Botulinum toxin type A and casting versus botulinum toxin type A or casting in treating equinus gait of children with cerebral palsy

Abstract

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

The primary purpose of this review is to determine if the combination of BoNTA and casting results in better outcomes than if casting alone or BoNTA alone or no intervention is carried out for children with CP presenting with equinus gait. The review considered outcomes from domains of body structure and function , activity limitation, as well as participation/contextual factors.

Background

Cerebral palsy (CP) refers to "a group of permanent disorders of the development of movement and posture, causing activity limitation, that are attributed to non progressive disturbances that occurred in the developing fetal or infant brain. The motor disorders of cerebral palsy are often accompanied by disturbances of sensation, perception, cognition, communication, and behaviour; by epilepsy, and by secondary musculoskeletal problems" (Rosenbaum 2006). CP is the most common physical disabling condition in childhood (Pirpiris 2006) occurring in 2‐2.5 per 1,000 live births (Stanley 2000). The incidence of CP is higher in premature infants (Stanley 2000) and in twin births (Nelson 1995).

Equinus gait (toe‐walking) is a common gait deviation in children with CP (Preiss 2003) and is due to the presence of spasticity in the calf muscles and decreased range of movement at the ankle joints. Equinus deformity is defined as a limitation of passive dorsiflexion leading to increased plantarflexion during stance and walking. The deformity can be dynamic or fixed. Dynamic equinus is attributed to spasticity in the gastrocnemius‐soleus and an imbalance between the action of this muscle and the tibialis anterior (Blackmore 2007). It is seen more commonly in younger children (Boyd 2000). Fixed equinus is caused by muscle contracture, attributed to faster growth in the tibia than the spastic calf muscle‐tendon (Blackmore 2007).The equinus deformity results in an unstable and inefficient gait pattern. Failure to address this problem at an early stage results in decreased walking ability and an increased risk of developing deformities. Goldstein (Goldstein 2001) observed that equinus deformity increases the risk of pain and early cessation of walking.

There are many management  approaches in the treatment of dynamic equinus including physiotherapy (i.e. strengthening and gait training) in combination with orthoses and night splinting (Boyd 2000). In the past 15 ‐ 20 years, botulinum toxin type A (BoNTA) and casting, separately or combined, have been increasingly used to affect change in both tone and muscle length in an attempt to increase range of motion and improve gait and function. These approaches address both neural and non‐neural components of hypertonia.

Casting (serial or continuous) provides the opportunity for the physiological adaptation of the muscle to prolonged stretch (Brouwer 1998). Serial casting refers to a sequence of consecutive below‐knee cast applications and removals, each one setting the joint in another increment of desired range of movement. Tissue elongation is achieved through physiological adaptation to prolonged stretch, each cast contributes to an incremental increase into the desired range of motion (Blackmore 2007).Several studies have investigated the effectiveness of casting in children with CP (Watt 1986, Brouwer 1998, Glanzman 2004, McNee 2007). Three weeks of serial casting in seven children with CP resulted in increased dorsiflexion ROM post casting which persisted at six‐week follow‐up (Brouwer 1998). Watt and colleagues (Watt 1986) reported significant changes in dorsiflexion range of motion (ROM) and gait immediately after three weeks in a single cast; however, these changes were not present at a five‐month follow‐up. Similarly, a retrospective study of 20 children with CP showed that serial casting was effective in improving ROM and gait, again, the effects being temporary (Cottalorda 2000). A recent crossover trial of serial casting in nine children with CP also demonstrated improvements in active and passive dorsiflexion ROM that dissipated by the 12‐week follow‐up and were not accompanied by functional changes (McNee 2007). Overall, it would appear that using a single cast or a series of casts elicits a positive, but temporary, impact on dorsiflexion ROM without significant effects on gait velocity (Ackman 2005) walking distance or function (McNee 2007).

Botulinum toxin A (BoNTA) has also been used in the management of equinus gait. Botulinum toxin is produced mainly by a bacteria called clostridium botulinum. The toxin is synthesized in seven serotypes, designated A‐G (Simpson 1996). The most commonly used serotype is type A however serotypes B and F have also been used clinically (Aoki 2001). When injected into the muscle, BoNTA blocks the release of acetylcholine, causing temporary paralysis of the muscle and a localized reduction in spasticity (Fehlings 2005). The response occurs within 2‐3 days, the effect reaching its maximum at two weeks and waning by 3‐6 months (Fehlings 2005). Studies examining the effectiveness of BoNTA in improving equinus gait have reported significant decreases in muscle tone (Corry 1998; Love 2001; Reddihough 2002, Satila 2008, Hu 2009, Lee 2009, Tedroff 2010 ) significant improvements in active (Sutherland 1999, Koman 2000; Satila 2008, Hu 2009) and passive ROM (Corry 1998, Sutherland 1999, Reddihough 2002, Baker 2002, Hu 2009, Lee 2009) and significant improvements in gait (Koman 1994, Corry 1998, Koman 2000, Lee 2009; Balaban 2012). More recently, attention has turned to the combination of casting and BoNTA as a management option.

Serial casting or the use of a single cast has been used as an adjunct to BoNTA in order to gain additional ROM ( Desloovere 2001a, Booth 2003, Glanzman 2004, Park 2010, Lee 2011). The use of casting after BoNTA injections has been shown to result in improved ankle kinetics (Molenaers 1999) and walking performance (Desloovere 2001a). Booth and colleagues (Booth 2003) reported that the time to achieve the desired ROM was reduced using BoNTA followed by serial casting than by serial casting alone. In a trial comparing BoNTA followed by physiotherapy and orthotics with and without the addition of casting immediately following BoNTA, greater improvements in time‐distance gait parameters and spasticity were found with the addition of casting (Corry 1998). Two recent studies comparing BoNTA alone, casting alone and the combination of both, reported no significant changes in passive ROM, ankle kinematics, spasticity and dorsiflexor strength in the BoNTA only group but significant changes in the other two groups (Glanzman 2004, Ackman 2005). However; not all of these studies are controlled studies. A systematic review of randomised controlled studies can determine if the combination of BoNTA and casting results in better outcomes than if casting alone or BoNTA alone is carried out for children with CP presenting with equinus gait.

Objectives

The primary purpose of this review is to determine if the combination of BoNTA and casting results in better outcomes than if casting alone or BoNTA alone or no intervention is carried out for children with CP presenting with equinus gait. The review considered outcomes from domains of body structure and function , activity limitation, as well as participation/contextual factors.

Methods

Criteria for considering studies for this review

Types of studies

Parallel group, cross‐over and cluster randomised controlled trials (RCTs) were sought for this review.

Types of participants

This review concerns children with cerebral palsy, under eighteen years of age, who present with equinus gait.

Types of interventions

Trials that included:

• administration of BoNTA, followed by single or serial casting, or;

• administration of BoNTA alone, or;

• administration of single or serial casting alone.

Comparators to these interventions were:

• BoNTA and single or serial casting to BoNTA alone,

• BoNTA and single or serial casting to single or serial casting alone,

• BoNTA alone to single or serial casting alone,

• BoNTA to placebo/control,

• Single or serial casting to no casting.

Types of outcome measures

To structure the outcome measures that were considered we used the International Classification of Functioning, Disability and Health (ICF), a member of a family of classifications developed by the World Health Organization (Ustun 2003). Within the ICF framework, our ability to function is described in terms of the interplay between body functions and structures (i.e., physiological functions of body systems (including psychological) and anatomical parts such as organs, limbs and their components), activity (i.e., execution of a task or action by an individual) and participation (i.e., involvement of an individual in life situations) and contextual factors (i.e. environmental and personal factors). We included the following outcome measures:

Body Structures/Body Functions Spasticity Measures

• Ashworth Scale (AS)

• Modified Ashworth Scale (MAS)

Range of Motion Measures

• Passive Range of Motion (PROM)

• Active Range of Motion (AROM)

• Modified Tardieu Scale (MTS)

Gait Measures

• Temporal‐spatial parameters (walking speed, cadence, stride length, step length); Physician rating scale (PRS)

Activity

• Gross Motor Function Measure (GMFM)

Participation/Contextual Factors

• Satisfaction portion of COPM

• Parent satisfaction questionnaires.

Assessments carried out in the short‐term (1‐6 months) and the longer term (12‐24 months) were of interest for the\is review,

Search methods for identification of studies

We searched the following bibliographic databases using the search strategies outlined below: PUBMED (1950's‐January 2013), Cochrane Library (2013), EMBASE (1974‐January 2013), CINAHL (1982‐January 2013), WEB OF SCIENCE (1980‐January 2013), PEDRO (1929‐January 2013), NLM GATEWAY, LILACS (1982‐January 2013), Dissertations and Theses ‐ Proquest (1980‐January 2013), Index to Theses with abstracts accepted for higher degrees by the universities of Great Britain and Ireland and the Council (1716‐January 2013), Theses Canada Portal (1965‐January 2013) Appendix 1; Appendix 2; Appendix 3; Appendix 4; Appendix 5; Appendix 6; Appendix 7; Appendix 8; Appendix 9.

General notes: Where applicable all thesaural subject terms in databases were exploded. All languages were included and no methodological design filter was employed.

In addition to the database searches, we contacted key investigators in the field and pharmaceutical companies involved in research in this area for additional trials. We had responses from Richard Baker (Australia), HK Graham (Australia), LA Koman (USA) indicating that they were not currently involved in research in this area. We did not have a response from K. Desloovere or G. Molenears (Belgium).On review of the Allergan web site the only ongoing trial is related to stroke. The senior medical information officer at Ipsen offices (Dysport) has forwarded the inquiry to her colleagues and we did not receive a response.The reference lists of relevant publications were reviewed for additional publications. No unpublished studies were found.

Data collection and analysis

Identification and selection of relevant studies

Two review authors (BK and JW) independently screened the titles and abstracts of the citations produced by the literature search to determine if the inclusion criteria were met. A total of seven hundred and seventy one titles and abstracts were reviewed. When titles and abstracts suggested a study was potentially eligible for inclusion, a full paper copy of the report was obtained. Thirty six papers were retrieved. Twenty were excluded for reasons outlined in Characteristics of excluded studies. Disagreements between the two authors regarding the study's eligibility were resolved through negotiation with a third review author (MML).

Data extraction

Two review authors (BK and JW) extracted the data using a standard extraction form. Data were extracted at baseline and at follow‐up assessment periods post intervention as well as at the end of the intervention period. Any discrepancy between the data extracted by the two review authors was resolved by negotiation arbitrated by a third author (MML).

Assessment of methodological quality of included studies We assessed the methodological rigor of each included trial in order to prevent systematic errors (bias) in the review. We used the Cochrane Risk of Bias Tool for this. Sources of bias looked for included selection bias (blinding of randomisation), performance bias (blinding of intervention), attrition bias (completeness of follow‐up), detection bias (blinding of outcome measures) and selective reporting of results. Allocation concealment was classified as adequate (A), unclear (B), inadequate (C), or not used (D). Any differences between the review authors were resolved through discussion with a third author (MML) until consensus was reached. The results of quality ratings are presented with the table of included studies.

Measures of Treatment effect All trials included in the systematic review were entered into Review Manager 5. One randomised cross‐over trial was included in this review. Cross‐over trials are considered suitable for evaluating interventions in the treatment of chronic, stable conditions Deeks 2006. In this study McNee 2007 a washout period of 12 weeks was used based on a study by Corry 1998.

In order to perform a meta‐analysis of continuous data we needed to extract the mean values of the outcomes, the standard deviations (SDs) and the number of participants in each of the groups.When means and confidence intervals were provided, SDs were calculated by first calculating the standard error (SE = (upper limit‐lower limit)/3.92 for 95% CI) and then SD was calculated from the SE (SD = SE/√1/Ne +1/Nc where Ne =sample size in experimental group and Nc = sample size in the control group).

For continuous outcomes a weighted mean difference (WMD) or standardized mean difference (SMD) between treatment and control groups was calculated, when appropriate. A risk ratio method was used to combine dichotomous outcome data. The results from the various studies were tested for heterogeneity using the chi‐square statistic, with a significance level of P= 0.05 and the I² statistic where 50% indicates substantial heterogeneity Deeks 2006. Where heterogeneity was significant, meta‐analysis was performed using the random‐effects model. Overall effects were only estimated among groups of trials that used the same intervention. A sensitivity analysis was conducted to evaluate the robustness of the meta‐analysis examining range of motion. When meta‐analysis was not possible a descriptive summary of the results of individual studies was compiled.

Feedback

Questions from the Authors to Reviewers

Summary

1. We did receive useful feedback from two reviewers and I will attach our responses here:

 

Internal Reviewer	Change
In defining equinus in the first sentence of ‘Plain language summary: Background, para 2’ the authors appear to assume that both forms of contracture are always present: ‘Equinus gait is due to spasticity in the calf muscles and decreased ROM at the ankle joints’. Since this is not always true, particularly in younger subjects, this should read:  ‘ spasticity in the calf muscles and/or decreased ROM at the ankle joints’.  Although the most common cause of fixed contracture in cerebral palsy is unequal muscle/bone growth, there are other possible causes of fixed contracture.	‘... spasticity in the calf muscles and/or decreased ROM at the ankle joints’.    most commonly caused by....
Equinus gait may be the result of a fixed contracture and/or a dynamic contracture, failure to clearly differentiate between these two causes (which may co‐exist in one individual) may be the reason that the authors fail to reach generalizable conclusions, since the ‘best’ intervention is likely to vary with type of contracture.   Botulinum toxin A (BoNTA) can only be useful in dynamic contracture, whereas casting addresses primarily, though not exclusively, fixed contracture.	Add to background. We have, I think, made statements explaining the difference between the two..is this enough? “Equinus deformity is defined as a limitation of passive dorsiflexion  leading to increased plantarflexion during stance and walking. The deformity can be dynamic or fixed. Dynamic equinus is attributed to spasticity in the gastrocnemius‐soleus and an imbalance between the action of this muscle and the tibialis anterior (Blackmore 2007). It is seen more commonly in younger children (Boyd 2000). Fixed equinus is caused by muscle contracture, attributed to faster growth in the tibia than the spastic calf muscle‐tendon (Blackmore 2007).” All the studies included specifically state that the participants have a dynamic contracture so I have edited the”types of participants”
Under Types of Subjects, subjects should therefore be characterized by the origins of their equinus gait:  fixed, dynamic or both forms of contracture, and the effects of the interventions (on both range of motion and spasticity as well as higher level outcomes) examined for each group of subjects.	Added to Type of Participant “dynamic” but need to describe for each study????
Concerning the interventions.  The authors mention in their conclusion, that the treatments are very heterogeneous and discuss the heterogeneity of BoNTA administration.  However casting may be applied in many ways which may broadly be categorized by the frequency and duration of applying the cast.  The majority of studies addressed in this review consider serial casting, and it may be advisable to limit the review in this respect:  if the effects of a single cast were to differ from those of serial casting (as might be anticipated for an intervention that aims to apply progressive stretch) it is not useful to combine the two as suggested in the last para of Plain Language Summary, Background.	Add to background and or discussion.   I have mentioned it broadly in the Background and feel that a more detail could be addressed in the discussion. 4/7 used serial casts
There are two commercially available forms of BoNTA, one of which has the trade name Botox which is often abbreviated BTX‐A.  When referring to the generic toxin it is therefore preferable to use the acronym BoNTA.	Changed to BoNTA throughout
P3/39, 1st para, line 8:  There are 2 further studies that might be added to those of Corry, Love and of Reddihough:  Ubhi 2000 and Bjornson 2007.	Added Bjornson 2007, however Ubhi did not examine tone
p.3/39 Types of interventions:  3 criteria are listed:  are all 2 necessary, any one or any two? (i.e. should the last criterion be prefaced by an ‘and’ or an ‘or’ or something more complicated?  From the studies selected the answers appears to be any one criterion (provided there was at least one comparison group), this needs to be specified.	Changes made.
P3/39 Under search methods:  the authors have contacted a number of researchers to ascertain whether any relevant trials are in progress.  With the increasing pressure to register trials before their initiation by many leading journals making this a pre‐requisite to publication, it should now be more thorough (as well as easier) and to examine clinical trial registers.	LILACS search strategy included clinical trials registry
Pp5‐10/39 In reporting the 20 studies, it would be helpful to indicate the GMFCS range of each of the study samples, particularly since Reddihough et al suggest that this may be a determinant of the effect of the therapies being studied.	Added GMFCS to descriptions and tables of included studies
p.10/39, ‘Risk of bias in included studies para 3, 2nd and 3rd sentences:  would read more clearly and accurately with: ‘Eight of the 20 trial gave insufficient information to judge the adequacy of allocation concealment (refs.).  The remaining trials did not describe any concealment techniques (refs.).’	Wording changed
p.13/39 Discussion, 2nd  sentence:  I do not understand the phrase ‘post‐management care’;  why not simply, ‘care’ or ‘management’?	Reworded.
There are a great number of minor typographical errors, below I indicate only a few.  It will need careful proof reading	Corrected
p.2 end of para 2 what does the ‘A’mean?	Removed “A”
Acronyms should be spelt out at first use:  e.g.. PT, ROM, PROM, GMFM (defined on p. 4 1st use on p.2)	Acronyms written out
Bottom p. 2.  Ackman citation needs attention.	Second reference removed.
p.3 para 2; 7th line, ‘the’ is repeated	Corrected
p. 8 para 1: line 8, the differences .. were.    penultimate line:  effective is missing its first ‘e’.	Corrected or section removed?
p.8.   In reporting Flett 1999, the first sentence is difficult to read on account of the amount of  material in parentheses.  I suggest that the 1st sentence is completed and the material in parentheses presented in a 2nd sentence.	Changed to sex ratio and mean age and range of total sample.
External Reviewer (Smeulders)
These studies should fall outside the scope of this review. PT is not a research question. Results of these studies should only include the Botox alone group or casting.	Deleted 8 studies comparing BTX‐A and PT
Secondly, I miss the critical appraisal of the included studies, Many of them contradict each other, so which one should we believe? What do the authors think of the results and conclusions of the separate studies?	Add critical appraisal in discussion.
Used outcome measures. Which outcome measures would be best to test an improved equinus gait? Where do these outcome measures come from? Did the authors beforehand select these outcome measures?why these measures? It would be interesting to include a quality review of the measures used. Many of them are not standard.	Having read the papers we then selected the most commonly reported outcome measures in an effort to combine the results. Reliability and validity of the outcomes added to discussion
Results: description of Studies I would suggest to restrict to a description of the study only aqt this part, and save the conclusions of the study for later. Furthermore, the conclusions of the separate studies that do not concern casting and Botox should not be mentioned, as this is confusing.	Conclusions of the separate studies that do not concern casting and Botox have been removed
Results: Botox vs casting: In this subheading only casting vs Botox is discussed. I would suggest to change the subheading and discuss the three conditions in one paragraph.
BTX‐A only versus Casting only Corry These results show that the number of statistical tests were far too great for the small sample sizes. That makes the significance level unstable (as is shown by significant improvement in one group, but no differences between groups. I would suggest to select the single most important outcome measure (based on theoretical considerations) and leave the rest out.	Critique results.
Results: BTX‐A only versus Casting Corry I do not agree with this conclusion, do the others agree? That should be in the discussion	Discuss shortcoming of Corry in discussion section
Results: BTX‐A and casting versus BTX‐A only Summary: Why is pooling not an option. Should for instance the follow‐up time be of such importance? Does it really matter that some studies had follow up time of 6 weeks and some of 8 weeks and others 12 months? Would ‘’short term’’ and ‘’long term’’ follow‐up be possible? Furthermore, If the outcome measures are reviewed critically, would it be possible to pool some of them?	Consider pooling
Results: Casting versus no Cast Summary: All studies had 12 months follow up time
Results of Desloovere 2001 Again, this is not a research question. Data from this study may be used to pool with the ‘Botox and casting group’.	Use as exp group the group that had Botox first, then casting
In the discussion, it may be questioned whether differences between administration order of Botox/casting would be of importance.	Added to discussion
Discussion ‐ One trial is counted three times[MS1] .	Reworded
Discussion ‐ In general these are outcomes at the function level and the activity level, after critical appraisal of the outcome measures used, it should be possible to pool at least some of them. And when outcome measures are considered not valid, the study should be excluded.	Unclear what review is suggesting difficult to pool data generated by different outcome measures
Discussion: But does the treatment protocol matter? I would suggest to try to pool the data of all patients that underwent: 1 casting, 2 casting and Botox, or 3 Botox only. The rest of the data should not be used in this review.	Unclear what reviewer as suggesting pooling all of these data would preclude drawing conclusions about the most effective intervention.

 

Appendices

Appendix 1. Medline search strategy

1.cerebral palsy [mesh] 2.cerebral pals* 3.dipleg* or hemipleg* 4.equinus deformity [mesh] 5.equinus 6.muscle spasticity [mesh] 7.spastic* 8.gastroc* 9.soleus 10.#1 or #2 or #3 or #4 or #5 or #6 or #7 or #8 or #9 11.botulinum toxins [mesh] (includes type A) 12.botulinum or botox or btx‐a or dysport 13.casts or casting 14.#11 or #12 or #13 15.#10 and #14 16.Limit to all child (includes adolescents up to aged 18)

Appendix 2. Cochrane Library search strategy

1.cerebral palsy [mesh] 2.cerebral pals* 3.dipleg* or hemipleg* 4.equinus deformity [mesh] 5.equinus 6.muscle spasticity [mesh] 7.spastic* 8.gastroc* 9.soleus 10.#1 or #2 or #3 or #4 or #5 or #6 or #7 or #8 or #9 11.botulinum toxins [mesh] (includes type A) 12.botulinum or botox or btx‐a or dysport 13.casts or casting 14.#11 or #12 or #13 15.#10 and #14

Appendix 3. EMBASE search strategy

1.cerebral palsy [Emtree] 2.cerebral pals* 3.dipleg* or hemipleg* 4.pes equinus [Emtree] 5.equinus 6.spasticity [Emtree] 7.spastic* 8.gastroc* 9.soleus 10.#1 or #2 or #3 or #4 or #5 or #6 or #7 or #8 or #9 11.botulinum toxin [Emtree] 12.botulinum or botox or btx‐a or dysport 13.casts or casting 14.#11 or #12 or #13 15.#10 and #14 16.Limit to all child and adolescents

Appendix 4. CINAHL search strategy

1.cerebral palsy [CINAHL heading] 2.cerebral pals* 3.dipleg* or hemipleg* 4.equinus deformity [CINAHL heading] 5.equinus 6.muscle spasticity [CINAHL heading] 7.spastic* 8.gastroc* 9.soleus 10.#1 or #2 or #3 or #4 or #5 or #6 or #7 or #8 or #9 11.botulinum toxins [CINAHL heading] 12.botulinum or botox or btx‐a or dysport 13.casts or casting 14.#11 or #12 or #13 15.#10 and #14 16.Limit to all child and adolescence

Appendix 5. WEB OF SCIENCE search strategy

1.cerebral pals* 2.dipleg* or hemipleg* 3.equinus 4.spastic* 5.gastroc* 6.soleus 7.#1 or #2 or #3 or #4 or #5 or #6 8.botulinum or botox or btx‐a or dysport 9.casts or casting 10.#7 or #8 or #9 11.#7 and #10 12.#11 and child* or adolesc*

Appendix 6. PEDRO search strategy

First search strategy: "cerebral pals" and "botulinum toxin" Second strategy "cerebral pals" and cast Third search strategy: Spastic and botulinum

Appendix 7. LILACS search strategy

cerebral palsy OR diplegia OR hemiplegia OR equinus OR spasticity OR gastrocnemus OR soleus AND botulinum OR botox or btx‐a OR dysport OR casts OR casting Found zero results; however, the alternate search strategy below found 14 references: cerebral palsy and botulinum [Palavras] Various combinations of terms were used to ensure nothing relevant was missed.

Appendix 8. Other search strategies

Digital Dissertations

cerebral palsy OR diplegia OR hemiplegia OR equinus OR spasticity OR gastrocnemus OR soleus AND botulinum OR botox or btx‐a OR dysport OR casts OR casting

Index to Theses with abstracts accepted for higher degrees by the universities of Great Britain and Ireland and the Council

(cerebral pals* OR dipleg* OR hemipleg* OR equinus OR spastic* OR gastroc* OR soleus) and (botulinum OR botox or btx‐a OR dysport OR casts OR casting)

Theses Canada Portal

Cerebral palsy and botulinum

Appendix 9. NLM Gateway search strategy (includes the ClinicalTrials.gov registry)

(cerebral pals* OR dipleg* OR hemipleg* OR equinus OR spastic* OR gastroc* OR soleus) and (botulinum OR botox or btx‐a OR dysport OR casts OR casting)

What's new

Last assessed as up‐to‐date: 13 March 2008.

Date	Event	Description
30 March 2014	Amended	Incorporated Feb 2014 Reviewers suggestions
28 July 2013	Amended	Peer Referee's recommendations (received May 2013) added. Authors' comments will be sent separately to Movement Disorders Group.
22 March 2013	Amended	amended and reviewer's recommendations added.
23 May 2011	Amended	Converted to new review format.

Contributions of authors

Tim Ruggles carried out the literature search.

Barbara Kelly and Judie Woodward screened the titles and abstracts of the citations. They also extracted the data.

Barbara Kelly wrote the review.

Marilyn MacKay Lyons assisted in the writing and editing of the review. She also assisted with responding the reviewers comments.

Declarations of interest

Two of the authors of this review are currently carrying out a randomised controlled trial to determine an effective casting protocol post‐BoNTA injections to treat equinus gait in children with CP.

Notes

The protocol is out of date and does not meet the current methodological standards of The Cochrane Collaboration.

Withdrawn from publication for reasons stated in the review