups: Untreated period, functional therapy, traditional therapy, play, infant stimulation, motherchild interaction instruction, NDT with lesser intensity, CE, skill practice, functional skills OT L: 25 days to 12 mos S: nr Ambulatory status, Bayley Scales (motor, mental), clinical observation, COPM, DDST, ne motor skills, gait measurements, Gesell Developmental Schedules, GMFM, goniometry, MAS, modied MCMDST and GS, Motor Development Evaluation Form, neurological signs scale (nonstandardized), PFMS, PCI, QUEST, ROM, SBIS, self-care evaluation (nonstandardized), strength, the Hollingshead Four-Factor Index for Social Position, the Maternal Observation Interview, Vanguard spirometer, Videotape, VSMS Attainment or observation dened skills, Bayley Scales (mental, motor), Biofeedback instrument, bracing recommended, CITQ, COPM, DDST, DMIB, GM, GMFM, goniometer, HOME, kinematic analyses, neurological examination, PAS, pedographs, PFMS, Q (parent satisfaction), Q (social activities and home management), Q (motor development), QUEST, rate of movements, rated observation (of automatic reexes, GM activities, position), Rating Scale, RMCRE, ROM Scale, surgery recommended, video analysis, VSMS, WDRP Bayley Scales (mental, motor), Childrens Hands Skills Survey, DDST, GM evaluation, Hartwell Motor Age Test, Minear (1956) Classication system, Motor Development Evaluation Form, Physical therapy (based on the work of Holt 1976), Preschool Functional Activity Test, Standardised Neurologic Examination, WDRP 9 6 studies with control groups, 3 descriptive studies (n 451) Children with diagnosed CP Age: 5 yrs 5 4 RCTs, 1 other design (n 174) Children with CP Age data not available PT, unspecied individualized treatment, facilitation, neurophysiologic PT and parent-provided therapy, direct treatment, early PT using Vojta Control groups: Functional approach, passive ROM exercises, developmental stimulation by parents, no treatment, supervised management by teacher or aide, late PT using Vojta L: 6 wks to 4 yrs S: Clinic, home, or nr Bobath NDT, some rehabilitation and surgery, NDT conducted by parents Control groups: Bobath NDT by teachers, no treatment, rehabilitation program (nature unspecied), passive ROM exercises, Vojta by parents, therapy by parents L: 612 mos S: Home or nr GM milestones: Sitting, crawling, walking, home management (nonstandardized), motor and mental quotients, motor function (nonstandardized), neurological status (nonstandardized), ROM, social development quotient Inconsistent results. Positive results for NDT compared with no treatment, NDT by parents compared with passive ROM exercises, and treatment ensuing before compared to subjects 6 mos of age treated with Vojta technique by parents (three studies). Two studies reported no effects for NDT (Bobath NDT by physiotherapist compared to NDT by parents, and NDT compared with stimulation by parents with supervision by therapists)

Review (Year)

Number of Studies, Excluded Studies, a and Reasons for Exclusion (n)

Comprehensive physiotherapy Brown and Burns26 (2001)

17 7 excluded Non-CP population (suspected CP or high-risk infants) (n 519)

Inconsistent results. Six studies reported a benet from using the NDT approach. Three studies reported no benet, and one study did not specically examine effects of NDT

Butler and Darrah27 (2001)

21

Parette et al.28 (1991)

13 4 excluded Non-CP population (n 102)

NDT: Immediate improvement in dynamic ROM, no consistent evidence that NDT changed abnormal motoric responses, slowed or prevented contractures, facilitated more normal motor development or functional motor activities. No clear evidence that NDT produced other benets such as enhancement of social emotional, language, or cognitive domains of development, better home environments, improved parent-child interactions, or greater parent satisfaction. No benet of more intensive therapy Inconsistent results. Seven studies reported positive ndings of PT, unspecied individualized treatment, facilitation, individual therapy or neurophysiologic PT of early PT using Vojta, while there were no differences between groups in two studies that compared direct treatment with supervised, and PT with no treatment

Effectiveness of Interventions in CP

Tirosh and Rabino29 (1989)

9 4 excluded Non-CP population (n 195)

495

496
Studies on Children with Diagnosed CP Study Designs (n) Population Interventions, Length (L), and Settings (S) Outcome Measures Results 9 1 RCT, 8 other designs (repeated measures, single group) (n 124) Patients with CP (spastic, ataxic, dystonic and mixed types; di-, hemi-, and quadriplegic; from ambulatory to nonambulatory or not specied, or CP-ISRA classes 78 Age: 420 yrs Isokinetic, concentric and eccentric, isometric, free-weight, or xed-weight exercises L: 610 wks, sessions 37 times per week. Repetitions ranged from 1 set and 6 repetitions to 34 sets and 510 repetitions S: Home-, clinic-, or community-based programs 12-min wheelchair test, dynamic tapping, EEI, GMFM, heart rate, MAS, perceived competence, ROM (ankle, knee), SelfPerception Prole for Adolescents, SelfPerception Prole for Children, spasticity, strength with isokinetic dynamometer, (change in weight over 6RM, maximum isometric contraction, maximum voluntary contraction, mean isometric maximal contraction, rate of torque development, peak torque and work) 7 1 RCT, 6 case series (n 74) Children and adolescents with CP Age: 626 yrs Isokinetic, concentric, eccentric, or isometric exercises with Cybex, free weights, KIN-COM, Nautilus equipment, training machine, pulley, or free weights Control groups: Isokinetic training without resistance, or noneL: 210 wks S: nr Cybex II (peak torque, torque development, endurance, movement time), degree of knee exion at heel strike, energy expenditure, gait analyses, GMFM, handheld dynamometer (quadriceps and hamstring), KIN-COM (peak torque, work of quadriceps and hamstrings), obstacle course video, ROM (knee extension), stride length, tensiometer (quadriceps), timed ne motor tasks 3 2 RCT, 1 CCT (n 94) Children with CP, hemiplegic, or asymmetric involvement of the upper extremities Age: 7 mos to 8 yrs Casting of the less involved hand and training (shaping), fabric glove with a built in stiff volar plastic splint and a motor learning treatment program, scotchcast from below elbow to ngertips and individualized functional OT, stretching, and ADL practices L: 27 days/wk, 3 wks to 2 mos S: Preschool or nr Assisting Hand Assessment, Box and Blocks test, CAUT, PMAL, EDPT, EBS, QUEST, WeeFIM 1 RCT: No effects on QUEST or CAUT, positive effects on PMAL frequency of use subscale and EBS at 3 and 6 wks 1 CCT: Improvement in bimanual performance at 2 and 6 mos 1 RCT: Improvement in self care component on the WeeFIM at 6 wks. No effects in all other measures

APPENDIX B Continued

Review (Year)

Number of Studies, Excluded Studies, a and Reasons for Exclusion (n)

Anttila et al.

Strength training Dodd et al.30 (2002)

10 1 excluded Population aged 1247 (n 6) 4 reviews

Darrah et al.31 (1997)

Impairment: Increases in muscle strength (effect size range from 1.16 to 5.27) (8 of 9 studies), nonsignicant results (1 study, small population). No change in muscle spasticity (2 studies), increase in ROM (3 of 4 studies), no change in ROM (1 of 4 studies) Activity (4 studies): Increases in walking, running, and jumping (GMFM dimension E) (2 studies), and in standing (1 study), contradictory results on walking speed (2 studies) Participation: Not measured Contextual factors: Nonsignicant and positive results with individual exercise programs (6 studies), positive results with group program (1 study), nonsignicant and positive results of home- and clinicbased programs (3 studies), positive results of a community-based program (1 study) All studies reported positive effects on muscle strengthening on various outcomes (movement time, increased rate of torque development, quadriceps strength, stride length, peak torque and work, strength and speed of movement)

Am. J. Phys. Med. Rehabil.

Constraint-induced movement therapy Hoare et al.32 (2007)

Vol. 87, No. 6

June 2008
Studies on Children with Diagnosed CP Study Designs (n) Population Interventions, Length (L), and Settings (S) Outcome Measures Results 4 1 crossover cohort study, 3 case studies (1 case series, 1 ABA, 1 BA design (n 45) 33 healthy Spastic quadriplegia, diplegia, hemiplegia. Severity: not provided, could not sit independently, all could walk or GMFCS levels I and II Age: 2.515 yrs NDT (to improve trunk and shoulder control, smoothness and efciency of movement, ability to initiate reach); rocker platform movements; massed practice on a moving platform; microprocessor-controlled articial saddle riding using Brunel Active Balance Saddle Control groups: Practice or none L: 45 sessions during 1 wk, nr, 46 perturbations per minute with rest after 20 25 mins, 100 perturbations per day for 5 days S: nr Manual stretch, sinusoidal stretch by specically designed apparatus, standing in tilt-table L: Manual stretch with hold 60 secs or 15 20 mins, 35 repetitions for each movement or joint, 1 time per week to 3 times per day, 5 days/wk; sinusoidal stretch for 30 mins per session, 3 times per week for average 42 days; standing in tilt-table 30 mins S: nr Area (where control became decient during static sitting), autoregressive modeling for feetrelated displacement, center-of-pressure sway area per second, GMFM (standing), independent sitting for 30 secs, Kinematic analysis (Watsmart) of displacement of the head and trunk during reach, manual perturbation and voluntary movement; modied Posture Assessment Scale scored from video (head, neck, shoulder, scapula and trunk items only), postural sway, time to stabilization after a perturbation Externally generated movement on development of postural control: Improved reactive balance, GMFM (1 study), postural sway (1 study), independent sitting balance segmental level of control (1 study) NDT vs. practice (1 wk): no effects on posture and kinematic analyses 5 3 RCT , 1 beforeafter design, 1 multiple single-subject design (n 69) 20 controls Children with CP (spastic, diplegia, hemiplegia, quadriplegia) Age: 319 yrs Passive torque (ankle joint), EMG, goniometer, Kin-Com dynamometer and surface electrodes (in terms of torque, ankle angle and EMG), Spastic Locomotion Disorder Index, specially designed apparatus (measuring passive hip adduction angle, hypertonicity), tonic stretch reex, video recording (gait analysis) ROM: One RCT showed no between-group differences in ROM (triceps surae). Observational studies showed reduction in knee exion contracture (one study), or no reduction in passive hip abduction (one study) Spasticity: Three RCTs and one singlesubject study showed a reduction in spasticity Gait: No changes in gait pattern after 30 mins of stretching on a tilt table (one study) Balance and equilibrium skills, bilateral activities, function and coordination of extremity during bilateral activities, GMFM, Leisures Activity Inventory, Matrinek-Zaickowsly Self-Concept Scale, promotion of positive self-image. Rosenbergs Self-Esteem Scale, ROM (shoulder joint), SWIM, VC, water orientation Aquatic exercise (1 time per week) by Halliwick method, combined land and aquatic exercise, swimming and water safety instruction Control groups: Aquatic exercise (2 times per week) or none L: 6 wks to 6 mos; 1 time per week, sessions: 3055 mins S: nr Inconsistent results. No differences between conventional aquatics program and adapted aquatics program (one RCT). One study: Likely improvements in function, improvement in orientation skills, no improvement of self-concept. Case reports: Improvement in SWIM, GMFM, standing, balance, VC, shoulder exion and abduction ROM, arm swing during ambulation, use of right extremity in ADL, back and side stroke, and crawl Children or adolescents with CP, spastic, di-, hemi-, quadriplegia, ataxia, athetosis Age: 517.3 yrs 5 1 RCT, 1 quasiexperimental study (same study in 2 reports), 2 case reports (n 68)

APPENDIX B Continued

Review (Year)

Number of Studies, Excluded Studies, a and Reasons for Exclusion (n)

Postural control Harris and Roxborough33 (2005)

12 8 excluded Interventions on seating device, orthoses. or lycra garment (n 100)

Soft-tissue treatment Pin et al.34 (2006)

7 2 excluded Non-CP population (spasticity in lower limbs, severe physical and cognitive impairment, and decreased joint ranges of lower limbs) (n 20)

Effectiveness of Interventions in CP

Hydrotherapy Getz et al.35 (2006)

11 6 excluded Non-CP population (neurological dysfunction, spinal muscle atrophy, Rett syndrome, progressive muscular dystrophy, highrisk infants) (n 59)

497

498
Studies on Children with Diagnosed CP Study Designs (n) Population Interventions, Length (L), and Settings (S) Outcome Measures Results 9 3 RCT, 4 quasiexperimental, 2 case studies (n 99) 9 nondisabled controls Children with CP (spastic, di-, and hemiplegia, moderate to severe) Age: 2.312 yrs Hippotherapy (and riding therapy) by physical or occupational therapist, therapeutic horseback riding (THR) by a trained riding instructor L: From 1 session (8 mins) to 6 mos; sessions from one 8-min visit to 1 hr, 2 times per week S: nr Bertoti Posture Assessment Scale, BOTMP, CBC, EMG (trunk ab/adductors), energy expenditure, gait (velocity, cadence, average stride length), GMFCS, GMFM, HSPS, kinematic analyses by peak 5 motion analyzer, MAS, PDMS, PEDI, Qualitative information, trunk coordination (kinematics), VABS ADL/ socialization, WeeFIM 9 3 cohort designs, 1 single-case design, 2 beforeafter designs, 3 case studies (n 85) 7 without CP Children with CP (spastic, mild to severe, ambulatory to nonambulatory, di-, hemi, and quadriplegia) Age data not available Hippotherapy by licensed health professionals, recreational horsebackriding therapy by riding instructors L: 626 wks; sessions for 45 120 mins, 1 time per week and 3060 mins, 2 times per week S: nr GMFM, kinematic measurement from videography, PEDI, test for scoring posture with a four-point scale for ve measurements (head/neck, shoulder/scapula, trunk, spine, pelvis), the riders lateral trunk displacement, weight bearing of arms and legs 14 1 RCT, 1 CCT, 3 cohort studies with concurrent controls, 2 cohort studies with historical controls, 7 case series (n 1038) Children with CP (di-, hemi-, and quadriplegic; mild, moderate, or severe; spastic or dystonic; intellectual disability: mild, moderate, severe, or normal) Age: 113 yrs CE by physical or occupational therapist, teacher, professional trained therapist teachers, conductors, nurses, mothers, caregivers. Rhythmical intention Control groups: Training programs in orthopedic residential schools, centerbased early intervention, individual PT, special education, individual therapy, traditional therapy, or none L: nr, session lengths ranged from 2.8 hrs/ wk to 13.5 hrs/day S: Various schools, Peto Institute, spastic centre, Move and Walk Institute Basic Math test, CMMS, CRT, DMT, DP2 CAS, DP2 PAS, DP2 SHAS, DSI, ECFAT, GMFM, survey to parents, observation, parent perception of goal achievement, PIAT, PPVT, QRS, rating scale, RDLS, SB subtest, task analysis, task series, VAB BR (video), VAB CR, VABS CE, VABS IE, video ratings, VLDS, WBSI, WPPSI

APPENDIX B Continued

Review (Year)

Number of Studies, Excluded Studies, a and Reasons for Exclusion (n)

Anttila et al.

Hippotherapy Snider et al.36 (2007)

Sterba37 (2007)

11 2 excluded Non-CP population (various developmentally delayed children, down syndrome, autism, spina bina, and traumatic brain injury) (n 26)

Hippotherapy: Increased muscle symmetry of the trunk and hip ab/adductors (one RCT); improved posture (one study), GMFM-E (one study), GMFM and PEDI (one study). energy expenditure (one study), and trunk coordination (one study) Therapeutic horseback riding: Positive changes in GMFM, no changes in muscle tone (one RCT), improvements in grasping (PDMS), but not for posture, self-esteem, or global behavior (one RCT); improvement in GMFM (one study) and more likely to respond with normal equilibrium reactions to the pelvic displacement of the horse in diplegia, than in children with quadriplegia (one study) Hippotherapy (5 studies): Improved posture (one study), weight bearing (one study), gross motor function (GMFM-E) (one study), coordination and functional mobility (PEDI) (one study), and in functioning (GMFM and PEDI) (one study) HBRT (3 studies): Inconsistent results. No changes in GMFM (one study), improvements in walking, running, and jumping (GMFM dimension E) (one study) and in GMFM total score (one study) The majority of the results of the methodologically stronger studies reveal no difference in outcome between CE intervention group and the control group or prepost CE group results (of 20 outcomes, 10 favored CE and 10 favored the control group; no outcome of interest showed improvement in the CE group across studies). The majority of outcomes of interest in methodologically weaker studies showed some improvement for the CE group.

Conductive education Darrah et al.38(2003)

Am. J. Phys. Med. Rehabil.

15 1 excluded Non-CP population (mothers of children with CP) (n 36)

Vol. 87, No. 6

APPENDIX B Continued
Studies on Children with Diagnosed CP Study Designs (n) 6 1 RCT, 3 CCTs, 2 descriptive studies (n 203) Children with CP Age: 112 yrs CE, or programs based on CE L: Ranged from 26 wks to 2 yrs; sessions from 3 hrs/wk to 36 hrs/day (nr for all studies) S: UK Birmingham- and Manchester-area special schools, and 5 Australian schools in Tongala and Melbourne, Victoria; early interventions in Brisbane CE or programs based on CE principles L: 26 wks to 2 yrs; sessions: 13.5 hrs/day (1 study). Other session lengths nr S: nr nr CMMS, GMFM, Parenting Stress Index, PPVT, QRS, RDLS, VAB, WPSSI Population Interventions, Length (L), and Settings (S) Outcome Measures Results

June 2008
8 1 RCT, 7 studies with control groups (n 243) Children with CP Age: 115 yrs (age nr from all studies) 5 5 quasi-experimental studies (n 244) Children with CP Age: 313 yrs nr from all studies. Feeding and drinking program, daily program, plinth, sitting, hand, standing and walking programs L: 622 mos S: nr Bayley Scales, checklist of skills (basic motor, ne nger, GM, eating, grooming, helping, social responsibilities, nonstandardized), CWPAC, Comparative Appraisals Scale, ECFAT, EDPT, modied VAB, ParentChild Interaction Scale Q (nonstandardized), QRS, RDLS, RDLSZinkin, SBIS, VAB 8 6 RCTs, 1 CCT, 1 other design. (n 334) CP (spastic, diplegic, or nr) Age: 1.58 yrs Intensive NDT with or without cast, individually sensory perceptual motor training (SPM) home program, tracing auditory feedback, dressing and undressing during play, facilitation sensory organization postural stability, verbal instruction, distance training urban/rural Control groups: Regular NDT with or without cast, regular OT, group SPM home program, tracing without feedback, functional approach self-care, NDT instruction, motherchild group training urban/rural, or none L: Sessions varied from 2 10 mins, 5 times per week, to 3 1.5 hrs, 1 time per week; intervention length: 4 wks to 9 mos S: Outpatients, home or school Bayley Scales (motor), DDST, Independent Behaviour Assessment Scale (IBAS), KleinBell scale, observation of independent play, PFMS, Physical ability test, QUEST, SCMAT

Review (Year)

Number of Studies, Excluded Studies, a and Reasons for Exclusion (n)

Ludwig et al. (2000)

39

9 3 excluded Non-CP population (parents) (n 57)

Pedersen40 (2000)

Inconsistent results. No differences between the groups (three studies), improved motor performance and parental coping in CE group, improved cognitive variables in control group (one study); positive results in selected skills (one study); in mobility and eating skills (one study); decreased mobility in CE group (one study) Inconsistent results. Positive results in favor of CE on social skills (one study), motor skills , parental coping (one study) and motor, social and cognitive skills (one study); no effects (ve studies) Inconsistent results. From gross motor, ADL, play, language, and personal/social skills and parent outcomes one variable of each supported CE signicantly. From gross motor two variables and from ADL and personal/social skills and parent outcomes one variable supported comparison groups signicantly. Nonsignicant differences where in one GM and play variables, in two cognitive, language and personal/social skills variables, and three ne motor and ADL variables NDT casting: No between-group differences in dexterity or upper extremity function (two RCTs) Sensorimotor functions: Nonsignicant result (one RCT), nr (one CCT) Training of skills: No improvement in functional ability (one study) Training of sensorimotor function vs training of skills: Nonsignicant results on motor skills (one RCT) Parental counseling: Nonsignicant results on functional ability (two RCTs)

French and Nommensen41 (1992)

9 1 excluded Non-CP population (children with mental retardation or multiple handicaps) (n 10) 6 1 excluded Non-CP population (children with profound, multiple handicaps) (n 10)

Various interventions Steultjens et al.42 (2004)

Effectiveness of Interventions in CP

17 9 excluded Interventions on assistive devices or splints (n 102)

499

500
Studies on Children with Diagnosed CP Study Designs (n) 13 3 RCTs, 10 other designs with prospective data (n ?) Children with CP (quadri-, di-, and hemiplegic) Age: 1.518 yrsb PT, OT, NDT, motor learning, CE, strength training, constraint-induced therapy Control groups: nr L: nr S: nr COPM, PEDI, PFMS, QUEST, ROM (wrist), VAB Population Interventions, Length (L), and Settings (S) Outcome Measures Results 9 1 RCT (in 2 articles), and 8 quasiexperimental designs (n 399) Children with CP (mild to severe spastic diplegia in 2 studies; data nr for others) Age: 17 yrs Group 1: Multidomain developmental stimulation programs (children placed in conned space to inhibit abnormal motor movements and to promote social and language development), NDT, motor, cognitive, language, and sensory activities stimulation program, interdisciplinary developmental stimulation program (gross and ne motor, communication, cognitive, social, attention and self-help) Group 2: CE compared with traditional special education, individual PT, aimoriented management, eclectic PT and NDT L: 412 mos (group 1), 124 mos (group 2) S: nr Bayley scales, CITQ, DP 2, GMFM, HOME, MFD, movement scale (nonstandardized), observation and parental report of motor skill attainment, Parent Q, PTI, QRS, RMCRE, Schedule of Growing Skills, semistructured interviews, VAB, VSMS 19 5 RCTs, 1 nonrandomized group comparisons, 3 beforeafter treatment comparisons, 4 descriptive studies, 4 single case, 2 case studies (n 270) Children with CP (spastic, athetosis type, hemiplegic, triplegic, quadriplegic, dyskinesia, all categories, or nr) Age: 8 mos to 22 yrs Bobath PT, facilitation, NDT, vestibular stimulation, direct therapy, semicircular canal therapy, interaction sessions, balance training orthosis, exercise, PT head support device, visual training, reverse tailor sitting, head positioning, swimming L: 4 wks to 6 yrs (nr for all studies). Sessions: 15 mins, 4 times per week (1 study); 30 mins, 2 times per week (1 study). Other session lengths nr S: nr nr

APPENDIX B Continued

Review (Year)

Number of Studies, Excluded Studies, a and Reasons for Exclusion (n)

Anttila et al.

Boyd et al.

43

(2001)

56 43 excluded Interventions on orthoses, splints, lycra suits, special seating, electrophysical agents, or drugs (n ?)

Woolfson44 (1999)

Hur 45(1995)

37 18 excluded Non-CP population (evidence of brain insults, signs of cerebral neuromotor disturbances, parents); biofeedback, casts, braces, oral motor control, surveys (n 457)

Inconsistent results Impairment outcomes: Positive ndings (ve studies), no difference between groups (two studies) Activity outcomes: Positive results (seven studies), no difference between groups (three studies) Participation outcomes: Positive effects (two studies), no difference between groups (three studies) Group 1: Improvement in infant stimulation group in some outcomes compared with NDT group, no differences in psychosocial variables (one RCT); progress in motor activity, ne motor skills, social skills, communication, comprehension, visual and auditory understanding, interactive social, and self-help (four studies) Group 2: No between-group differences (three studies); greater mothers satisfaction with help in CE group, improvement in object transfer, postural independence, hip mobility, form discrimination and activities of daily living in control group (one study); improvements in motor skills, ADL and parental coping in CE groups, and in cognitive skills and social interaction in the control group (one study) RCTs: No between-group differences (three RCTs), improvement in all areas (one RCT), or in motor, social, and home management (one RCT) Other designs: Positive resultsin various outcomes (13 studies), negative results (one study)

Am. J. Phys. Med. Rehabil.

Vol. 87, No. 6

June 2008
Studies on Children with Diagnosed CP Study Designs (n) 26 6 RCTs and 2 comparison group designs, 13 singlesubject designs, 5 case studies (n ?) Severe (51%) and moderate (34%) motor disabilities. High tone problems (70%), quadriplegia (66%) Age data not available Training of ne and GM skills (head control, global motor functions, reexes, weight shift/bearing, postural stability, upper extremity) L: nr S: nr Bayley scales (motor), Cattel Scale, developmental measure of motor function reexes, developmental screening (with/without reliability), direct observation, EIDP, EMG recording, GM evaluation, Kreutzberg Reex, mechanical count (no validity check), mechanical device, motor development, motor milestones, Motor Skills Test, neurological examination, PFMS, reex test, ROM scale, SBIS, subject report, VSMS, WDRP Population Interventions, Length (L), and Settings (S) Outcome Measures Results Neuromotor interventions: No difference (ve studies), positive results (two studies), three of seven children beneted (one study), no short-term advantages (one study) Sensory stimulation: No differences (one study), improvements (three studies) Behavior programming: No improvements (one study); improvements (11 studies) Natural context treatment: Positive results in accelerated acquisition (one study), and in general positive changes (one study)

APPENDIX B Continued

Review (Year)

Number of Studies, Excluded Studies, a and Reasons for Exclusion (n)

Horn46 (1991)

28 2 excluded Non-CP (high risk infants) (n ?)

Descriptions of study designs are reported here as described in the reviews; b characteristics of population reported only for RCTs. ADL, activities of daily living; CP, cerebral palsy; RCT, randomized controlled trial; CCT, clinical controlled trial; NDT, neurodevelopmental therapy; CE, conductive education; PT, physiotherapy; OT, occupational therapy; nr, not reported; CP-ISRA, Cerebral Palsy International Sport and Recreation Association Classication System (eight levels); EMG, electromyography; GM, gross motor; Q, Questionnaire; BOTMP, Bruininsks-Oseretsky Test of Motor Prociency; CAUT, Child Arm Use Test; CBC, Child Behavior Checklist; CHQ, Child Health Questionnaire; CITQ, Carey Infant Temperament Questionnaire; CMMS, Columbia Mental Maturity Scale; COPM, Canadian Occupational Performance Measure; CRT, Comprehensive Reading Test; CWPAC, Cheyne Walk Physical Ability Chart; DDST, Denver Developmental Screening Test; DMIB, Dictionary of Mother-Infant Behaviours; DMT, Diagnostic Mathematical Task; DP2 CAS, Developmental Prole 2 Communication Age Scale; DP2 SAS, Developmental prole 2 Social Age Scale; DP2 SHAS, Developmental Prole 2 Self-help Age Scale; EBS, Emerging Behaviours Scale; ECFAT, Eau-Claire Functional Abilities Test; EDPT, Erhardt Developmental Prehension Test; EEI, Energy Expenditure Index; EIDP, Early Intervention Development Prole; GMDS, Grifths Mental Developmental Scale; GMFCS, Gross Motor Classication System; GMFM, Gross Motor Function measure; GS, Gidoni Scale of Gross Motor Development; HOME, Home Observation for Measurement of the Environment; HSPS, Harter Self-Perception Scale; MAS, Modied Ashworth scale; MCMDST, Milani-Comparetti Motor Development Screening Test; MFD, Munich Functional Diagnostic; MI, Malaise Inventory; PAS, Postural Assessment Scale; PCI, Physiological Cost Index; PDMS, Peabody Developmental Motor Scales; PEDI, Pediatric Evaluation of Disability Inventory; PFMS, Peabody Fine Motor Scale; PIAT, Peabody Individual Achievement Test; PMAL, Pediatric Motor Activity Log; PPVT, Peabody Picture Vocabulary Test; PTI, Pictorial Test of Intelligence; QRS, Questionnaire on Resources and Stress; QUEST, Quality of Upper Extremity Skills Scale; RDLS, Reynell Developmental Language Scale; RM, repetition maximum; RMCRE, Roth Mother-Child Relationship Evaluation; ROM, range of motion; SCMAT, Southern California motor accuracy test; SBIS, Stanford-Binet Intelligence Scale; SWIM, Swimming with Independent Measurement; VAB, Vulpe Assessment Battery (BR, behavior rating; CR, caregiver rating); VABS, Vineland Adaptive Behaviour Scales (IE, Interview edition; CE, classroom edition); VC, Vital capacity; VLDS, Verbal Language Developmental Scale; VSMS, Vineland Social Maturity Scale; WeeFIM, Functional Independence Measurement for Children; WBSI, Wolfe-Bleuel Sozialization Inventory; WDRP, Wilson Developmental Reex Prole; WGME, Wolanski Gross Motor Evaluation; WPSSI, Weschler Pre-School Scale of Intelligence (revised).

Effectiveness of Interventions in CP

501