A systematic review of reported outcomes following Ponseti correction of idiopathic club foot.

AIMS: To analyze outcomes reported in studies of Ponseti correction of idiopathic clubfoot.
METHODS: A systematic review of the literature was performed to identify a list of outcomes and outcome tools reported in the literature. A total of 865 studies were screened following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, and 124 trials were included in the analysis. Data extraction was completed by two researchers for each trial. Each outcome tool was assigned to one of the five core areas defined by the Outcome Measures Recommended for use in Randomized Clinical Trials (OMERACT). Bias assessment was not deemed necessary for the purpose of this paper.
RESULTS: In total, 20 isolated outcomes and 16 outcome tools were identified representing five OMERACT domains. Most outcome tools were appropriately designed for children of walking age but have not been embraced in the literature. The most commonly reported isolated outcomes are subjective and qualitative. The quantitative outcomes most commonly used are ankle range of motion (ROM), foot position in standing, and muscle function.
CONCLUSIONS: There is a diverse range of outcomes reported in studies of Ponseti correction of clubfoot. Until outcomes can be reported unequivocally and consistently, research in this area will be limited. Completing the process of establishing and validating COS is the much-needed next step.Cite this article: Bone Joint Open 2020;1-8:457-464.

Introduction

The Ponseti method has revolutionized the care of children with clubfoot. Primary correction rate is generally high and fewer surgical procedures are required to correct the deformity.[1-3] The Ponseti method is low cost, can be delivered successfully by various members of the healthcare team,[4-10] and has been adopted worldwide with good results. Although primary correction can be achieved, there is less certainty around the long-term outcomes of treatment. Data on relapse rates, functional outcome, and quality of life is inconsistently defined[11,12] The two most commonly used assessment methods of club foot are the Pirani[13] and Diméglio[14] scores. These are validated and repeatable assessments of the foot at initial presentation with good interobserver repeatability between orthopaedic surgeons, physiotherapists, and trained Ponseti practitioners.[5-9] Other early classification systems[15,16] aimed to provide an assessment of the foot during treatment or after initial correction. These classification systems make a descriptive, static assessment of the foot. They do not consider function, pain, patient or parent satisfaction, or psychological wellbeing of the child or caregiver. They are also not validated for use in older children but can possibly still be helpful in static assessment of the primary correction. There has been a noticeable effort in the literature to develop tools to assess outcomes of treatment of club foot across a broader range of domains.[17-21] These systems combine aspects of physical examination, gait assessment and parental and child reported outcomes (PROMS).

Most recently, assessment tools have been developed and specifically designed for low-income healthcare settings[22,23] showing good interobserver repeatability between physiotherapists and Ponseti practitioners.[24] Other classification systems which integrate gait analysis and paedobarography, [25-29] or isokinetic muscle strength assessment,[26,30] provide additional information but may not be feasible in most settings. The use of plain radiographs to assess bone/joint alignment, as an adjunct in the assessment of clubfoot correction,[31,32] has shown poor correlation with function and relapse.[33] There is a risk of creating a ‘flooded marketplace’ of clubfoot outcome measures, with no clear guidance or evidence on which tool is best utilized in each clinical environment at each stage of treatment.

Measuring and comparing clubfoot treatment success as well as indications for further intervention is currently hindered by the lack of agreed outcomes by the treating community. Core Outcome Sets (COS), defined as the minimum standardized collection of outcomes to be measured and reported in all research for a specific clinical area,[34] have been developed for several musculoskeletal conditions[35,36] but not for clubfoot management. In the absence of a COS we have used the COMET (Core Outcome Measures in Effectiveness Trials) initiative principles[37] to identify the outcome tools reported following Ponseti correction of idiopathic clubfoot and analyze them according to domains and acceptance.

Methods

A systematic review was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.[38] It was prospectively registered on the PROSPERO international prospective register of systematic reviews (CRD42018089517) and on the Core Outcome Measures in Effectiveness Trials (COMET) database.

The National Institute for Health and Care Excellence Databases Advanced Search interface was used to search EMBASE, Medline and CINAHL databases between 01 January 2012 and 01 May 2019. The search strategy terms are listed in the supplementary material. The original registration related to a project developing a core outcome set for the assessment of congenital talipes equionvarus (CTEV). An initial study was limited to assessing relapse as an outcome measure[11] and the search dates were later extended for the purpose of this study which looks at other methods of assessment. Included were studies published in English, which addressed primary correction of idiopathic clubfoot using the Ponseti method. All study designs were considered.

Click here for additional data file.

Study titles, abstracts and full texts were screened independently by two authors (KH, AF). A senior author (YG) was consulted in the case of disagreement over the suitability of a text for inclusion. Disagreements were resolved by consensus. Data were extracted with use of the Covidence systematic review software (Veritas Health Innovation, Melbourne, Australia) and a data collection table on Microsoft Excel, (Microsoft, Redmond, Washington, USA).

All clubfoot specific tools and classification systems were included, as well as isolated outcomes when no specific tool was used. For each clubfoot-specific assessment instrument, a separate literature search was performed to identify papers reporting the instrument’s interobserver repeatability. Repeatability was presented as reported by the relevant paper. Categorization of each verbatim outcome definition to an outcome name, and each outcome name to an outcome domain, has been performed independently by two researchers from different backgrounds (KH, AF). These two researchers worked on this process and two senior authors (YG, DME) were available to resolve differences and make final decisions.

Each outcome term was assigned to one of the five core areas defined by the Outcome Measures Recommended for use in Randomized Clinical Trials (OMERACT) Filter 2.0.[39] The OMERACT framework describes five key domains; adverse events, life impact, resource use, pathophysiological manifestations, and death. The sixth domain of ‘technical considerations and feasibility of use in clinical practice’, suggested by Dorman et al[36] for technical or surgical outcomes relevant to surgeons, was also included. Death was deemed nonapplicable for this analysis. As the purpose of this study was to collect all outcomes of treatment regardless of data quality, a bias assessment or appraisal of methodological quality of data has not been included.

Results

The initial literature search resulted in a total of 865 articles. The PRISMA flowchart for study selection is presented in Figure 1. A total of 124 studies fulfilled the inclusion criteria and were analyzed (Supplementary Material). The 124 included studies contributed 12,376 patients with 16,935 clubfeet. The demographic data is presented in Table I.

Fig. 1

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flowchart for study selection.

Table I.

Demographic data for included studies and participants.

Characteristic	Total
Studies, n	124 (100)
Evidence level n (%)
1b	7 (6)
2b	92 (75)
3b	3 (2)
4	21 (17)
Continent of origin of publication n (%)
Asia	56 (46)
Europe	31 (25)
North America	23 (19)
Africa	8 (7)
Australasia	3 (2)
South America	3 (2)
Patient demographics
Female, n %	32 (26)
Mean age at start of casting, wks (range)	32.3 (0.3 to 384)
Ponseti treatment
Mean Achilles tenotomies performed, % (range)	75.8 (22.4 to 100)
Mean follow-up period, mths (range)	47.4 (2.5 to 360)
Ponseti practitioner, n (%)	n = 61
Orthopaedic surgeon	50 (82)
Physiotherapist	9 (15)
Plaster technician	2 (3)

These studies presented a wide variety of isolated outcomes and outcome scores. The list of the 20 verbatim outcomes generated along with the number of times they have been used is available in Table II.

Table II.

Clubfoot verbatim outcomes.

Outcome measure, n	Studies reporting outcome (n = 124)
Heel (hindfoot) position	23
Forefoot position	16
Footwear	6
Foot size	1
Calf size	1
Foot position in standing	5
Foot position in walking	4
Ankle ROM, passive and active	34
Subtalar ROM, passive and active	12
Pre-chosen gait parameters (squatting, stair climbing, walking)	10
Qualitative muscle activity and function	10
Quantitative muscle strength	3
Pain	7
‘Need’ for surgical intervention (surgeon’s perspective)	44
Recurrence (qualitative)	67
Recurrence (quantitative)	3
Angles from plain radiographs	18
Paedobarograph	3
Gait analysis	8
PROMs (parents)	28
PROMs (patients)	11
PROMs (both patients and parents)	4

PROMs, parental and child reported outcomes; ROM, range of motion.

The outcome scores identified, of which there were 16, are presented in Table III. Each score was labelled as containing descriptive, functional, and/or prognostic elements. The stage at data collection and the repeatability of the score, if reported, were also documented. All scores presented a descriptive component, and the majority (13/16; 81%) offered an additional functional component with variable detail, but only three systems incorporated a prognostic component.

Table III.

Clubfoot-specific outcome tools presented in chronological order, with most recent first.

Outcome instrument (year)	Parameters assessed	Descriptive/ Functional/ Prognostic	Stage at treatment to be used	Repeatability/predictive value	Studies, n*
PBS tool[18] (2019)	Foot position in standingPrechosen gait parametersAnkle ROMSubtalar ROM	D/F	Walking age	Interobserver agreement: 0.93[19]	1
Assessing Clubfoot Treatment tool[40] (2017)	Ankle ROMPainFootwearParental PROM	D/F	Walking age	Detecting need for further intervention:Sensitivity: 79.2% Specificity: 100%[23,25]	2
Evertor score[41] (2014)	Ankle ROM and subjective muscle strength	D/F/P	After initial treatment	Predicting recurrence and need for further interventionPPV and NPV	4
Bangla clubfoot tool assessment[23,42] (2014)	Parental PROM of foot appearance, footwear, pain, functionPrechosen gait parametersFoot position in standingAnkle ROM	D/F/P	Walking age	Predicting need for referral for further treatment:Interobserver agreement: 0.92Sensitivity: 79.2% Specificity: 79.5% [24,25]	4
Bhaskar relapse assessment tool[43] (2013)	Ankle and foot ROMPrechosen gait parametersFoot position	D/F	Walking age		3
IMAR Clubfoot Scale[25] (2009)	Ankle ROMFoot position in standingFoot and calf appearanceGait (GAITRite system)Paedobarograph (dynamic and static)Parental PROM	D/F	Walking age	Interobserver agreement: 0.796[35]	2
Richards Classification[44] (2008)	Ankle ROMNeed for surgical correction	D/F	After initial treatment		3
Clubfoot Assessment Protocol[21] (2005)	Ankle/foot ROMMuscle strengthFoot positionPrechosen gait parameters	D/F	After initial treatment	Interobserver agreement range: 0.35 to 0.38Interobserver agreement range: 0.54 to 1.00[22]	1
International Clubfoot Study Group classification system[31] (2003)	Foot positionAnkle/foot ROMMuscle functionPrechosen gait parametersPainAngles from plain radiographs	D/F	Walking age	Interobserver agreement: 0.73[45]	3
Roye’s disease-specific instrument[19] (2001)	Parental PROM of foot appearance, footwear, pain, function, gait	D/F/P	Walking age	Internal consistency reliability of 0.74 to 0.85 (Cronbach’s α)[21] Predicting need for referral for further treatment:Sensitivity: 31.8% Specificity: 100%[25]	5
Ezra clubfoot score[46] (2000)	Ankle/subtalar ROMHind/forefoot position in standingTibialis Anterior functionPrechosen gait parametersFootwearFunctional limitationsPainParental PROM	D/F	Walking age		3
Pirani Score[13] (1999)	Hind/mid/forefoot positionAnkle dorsiflexion	D	Before, during and after initial correction	Interobserver agreement: 0.90[9] Pearson r : 0.89[47]	82
Diméglio Score[14] (1995)	Hind/forefoot positionAnkle/subtalar ROM	D/F	Before, during and after initial correction	Interobserver agreement: 0.83[9] Pearson r : 0.85[48]	36
Catterall Classification[16] (1991)	Hind/forefoot position	D	During treatment	Interobserver agreement range: 0.15 to 0.4[48]	4
Harrold and Walker Classification[15] (1983)	Ankle ROM Hindfoot position	D	Pretreatment	Interobserver agreement range: 0.4 to 0.7[48]	2
Laaveg and Ponseti functional rating system[17] (1980)	Parental PROMPainFoot position in standingAnkle/foot ROMPrechosen gait parameters	D/F	Walking age		5

Absolute number out of 124 papers

PBS, Pirani/Böhm/Sinclair; PROM, parental and child reported outcomes; ROM, range of motion

Seven assessment tools were relevant to the peritreatment period and nine were designed for walking age children, and therefore required an appropriate follow-up time period. The individual outcomes distributed as per OMERACT Framework are presented in Table IV. The identified outcome instruments are mapped in Table V according to the OMERACT domains, defining which outcome score falls under each domain.

Table IV.

Outcomes as per Outcome Measures Recommended for use in Randomized Clinical Trials framework.

Core area	Core domains	Outcomes
Adverse event	Adverse events	Relapse, residual deformity
Life impact	Quality of life	Quality of life, squatting, stair climbing, pain, participation in sports activities, functional score, footwear, foot appearance
Resource use	Economic/hospital/ need for intervention	Further surgery, further casting, appropriate for low income setting
Pathophysiological manifestations	Musculoskeletal	Muscle strength, ankle and subtalar range of movement, gait analysis, paedobarograph
Death	N/A	N/A
Technical considerations	Technical considerations	Radiological measurements, Feasibility of use in clinical setting

N/A, not applicable.

Table V.

Mapping of outcome instruments into Outcome Measures Recommended for use in Randomized Clinical Trials domains

Outcome instrument (year)	Domains
	Adverse event	Life impact	Resource use	Pathophysiological manifestations	Technical considerations
PBS tool[18] (2019)				X	X
Assessing Clubfoot Treatment tool[40] (2017)		X	X	X	X
Evertor score[41] (2014)			X	X	X
Bangla Clubfoot Tool[23,42] (2014)		X	X	X	X
Bhaskar relapse assessment tool[43] (2013)	X	X		X
IMAR-Clubfoot scale[25] (2009)		X		X
Richards classification[44] (2008)			X	X
Clubfoot Assessment Protocol[21] (2005)		X		X
International Clubfoot Study Group[31] (2003)		X		X	X
Roye’s disease-specific instrument[19] (2001)		X
Ezra clubfoot score[46] (2000)		X		X	X
Pirani score[13] (1999)			X	X	X
Diméglio score[14] (1995)			X	X	X
Catterall classification[16] (1991)			X	X
Harrold and Walker classification[15] (1983)			X	X
Laaveg and Ponseti functional rating system[17] (1980)		X		X	X

PBS, Pirani/Böhm/Sinclair.

The core areas reported were principally distributed between three domains (pathophysiological manifestations, life impact, and resource use) with the most common quantitative outcome being ankle ROM (27%; 34/124). None of the studies reported death. The two outcome measures most commonly used are recurrence rates in 54% (67/124) (with or without a qualitative definition) and need for surgical intervention in 35% (43/124), again without any definition of which aspect of the deformity required treatment. The most reported outcome tools were Pirani (66%; 82/124) and Diméglio (29% 36/124). None of the outcome systems designed for after walking age were used frequently (between one and five times each). The outcomes systems designed for walking age included more functional components than the mostly descriptive tools designed for correction age (Table III). PROMs were used in 31% (39/124) of the studies, mostly parent reported (Table II). Outcomes from the life impact core area, such as foot appearance, foot and calf size and function, were not frequently used by the treating clinician, but are emphasized in PROMs.

Discussion

The Ponseti method has become the standard of care for idiopathic clubfoot correction. The number of reported studies has increased steeply in the last two decades and all show consistent success in initial correction. However the long-term outcomes are not clearly defined, therefore limiting high-quality systematic reviews and treatment decisions. This systematic review reports all individual outcomes and assessment tools used in patients who have undergone the Ponseti method and allocates outcomes to the six core areas as per the OMERACT framework.

The majority of assessment tools were appropriately designed and tested in ambulatory children. The assessment tools require a broad range of clinical expertise, time, parental input, and equipment. Some tools use PROMs alone, such as Roye’s disease-specific instrument.[19] PROMs combine clubfoot-related symptoms such as pain and shoe wear and other more holistic assessments such as involvement with play, missing time at school and emotional impact. Other outcome tools use clinical examination findings alone.[15,18,21] Different tools dictate different degrees of detail for clinical examination – ranging from binary ‘yes’ or ‘no’ outcomes (for example, for a plantigrade foot) to quantitative measures of range of movement.

Several tools use a combination of PROMs and examination findings.[17,23,31] Gait was assessed in a variety of ways, from simple parental-reported functional measures (such as ability to squat or run[23]), qualitative gait pattern observed in clinic,[19] or a comprehensive assessment in the gait laboratory.[25] While outcomes at a given end point are important, to the treating clinician it may be important to recognize the ‘failing’ foot and thus scores that contain a prognostic element may have more relevance.[19,23,41]

Most tools were developed in high-income countries with some notable exceptions.[22,23] It should be acknowledged that different parameters will be important to the parents and the child at different ages and different outcomes will have different significance in different parts of the world.

After excluding assessment tools that were not condition specific, we have identified 20 outcomes and 16 assessment tools and defined their acceptability and, where assessed, reported their repeatability (as stated in the initial paper). While the infant applicable classification systems of Pirani[13] and Diméglio[14] are widely accepted, none of the true outcome tools have been similarly accepted by the treating community. There were seven outcome systems found to be applicable to walking age children. They all include ankle ROM and foot position in standing. Three include muscle function assessment and five included various PROMS with different degrees of detail adaptable to the local setting.

The most commonly reported outcome measures are quantitative and surgeon subjective. As previously reported[11] recurrence rate was reported in more than half of the papers with a broad interpretation of the term. The broad term “need for surgical intervention” has been reported as an outcome in 35% of the papers. Both of these outcome measures cannot serve as a part of the COS and their use in publications is not helpful for either clinical decision-making or research.

This study does have limitations. The systematic review is limited to the English language and therefore might be missing valuable information. The focus of this study was to collect as many outcomes measures as possible and therefore no studies were excluded by their level of evidence. There was also no threshold for minimum follow-up and thus a variable follow-up distribution within the included studies.

Clubfoot in the walking age child must be assessed more comprehensively and more appropriately than using the popular infant applicable Pirani[13] and Diméglio[14] scores. These are validated and repeatable assessments of the foot at initial presentation with good interobserver repeatability between orthopaedic surgeons, physiotherapists, and trained Ponseti practitioners.[5-9] There is an obvious trade-off between a comprehensive, time-consuming tool that requires expertise and training and a short, basic tool that can be used by all in every setting. The absence of clear outcomes has resulted in a wide diversity of treatments being recommended at various stages in the management of clubfoot and for various indications.

The ‘ideal’ CTEV assessment tool would be easy to use, comprehensive, adaptable to local resources, and tailored to geographical area. It would provide a multifaceted assessment including a static and dynamic clinical examination and basic PROMs, adaptable to different stages of treatment, and validated and repeatable by all Ponseti practitioners. A COS will provide a basic uniform building block from the overall diverse outcome system.

We recommend that the global clubfoot community embrace one classification system for the assessment of clinical outcome. Currently, this validated tool is not available and collaborative efforts would be recommended in order to establish one. Currently, there are a large variety of outcome measures reported following clubfoot correction with no established COS. Until outcomes can be reported unequivocally and consistently, research in this area will be limited. Completing the process of establishing and validating COS is the much-needed next step.

Supplementary material

Search terms used.