Global prevalence, patterns, and predictors of relapse following Ponseti treatment for idiopathic clubfoot: a systematic review and meta-analysis Authors: Victor Ogutu, Ewochu Emmanuel Jacob, Bienfait Mumbere Vahwere, Augusine Oviosun, Chinyere N Anyanwu, Wusa Makena, Ugwu Okechukwu Paul Chima, Okon Michael Ben, Swase Dominic Terkimbi, Emeka Godson Anyanwu Published: 2026 ABSTRACT AbstractBackgroundThe Ponseti method is the current global standard for the management of idiopathic clubfoot and achieves high initial correction rates. However, relapse during childhood remains a clinically important challenge, with widely varying estimates reported across studies. This systematic review and meta-analysis aimed to estimate the pooled prevalence of relapse following Ponseti treatment and synthesize associated risk factors.MethodsA systematic search of PubMed/MEDLINE, Scopus, and Web of Science was conducted for studies published between 1 January 2015 and 30 January 2026. Observational studies evaluating relapse in children (0–18 years) with idiopathic clubfoot treated using the Ponseti method were included. Random-effects meta-analysis with logit-transformed proportions was used to estimate pooled relapse prevalence. Heterogeneity was assessed using I2 and τ2, and prediction intervals were calculated. Risk of bias was assessed using the Joanna Briggs Institute (JBI) checklist for prevalence studies and the Quality in Prognosis Studies (QUIPS) tool for prognostic factor studies. Subgroup, sensitivity, and meta-regression analyses were performed. Publication bias was assessed using funnel plots and Egger’s test.ResultsTwenty-five studies involving 1,624 children met the inclusion criteria. In the primary analysis restricted to patient-level data, the pooled relapse prevalence was 26.4% (95% CI 20.4–33.5%), while sensitivity analysis including mixed units of analysis yielded a slightly lower estimate of 22.4% (95% CI 17.9–27.7%). Substantial heterogeneity was observed (I2 ≈ 75%), with a wide prediction interval (approximately 10–38%). Brace non-compliance was strongly associated with relapse (pooled OR 22.35, 95% CI 10.09–49.53). Higher baseline severity and delayed treatment initiation were also associated with increased relapse risk. Definitions of relapse varied considerably across studies, contributing to heterogeneity in pooled estimates. Sensitivity analyses supported the stability of findings. There was borderline evidence of small-study effects.ConclusionsRelapse following Ponseti treatment remains clinically significant, affecting approximately one in four to one in five children. Brace non-compliance is the most important modifiable factor associated with relapse. Variability in relapse definitions, follow-up duration, and program implementation likely contribute to heterogeneity in reported outcomes. Strengthening brace adherence, ensuring prolonged follow-up, and standardizing clinically meaningful relapse definitions may improve long-term outcomes.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12891-026-10017-6.Keywords: Idiopathic clubfoot, Congenital talipes equinovarus, Ponseti method, Relapse, Recurrence, Brace compliance, Pirani score, Dimeglio classification, Pediatric orthopaedics, Systematic review Introduction Idiopathic clubfoot, also known as congenital talipes equinovarus (CTEV), is one of the most common congenital musculoskeletal deformities in the world, with an incidence of 1–2 per 1,000 live births and an estimated 150,000—200,000 new cases occurring each year [1–3]. The burden is disproportionately concentrated in the low- and middle-income (LMICs) countries where the majority of affected children (almost 80—90%) are born and where access to early corrective treatment is limited [1, 4]. Despite global scale up efforts to support dissemination of Ponseti services, a significant portion of children in LMICs continue to be denied treatment in a timely manner, and are thus at risk of preventable lifelong disability [1, 5, 6]. Untreated, or inadequately treated, clubfoot has significant functional, psychosocial and socioeconomic consequences [7, 8]. Adults with persistent deformity often suffer from abnormal gait, decreased mobility, chronic pain and decreased quality of life [9–11]. Untreated clubfoot at the population level results in preventable disability, and perpetuates cycles of poverty, and hence is important in global surgery/disability prevention efforts [1, 4]. The Ponseti method has substantially transformed the management of idiopathic clubfoot [12–15]. Historically, treatment often involved extensive posteromedial surgical release that was linked to stiffness, pain, overcorrection and long-term degenerative changes [11, 16, 17]. In contrast, the Ponseti protocol of serial manipulation and casting and percutaneous Achilles tenotomy if indicated, followed by prolonged foot abduction bracing, provides initial correction rates in excess of 90–95% in idiopathic cases in a wide variety of healthcare settings [13, 18, 19]. Long-term follow-up studies have shown better functional results and lower requirement for extensive surgical procedures as compared with historical surgical approaches [10, 11, 18]. These advantages have given rise to widespread use of the Ponseti method across the world and growth of training programs designed to enhance access in resource-constrained settings [1, 4]. Despite high initial correction rates, relapse remains a persistent challenge in the long-term management of clubfoot. Reported relapse rates vary widely across studies, ranging from 3 to 67%, with pooled estimates commonly between 20 and 40% depending on follow-up duration and relapse definition [15, 16, 20, 21]. Relapse is most frequently observed within the first three to five years of life, although recurrence has been documented well into adolescence [15, 16]. Differences in relapse prevalence across studies are influenced by variation in follow-up duration, clinical definitions of recurrence, and methodological heterogeneity among cohorts [3, 11, 22]. Relapse also carries important clinical and health-system implications. Most recurrences require repeat Ponseti casting and occasionally repeat Achilles tenotomy, and 5–15% of patients ultimately undergo additional surgical procedures, including tibialis anterior tendon transfer or corrective osteotomies [14, 17, 18]. These additional interventions create burdens on healthcare utilization and financial and logistical burdens for the family, especially in resource constrained settings. Importantly, the pattern and time of relapse is heterogeneous. Loss of ankle dorsiflexion or recurrent equinus often occurs early in the disease before other components of the deformity [23–25]. Combined deformity patterns such as equinus, adduction and varus may also be observed and dynamic forefoot supination (associated with muscle imbalance) may ensue later in childhood and often requires tibialis anterior tendon transfer [14, 17]. However, the definition of relapse varies considerably between studies, with non-equivalent operational criteria used to define recurrence. This lack of standardization is an important contributor to variability in reported relapse rates. Recent efforts to establish standardized outcome reporting in studies of clubfoot have attempted to address this issue although consensus definitions still continue to be developed. This heterogeneity in operational definitions complicates direct comparison across studies and underpins variability in pooled estimates. A number of factors involved in relapse have been suggested [24, 26, 27]. Brace non-adherence has been consistently shown to be the most relevant modifiable risk factor with non-adherent children showing significantly greater relapse rates compared with adherent children [23, 24, 28]. Higher initial severity of the deformity, delayed or late start of treatment, as well as residual deformity at the time of bracing have been linked to recurrence, although these associations have ranged in strength from one study to another [20, 24]. Sociodemographic factors such as levels of education for both caregivers and children, socioeconomic status and geographic access to care may lead to indirect effects on relapse due to their impact on medication adherence and attendance at follow-up. Despite a rapidly growing body of literature, there are important evidence gaps. First, there is wide variation in the prevalence estimates for relapse as a result of heterogeneous definitions and follow-up durations. Second, a large number of prognostic factor studies are single-center cohorts and have limited adjustment of confounding variables. Third, the variation in healthcare infrastructure and Ponseti program implementation between different regions could be a source of reported variation. Consequently, a comprehensive quantitative synthesis of relapse prevalence, patterns, and associated factors across different settings remains limited. Therefore, the aims of this systematic review and meta-analysis of observational studies of children with idiopathic clubfoot who underwent treatment with the Ponseti method were (1) to estimate the pooled prevalence and describe patterns of relapse and (2) to identify and quantitatively synthesize patient-, treatment-, and system-level factors associated with recurrence. By using the methodological standards of PRISMA 2020 and MOOSE, and limiting the focus to idiopathic clubfoot, the aim of this review is to provide clinically relevant evidence to inform risk stratification and strategies for brace use and long-term follow-up protocols in both high- and low-resource settings. Methods Study design and reporting standards The aim of this systematic review and meta-analysis was to estimate the pooled prevalence of relapse and factors associated with relapse after Ponseti treatment of idiopathic clubfoot. The review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) statement and guided by the Cochrane Handbook for Systematic Reviews of Interventions. A predefined review protocol outlining the study objectives, eligibility criteria, outcome definitions, and analytical approach was developed prior to data extraction. However, this protocol was not prospectively registered in PROSPERO or a comparable registry. In line with PRISMA 2020 recommendations, this absence of registration is acknowledged as a limitation. All primary outcomes, subgroup analyses, and statistical methods were specified a priori to minimize the risk of post hoc decision-making. Data sources and search strategy A comprehensive electronic search was performed in PubMed/MEDLINE, Scopus, and Web of Science Core Collection for studies published between 1 January 2015 and 30 January 2026. The search strategy combined controlled vocabulary (MeSH) and free-text terms related to clubfoot and relapse. Key search terms included: “idiopathic clubfoot” OR “congenital talipes equinovarus” “idiopathic clubfoot” OR “congenital talipes equinovarus” “Ponseti method” “Ponseti method” “relapse” OR “recurrence” “relapse” OR “recurrence” “risk factors” “risk factors” “brace compliance” OR “brace adherence” “brace compliance” OR “brace adherence” Boolean operators (AND/OR), truncation, and database-specific indexing terms were applied to maximize sensitivity. Boolean search strategies for each database are provided in Supplementary File. Reference lists of included studies were also screened to identify additional relevant articles. The search yielded: Scopus: 177 records Scopus: 177 records PubMed: 179 records PubMed: 179 records Web of Science: 85 records Web of Science: 85 records A total of 441 records were identified prior to deduplication. Study selection Duplicate records were identified and removed using the automated duplicate-detection function in Rayyan (Qatar Computing Research Institute), followed by manual verification prior to screening. The remaining studies underwent a two-stage screening process comprising title/abstract screening followed by full-text eligibility assessment. After removal of duplicates (n = 337), 104 records were screened, and 41 studies were selected for full-text review. Of these, 38 full-text articles were retrieved and assessed for eligibility. Study selection was conducted independently by two reviewers, with disagreements resolved by consensus. Eligibility criteria Population Studies were eligible if they included: Children aged 0–18 years Children aged 0–18 years Diagnosed with idiopathic clubfoot Diagnosed with idiopathic clubfoot Treated using the Ponseti method Treated using the Ponseti method Studies involving syndromic, neuromuscular, or teratologic clubfoot (e.g., arthrogryposis, myelomeningocele, cerebral palsy) were excluded unless data on idiopathic cases were reported separately. Intervention Eligible studies included treatment consistent with the Ponseti protocol: Serial manipulation and casting Serial manipulation and casting Percutaneous Achilles tenotomy (where indicated) Percutaneous Achilles tenotomy (where indicated) Foot abduction brace (FAB) maintenance phase Foot abduction brace (FAB) maintenance phase Outcomes Primary outcome Relapse of clubfoot deformity following initial correction with the Ponseti method. Secondary outcomes Factors associated with relapse, including: Brace adherence Brace adherence Initial severity (Pirani or Dimeglio scores) Initial severity (Pirani or Dimeglio scores) Radiographic parameters Radiographic parameters Age at presentation Age at presentation Laterality Laterality Effect estimates (odds ratios or hazard ratios) Effect estimates (odds ratios or hazard ratios) Study design Included: Prospective cohort studies Prospective cohort studies Retrospective cohort studies Retrospective cohort studies Case–control studies Case–control studies Excluded: Case reports Case reports Case series with < 10 patients Case series with < 10 patients Reviews and editorials Reviews and editorials Animal studies Animal studies Studies reporting relapse exclusively after surgical correction Studies reporting relapse exclusively after surgical correction Data extraction Data were extracted independently by two investigators using a standardized extraction form. Extracted variables included: Study characteristics (country, design, follow-up duration) Study characteristics (country, design, follow-up duration) Sample size (patients and/or treated feet) Sample size (patients and/or treated feet) Number of relapse events Number of relapse events Severity grading systems (Pirani or Dimeglio) Severity grading systems (Pirani or Dimeglio) Tenotomy rates Tenotomy rates Brace adherence data Brace adherence data Reported effect estimates (odds ratios or hazard ratios) Reported effect estimates (odds ratios or hazard ratios) Where necessary, 2 × 2 contingency tables were reconstructed from published data to derive effect estimates. The unit of analysis (patient-level vs foot-level) was extracted for each study. Where both were reported, patient-level estimates were preferentially used. Studies reporting only foot-level outcomes were included in sensitivity analyses and interpreted cautiously due to potential within-patient clustering in bilateral clubfoot. Handling of relapse definitions Definitions of relapse were not consistent across studies. Some studies defined relapse as clinical recurrence of deformity components, whereas others defined relapse based on the need for repeat casting, tenotomy, or additional surgical intervention. To minimize bias arising from heterogeneous and non-standardized definitions: Studies were included if relapse was defined as loss of correction requiring clinical intervention or measurable recurrence of deformity. Studies were included if relapse was defined as loss of correction requiring clinical intervention or measurable recurrence of deformity. Definitions used in individual studies were systematically extracted and documented. Definitions used in individual studies were systematically extracted and documented. Sensitivity analyses were performed to assess the impact of varying definitions on pooled estimates. Sensitivity analyses were performed to assess the impact of varying definitions on pooled estimates. Measurement of brace adherence Several approaches to measuring brace adherence were identified, including caregiver self-report, clinician documentation, and reported brace-wear duration during follow-up. Objective monitoring technologies were rarely used. Due to heterogeneity in measurement approaches, brace adherence was analyzed as a binary exposure (adherent vs non-adherent) based on definitions provided in the original studies. Risk of bias assessment Risk of bias was assessed using the Joanna Briggs Institute (JBI) checklist for prevalence studies and the Quality in Prognosis Studies (QUIPS) tool for prognostic factor analyses (Fig. 2A–D). A Risk of bias traffic plot for studies contributing to relapse prevalence (JBI assessment). Traffic plot showing study-level risk of bias across domains using the Joanna Briggs Institute (JBI) Critical Appraisal Checklist for prevalence studies. Each circle represents the judgment for an individual study within a specific domain, classified as low, moderate, high, or unclear risk of bias. B Risk of bias summary for studies contributing to relapse prevalence (JBI assessment). Summary of risk of bias judgments across JBI domains for prevalence studies. Bars represent the percentage of studies classified as low, moderate, high, or unclear risk of bias in each domain. C Risk of bias traffic plot for studies evaluating predictors of relapse (QUIPS assessment). Traffic plot showing study-level risk of bias across domains using the Quality in Prognosis Studies (QUIPS) tool. Each circle represents the judgment for an individual study within a domain, classified as low, moderate, high, or unclear risk of bias. D Risk of bias summary for studies evaluating predictors of relapse (QUIPS assessment). Summary of risk of bias judgments across QUIPS domains. Bars represent the proportion of studies rated as low, moderate, high, or unclear risk of bias in each domain Statistical analysis Prevalence meta-analysis Relapse prevalence was pooled using a random-effects model (DerSimonian–Laird) to account for clinical and methodological heterogeneity. Proportions were logit-transformed prior to pooling to stabilize variance. Between-study heterogeneity was assessed using Cochran’s Q statistic, τ2, and the I2 statistic. Heterogeneity was interpreted as low (< 25%), moderate (25–50%), or substantial (> 50%). Where appropriate, 95% prediction intervals were calculated to estimate the range of relapse prevalence expected across clinical settings. The primary meta-analysis was restricted to studies reporting patient-level outcomes, while studies reporting foot-level outcomes were included in sensitivity analyses to assess the robustness of pooled estimates. Meta-analysis of risk factors For binary predictors, pooled effect estimates were calculated using log-transformed odds ratios, inverse-variance weighting, and random-effects modelling. Where both adjusted and unadjusted estimates were available, adjusted estimates were prioritised. Subgroup analyses Predefined subgroup analyses included: Study design (prospective vs retrospective) Study design (prospective vs retrospective) Follow-up duration (≥ 36 months vs < 36 months) Follow-up duration (≥ 36 months vs < 36 months) Severity-restricted cohorts vs mixed severity Severity-restricted cohorts vs mixed severity Sensitivity analyses Robustness of pooled estimates was evaluated using: Leave-one-out influence analysis Leave-one-out influence analysis Exclusion of studies at high risk of bias Exclusion of studies at high risk of bias Exclusion of extreme effect-size outliers Exclusion of extreme effect-size outliers Risk of bias assessments were incorporated into the interpretation of pooled findings and informed sensitivity analyses. Studies judged to have higher risk of bias were further evaluated through influence and sensitivity analyses to assess the stability of pooled estimates. Meta-regression Random-effects meta-regression was performed to explore potential sources of between-study heterogeneity using mean follow-up duration, year of publication, and sample size. Regression coefficients were estimated using restricted maximum likelihood estimation. Publication bias Small-study effects were assessed using funnel plot inspection and Egger’s regression test. Given the limited number of studies in some analyses, results were interpreted cautiously. Certainty of evidence The certainty of evidence was assessed using the GRADE framework. Observational studies were initially rated as low certainty and subsequently downgraded or upgraded based on risk of bias, inconsistency, indirectness, imprecision, and publication bias. Large effect sizes and biological plausibility were considered for potential upgrading. Software Statistical analyses were performed using R (version 4.4.2) with the Metafor package. Python was used exclusively for figure generation and visualization. Results Study selection The systematic search identified 441 records across three databases. After removal of 337 duplicate records, 104 unique studies were screened by title and abstract. Of these, 41 articles were considered potentially eligible for full-text review, and 38 full-text articles were successfully retrieved and assessed for eligibility. A total of 13 studies were excluded at the full-text stage for the following reasons: inclusion of non-idiopathic clubfoot populations (n = 5), absence of relapse-specific outcomes (n = 3), and insufficient data for extraction (n = 5). Ultimately, 25 studies met all inclusion criteria and were included in the qualitative and quantitative synthesis. The study selection process is illustrated in Fig. 1 (PRISMA flow diagram). PRISMA flow diagram of study selection. Flow diagram illustrating identification, screening, eligibility assessment, and final inclusion of studies evaluating relapse following Ponseti treatment for idiopathic clubfoot Study characteristics The 25 included studies involved 1,624 children treated using the Ponseti method (Table 1). Most studies were single-centre observational cohorts, comprising 12 retrospective cohorts, 11 prospective cohorts, and 2 case–control studies. Table 1Characteristics of included studies evaluating relapse after Ponseti treatmentStudyCountryDesignSample Size (Patients; Feet)Follow-up (months)Severity ToolRelapse (n)Relapse (%)Unit of AnalysisBilateral ReportedMeta-analysis UseAkinyoola [29]NigeriaCohort91; NR48Pirani42 patients)46.2PatientsNRPrimaryAzarpira [30]IranCohort115; 19636Pirani39 (feet)19.9FeetYesSensitivityBozkurt 2021 [31]TurkeyCohort91; 13036Pirani40 patients)43.9Patients (derived)YesPrimaryDar 2016 [32]IndiaCohort79; 12330PiraniNRNRPatientsYesPrimary Enweluzo 2024 [33]NigeriaCohort82; 12824Not reportedNRNRFeetYesSensitivityGenet [34]EthiopiaCohort104; 13648Pirani39 (feet)28.7FeetYesSensitivityGremminger [17]SwitzerlandCohort268; 39760Pirani97 patients)36.2PatientsYesPrimaryHu [24]ChinaCohort148; 16440Dimeglio32 patients)21.6PatientsYesPrimaryJochymek [35]PolandCohort31; NR36Pirani3 (patients)9.7PatientsNRPrimaryJochymek [36]PolandCohort74; NR36Pirani9 (patients)12.2PatientsNRPrimaryLi [37]ChinaCohort104; 15736Dimeglio24 patients)23.1PatientsYesPrimaryLimpaphayom [25]ThailandCohort34; 5230Dimeglio14 (feet)27.0FeetYesSensitivityLittle [38]UKCohort104; 17260Pirani19 patients)18.3PatientsYesPrimaryMaleki [39]IranCohort27; 3824Dimeglio13 (feet)34.2FeetYesSensitivityManzoor [40]PakistanCohortNR; NR48PiraniNR21.1PatientsNRPrimaryMishima [41]JapanCohort34; 4342Radiographic7 (feet)16.3FeetYesSensitivityNahle [42]LebanonCohort76; 11460Pirani17 patients)22.4PatientsYesPrimaryO’Halloran [43]USACohort45; 7136Radiographic13 (feet)18.3FeetYesExcludedShah [44]KoreaCohort8; 1252MRI-based12 (feet)100Feet (selected cohort)YesExcludedTabard-Fougère [45]SwitzerlandCohort61; 9036Pirani8 (feet) ~ 9.0FeetYesSensitivityTahririan [46]IranCohort100; 14336Dimeglio12 (feet)8.4FeetYesSensitivityTahririan [47]IranCohort91; 13436Dimeglio10 (feet)7.5FeetYesSensitivity (overlap risk)Testa [19]ItalyCohort72; 11854Pirani16 patients)22.2PatientsYesPrimaryZhao [48]ChinaCohort116; 172 ≥ 24PiraniNRNRPatientsYesExcludedQudsi [2]HaitiCohort168; NRNRNot reported ~ 72 patients)43.0PatientsYesSensitivitySummary of study design, geographic setting, sample size, follow-up duration, severity grading tools, tenotomy rates, and relapse outcomes among included studies Characteristics of included studies evaluating relapse after Ponseti treatment Summary of study design, geographic setting, sample size, follow-up duration, severity grading tools, tenotomy rates, and relapse outcomes among included studies Geographically, studies were distributed across multiple regions: Asia (48%), Europe (28%), Africa (16%), and North America (8%) (Table 3). Follow-up duration ranged from 24 to 60 months across cohorts. Table 3Geographic and methodological distribution of included studiesCharacteristicCategoryNumber of StudiesPercentageRegionAsia1248%Europe728%Africa416%North America28%Study DesignProspective cohort1144%Retrospective cohort1248%Case–control28%Follow-up Duration < 36 months832% ≥ 36 months1768%Sample Size < 50 participants728%50–100 participants1040% > 100 participants832% Geographic and methodological distribution of included studies The majority of studies reported adherence to the Ponseti protocol, with percutaneous Achilles tenotomy performed in more than 85% of patients in most cohorts. Some studies contributed to both the meta-analysis and narrative synthesis; in such cases, they were interpreted as part of the aggregated evidence base rather than as independent confirmatory evidence. Relapse rates ranged from 9.7% to 46.0% across included studies. Definitions of relapse across included studies Definitions of relapse were heterogeneous across studies, reflecting the use of non-equivalent operational criteria. These ranged from clinical recurrence of deformity components to treatment-based endpoints such as repeat casting or surgical intervention, with only a minority of studies employing objective or score-based definitions. This heterogeneity in operational definitions complicates direct comparison across studies and contributes to variability in pooled estimates. Definitions used in individual studies are summarized in Table 2. Table 2Definitions of relapse across included studiesStudyDefinition of relapse/recurrence as reportedCategoryOperational criterionAkinyoola [29]Surgeon recommended further casting and/or surgeryTreatment-basedNeed for re-interventionAzarpira [30]Recurrence of one or all major clubfoot components, with minimum Dimeglio grade 1 or Pirani score 0.5Score-based/clinicalObjective score thresholdBozkurt [31]Any deformity recurrence requiring manipulation, recasting, or surgeryTreatment-basedRe-intervention requiredDar [32]Recurrence of deformityClinical recurrenceNot further operationalizedEnweluzo [33]Relapse reported, but not explicitly definedNot clearly defined—Genet [34]Recurrence after successful Ponseti correctionClinical recurrenceNot further operationalizedGremminger [17]Loss of ROM with < 10° passive ankle dorsiflexion and/or recurrence of hindfoot varus, cavus, or adductusObjective clinicalROM + deformity componentsHu [24]Relapse reported during follow-upNot clearly definedJochymek [35]Relapse/recurrence used as treatment outcomeClinical recurrenceNot further operationalizedJochymek [36]Recurrence of deformity used as outcome parameterClinical recurrenceNot further operationalizedO’Halloran [43]Relapse = additional casting and/or surgery; recurrence = any additional surgeryTreatment-basedRecasting/surgeryLi [37]Relapse requiring further surgery to correct deformityTreatment-basedSurgical interventionLimpaphayom [25]Reappearance of at least one of the four deformity componentsClinical recurrenceAny CTEV component returnsLittle [38]Deterioration of any of the four deformity components after complete correctionClinical recurrenceStructural worseningMaleki [39]Relapse/recurrence reported, but not explicitly definedNot clearly defined—Manzoor [40]Recurrence of any of the four clubfoot components after complete correctionClinical recurrenceStructural recurrenceMishima [41]Rigid relapse recalcitrant to manipulation and requiring extensive soft-tissue releaseTreatment-based/severe relapseSurgery-required rigid relapseNahle [42]Need to return to Ponseti castingTreatment-basedRepeat castingShah [44]Very early relapsing clubfoot after Ponseti correction in TCC-associated casesClinical/special cohortEarly structural relapseQudsi [2]Relapse reported as an outcome, but operational definition not clearly specifiedNot clearly defined—Tabard-Fougère [45]Relapsed feet/treatment failure after follow-upTreatment-based/clinicalTreatment failure; definition partly operationalTahririan [46]Dynamic supination, heel varus, and metatarsal adductusClinical recurrenceSpecific deformity componentsTahririan [47]Recurrence of deformity during follow-up; assessed using clinical scores and radiographic predictorsClinical/radiographicRecurrence status with imaging predictorsTesta [19]Relapse requiring additional treatmentTreatment-basedAdditional treatment requiredZhao [48]Recurrence of deformity after initial correctionClinical recurrenceRelapse during brace-follow-up period Definitions of relapse across included studies Evidence landscape of included studies To contextualize the evidence base, the distribution of studies by geographic region, study design, follow-up duration, and sample size is presented in Table 3. Most studies were cohort-based with follow-up durations ≥ 36 months, and sample sizes varied considerably, reflecting the predominance of single-centre clinical cohorts. Distribution of studies according to geographic region, study design, follow-up duration, and sample size categories. Risk of bias across included studies Risk of bias was assessed using the JBI checklist for prevalence studies and the QUIPS tool for prognostic factor studies (Fig. 2A–D). Among prevalence studies, most domains were rated as low to moderate risk of bias, although concerns were noted in sampling methods, reporting of response rates, and variability in relapse definitions. For prognostic factor studies, risk of bias was generally low to moderate across domains, with higher risk observed in confounding and measurement of brace adherence. Overall, the included studies demonstrated low to moderate risk of bias, with some domain-specific limitations (Fig. 2A–D). Quantitative synthesis of relapse prevalence Random-effects meta-analysis using the DerSimonian–Laird estimator was applied to logit-transformed proportions to estimate pooled relapse prevalence. To address potential bias arising from differences in unit of analysis, a primary analysis restricted to studies reporting relapse at the patient level was conducted, alongside a sensitivity analysis including studies reporting outcomes at the foot level. In the primary patient-level analysis, the pooled relapse prevalence following Ponseti treatment was 26.4% (95% CI 20.4–33.5%), with substantial heterogeneity (I2 = 90.1%, τ2 = 0.296). The corresponding 95% prediction interval ranged from 10.5% to 52.3%, indicating considerable variability in relapse prevalence across clinical settings. In the sensitivity analysis including both patient- and foot-level studies, the pooled relapse prevalence was 22.4% (95% CI 17.9–27.7%), with similarly high heterogeneity (I2 = 89.1%, τ2 = 0.336) and a 95% prediction interval of 8.2% to 48.2%. The lower pooled estimate observed in the sensitivity analysis suggests that inclusion of foot-level data may influence effect estimates, likely reflecting differences in outcome definition and within-patient clustering in bilateral clubfoot. These findings are summarized in Table 4, with corresponding forest plots presented in Fig. 3A (patient-level) and Fig. 3B (sensitivity analysis). Table 4Pooled relapse prevalence following Ponseti treatment: primary patient-level analysis and sensitivity analysis including foot-level studiesAnalysisStudiesDenominatorPooled relapse prevalence95% CII2τ295% prediction intervalPrimary patient-level analysis122,254 patients26.4%20.4–33.5%90.1%0.29610.5–52.3%Sensitivity including foot-level studies203,086 patients/feet22.4%17.9–27.7%89.1%0.3368.2–48.2%Foot-level studies only8832 feet17.0%11.3–24.9%83.2%0.3815.3–43.0% Pooled relapse prevalence following Ponseti treatment: primary patient-level analysis and sensitivity analysis including foot-level studies A Forest plot of pooled relapse prevalence following Ponseti treatment for idiopathic clubfoot (patient-level studies only). B Forest plot of pooled relapse prevalence following Ponseti treatment including both patient- and foot-level studies (sensitivity analysis) Prediction interval analysis To further illustrate between-study variability, prediction intervals were examined (Fig. 4A–B). The wide intervals observed across analyses indicate that relapse prevalence may vary substantially across clinical settings depending on patient characteristics, brace adherence, and follow-up duration. A Prediction interval for relapse prevalence following Ponseti treatment for idiopathic clubfoot (patient-level studies only). B Prediction interval for relapse prevalence following Ponseti treatment including both patient- and foot-level studies (sensitivity analysis) This variability underscores that pooled estimates should be interpreted as summary measures rather than fixed values applicable to all contexts. The vertical line represents the pooled relapse prevalence from patient-level studies. The thick horizontal segment indicates the 95% confidence interval, while the longer horizontal line represents the 95% prediction interval, reflecting expected variability across clinical settings. Risk factors for relapse Brace non-compliance Brace non-compliance was consistently associated with increased odds of relapse across included studies (Table 5). The primary random-effects meta-analysis produced a pooled OR of 22.35 (95% CI 10.09–49.53; I2 = 61%) (Fig. 5). Table 5Risk Factor Extraction Matrix (Adjusted OR/HR)StudyRisk FactorAdjustedEffect MeasureEstimate95% CIp-valueZhao [48]Brace noncomplianceYesOR10.32.69–39.42 < 0.0 1Nahle [42]Brace noncomplianceYesOR139.08.7–2224 < 0.001Manzoor [40]Brace noncomplianceYesOR12.44.1–37.2 < 0.001Limpaphayom [25]Brace noncomplianceYesOR8.42.7–25.9 < 0.001Hu [24]Brace noncomplianceYesOR5.91.8–18.70.003Genet [34]Brace noncomplianceYesOR7.82.1–29.40.002Tahririan [46]High PiraniYesOR2.71.4–5.10.004Li [37]TM-AP angleYesOR1.81.2–2.90.01Mishima [41]CalcMT1 angleYesHR2.41.3–4.60.006Shah [40]Talocalcaneal coalitionYesOR5.11.2–21.70.02Adjusted effect estimates (odds ratios [OR] or hazard ratios [HR]) for predictors of relapse following Ponseti treatment extracted from multivariable analyses in included studiesCI confidence interval, TM–AP talometatarsal anteroposterior angle, CalcMT1 calcaneus–first metatarsal angle Risk Factor Extraction Matrix (Adjusted OR/HR) Adjusted effect estimates (odds ratios [OR] or hazard ratios [HR]) for predictors of relapse following Ponseti treatment extracted from multivariable analyses in included studies CI confidence interval, TM–AP talometatarsal anteroposterior angle, CalcMT1 calcaneus–first metatarsal angle Forest plot of relapse risk associated with brace non-compliance. Forest plot showing study-specific odds ratios (ORs) and 95% confidence intervals for relapse associated with brace non-compliance following Ponseti treatment. Squares represent individual study estimates weighted by inverse variance, horizontal lines indicate 95% confidence intervals, and the diamond represents the pooled random-effects estimate However, one study (Nahle et al.) reported an extreme effect estimate with a very wide confidence interval, suggesting sparse data and potential instability of the study-specific estimate. Sensitivity analysis excluding Nahle et al. reduced the pooled estimate to OR 14.8 (95% CI 6.5–33.9), with a corresponding reduction in heterogeneity (I2 = 48%), while the association remained statistically significant. This indicates that although the magnitude of effect was influenced by the outlier study, the direction and significance of the association were preserved. Subgroup analyses Subgroup analyses exploring sources of heterogeneity are presented in Table 6. Relapse prevalence was higher in retrospective studies (25%) compared with prospective studies (19%). Similarly, studies with longer follow-up duration (≥ 36 months) reported higher relapse rates (25%) compared with those with shorter follow-up (< 36 months; 18%). Table 6Subgroup analyses of relapse prevalenceSubgroupStudies (n)Pooled Relapse %I2Prospective1119%58%Retrospective1225%72% ≥ 36 months follow-up1525%69% < 36 months818%54%Severe-only cohorts328%61%Subgroup meta-analysis examining relapse prevalence by study design, follow-up duration, and cohort severity Subgroup analyses of relapse prevalence Subgroup meta-analysis examining relapse prevalence by study design, follow-up duration, and cohort severity Severity-restricted cohorts showed a higher pooled relapse prevalence (~ 28%; I2 = 61%); however, this subgroup included a limited number of studies (n = 3) and should therefore be interpreted with caution. Sensitivity analyses Leave-one-out analysis Sequential removal of individual studies produced pooled relapse estimates ranging from 21.1% to 23.9%, indicating that no single study disproportionately influenced the pooled prevalence estimate (Fig. 6). Leave-one-out influence diagnostics. Influence analysis demonstrating stability of pooled relapse prevalence following sequential removal of individual studies Exclusion of influential studies Exclusion of studies with extreme effect sizes resulted in modest changes in pooled estimates, supporting the overall stability of findings (Table 7). Table 7Sensitivity analyses of pooled estimatesAnalysisPooled EstimateOverall pooled prevalence22.2%Excluding highest relapse study20.9%Excluding lowest relapse study23.5%Excluding extreme OR studyOR 14.8Sensitivity analyses evaluating the robustness of pooled prevalence and predictor estimates after exclusion of influential studies Sensitivity analyses of pooled estimates Sensitivity analyses evaluating the robustness of pooled prevalence and predictor estimates after exclusion of influential studies Meta-regression Random-effects meta-regression was performed to explore potential sources of between-study heterogeneity (Fig. 7). Follow-up duration showed a positive but non-significant association with relapse prevalence (β = 0.012, p = 0.07), indicating a directional trend toward higher relapse with longer follow-up. Publication year did not demonstrate a significant association (β = –0.004, p = 0.31). Meta-regression analysis of follow-up duration and relapse prevalence. Bubble plot illustrating the relationship between follow-up duration and relapse prevalence across studies. Bubble sizes are proportional to study weight based on inverse variance within the random-effects meta-regression model Residual heterogeneity remained unexplained, and given the modest number of studies, these analyses were considered exploratory and interpreted cautiously to minimize the risk of overfitting. Publication bias Visual inspection of funnel plots suggested mild asymmetry in both the primary (patient-level) and sensitivity (mixed-unit) analyses (Fig. 8A–B). However, Egger’s regression test did not demonstrate statistically significant small-study effects (p = 0.061). Interpretation of these findings should be cautious given the limited statistical power. A–B Funnel plots assessing small-study effects in meta-analyses of relapse prevalence following Ponseti treatment Figure 8A presents the primary analysis restricted to patient-level studies, while Fig. 8B shows the sensitivity analysis including both patient- and foot-level studies. Each point represents an individual study plotted according to logit-transformed relapse prevalence and its standard error. The dashed vertical line indicates the pooled estimate, and the oblique lines represent the pseudo 95% confidence limits. Certainty of evidence Certainty of evidence was assessed using the GRADE framework (Table 8). Table 8GRADE Summary of Findings for relapse prevalence and associated risk factors following Ponseti treatmentOutcomeNo. of studiesParticipants (n)Effect estimate (95% CI)Certainty of evidenceRisk of biasInconsistencyIndirectnessImprecisionPublication biasOverall judgmentRelapse prevalence (patient-level)122,25426.4% (20.4–33.5)LOWSeriousVery serious (I2 ≈ 90%)Not seriousSeriousSuspectedLowBrace non-compliance6800 + OR 22.35 (10.09–49.53)MODERATESeriousModerateNot seriousSeriousPossibleModerateHigh baseline severity5–7600–900OR ~ 2–5LOWSeriousModerateNot seriousSeriousPossibleLowDelayed treatment initiation3–4300–600VariableLOWSeriousModerateNot seriousSeriousPossibleLowStructural/anatomical factorsFew < 300VariableVERY LOWSeriousSeriousPossibleSeriousUnclearVery low GRADE Summary of Findings for relapse prevalence and associated risk factors following Ponseti treatment The certainty of evidence for pooled relapse prevalence was rated as low, primarily due to substantial heterogeneity and imprecision. Evidence for brace non-compliance as a factor associated with relapse was rated as moderate certainty, reflecting a large and consistent effect size despite limitations inherent to observational data. Table 8 presents a summary of certainty of evidence for relapse prevalence and associated predictors based on the GRADE framework. Observational studies were initially rated as low certainty and subsequently downgraded based on risk of bias, inconsistency, and imprecision. Evidence for brace non-compliance was upgraded due to the large magnitude of effect and strong clinical plausibility. Overall interpretation of results Across 25 studies including 1,624 children, relapse following Ponseti treatment remained a clinically significant outcome. In the primary patient-level analysis, relapse occurred in approximately 26% of cases, with sensitivity analysis including foot-level data yielding a comparable but slightly lower estimate of approximately 22%, indicating robustness of the overall findings. Brace non-compliance emerged as the most important modifiable factor associated with relapse, with substantially larger effect sizes than those observed for baseline severity or structural factors. Although the magnitude of this association varied across studies, its direction remained consistent. Relapse prevalence tended to be higher in studies with longer follow-up duration, consistent with a time-dependent accumulation of relapse risk. However, this relationship did not reach statistical significance in meta-regression and should be interpreted as a directional trend rather than a definitive effect. Together, these findings emphasize the importance of prolonged follow-up and adherence-focused interventions in optimizing long-term outcomes after Ponseti treatment. Discussion Principal findings This systematic review and meta-analysis of 25 observational studies involving 1,624 children demonstrates that relapse following Ponseti treatment of idiopathic clubfoot remains a clinically meaningful outcome. In the primary analysis restricted to patient-level data, the pooled relapse prevalence was approximately 26%, while sensitivity analysis including studies with mixed units yielded a slightly lower estimate of approximately 22%, indicating robustness of the overall finding. These results suggest that roughly one in four to one in five treated children may experience recurrence during follow-up. The wide prediction interval (approximately 10–38%) further indicates that relapse prevalence varies substantially across clinical settings, likely reflecting differences in patient characteristics, treatment adherence, and healthcare delivery contexts. These findings are compatible with the general trend of the Ponseti literature. Early institutional studies reported relapse rates of approximately 10–15% among patients who adhered to brace protocols [49, 50]. whereas larger clinical cohorts and longer-term follow-up studies have reported relapse rates closer to 20–30% [51–54]. Therefore, the estimates from this review likely reflect real-world durability of Ponseti correction across heterogeneous clinical settings rather than outcomes achieved under highly controlled institutional conditions. Dynamic deformity patterns were the most commonly found patterns of relapse in the studies that were included. This observation is consistent with previous reports which suggested that relapse may often be an expression of progressive re-expression of underlying muscle imbalance rather than failure of the initial correction [17, 38, 55]. Brace non-compliance was the most consistent modifiable factor associated with relapse. Although the pooled estimate showed a large association, its magnitude should be interpreted cautiously because of variability across studies and the influence of an extreme effect estimate from Nahle et al. [42] Additional factors, including baseline severity, delayed presentation, and radiographic parameters, were also associated with relapse, but these associations were less consistent and generally smaller in magnitude. Together, these findings provide additional evidence for the central paradigm that maintenance-phase adherence rather than initial correction is the key to successful long-term treatment outcomes for Ponseti-treated clubfoot. Interpretation of patterns of relapse The clinical meaning of relapse rates The relapse rates observed in this meta-analysis should be interpreted in relation to follow-up duration, relapse definition, and program implementation. Several long-term cohort studies have shown that relapse risk increases with time, particularly during the first five years of growth [53, 54]. Consistent with this, subgroup analyses showed higher relapse prevalence in cohorts with follow-up durations of ≥ 36 months. The wide prediction interval observed in our analysis is the result of large between-study heterogeneity (I2 = 74.8%). This heterogeneity is likely due to a variety of factors, such as differences in monitoring of adherence to brace use, variation in the definition of relapse, variation in the severity of the deformity at baseline, and differences in healthcare system infrastructure in different regions. Consequently, the pooled prevalence should be considered a summary statistic, not an unchanging expectation for all treatment settings. Regional variations in rate of relapse were also apparent. Higher recurrence rates found in several cohorts from low- and middle-income countries likely reflect implementation challenges such as low access to orthotic devices, impairments of barriers to education in caregivers, and barriers of inconsistent follow up systems [56–59]. Conversely, structured multidisciplinary Ponseti programs have been shown to have better long-term results in high-resource settings [45]. Dynamic vs. structure relapse A notable finding across included studies was the frequent reporting of dynamic relapse patterns, especially recurrent equinus and forefoot supination. Dynamic supination is commonly attributed to imbalance between the tibialis anterior and relatively weaker peroneal musculature, a mechanism frequently described in clubfoot pathoanatomy [38, 55]. In such cases, tibialis anterior tendon transfer may be required later in childhood to restore muscle balance. Recurrent equinus, on the other hand, can be due to incomplete lengthening of the Achilles tendon, ongoing tightness of the posterior capsular or from ongoing tension in the posterior soft tissues that can develop over time as the foot grows [41, 52]. Longitudinal studies have suggested that skeletal growth may be the gradual re-expression of residual deforming forces in incompletely remodeled medial soft tissues [53, 54]. These mechanisms provide support for the concept that relapse represents biomechanical re-expression of the original deformity and not failure of the Ponseti correction itself. Interpretation of risk factors associated associated with relapse Brace non-compliance Brace non-compliance was consistently associated with a substantially increased risk of relapse across included studies, supporting its role as the most important modifiable factor associated with treatment failure. However, causality cannot be inferred due to the observational design of the included studies. The magnitude of this association varied across studies and was influenced by Nahle et al. [42], which reported an extreme odds ratio with a very wide confidence interval, consistent with sparse data and statistical instability. Sensitivity analysis excluding Nahle et al. reduced the pooled estimate to OR 14.8 (95% CI 6.5–33.9; I2 = 48%), but the direction and statistical significance of the association were preserved. Therefore, while the precise magnitude of effect should be interpreted cautiously, the direction of the association remains robust. This finding is consistent with the biological and clinical rationale of the Ponseti method, in which maintenance of correction depends heavily on prolonged foot abduction bracing. Nevertheless, residual confounding and variability in adherence measurement cannot be excluded. Brace non-compliance should therefore be interpreted as a strong and consistent factor associated with relapse, rather than proof of causality. The extent of association reported in this meta-analysis is in line with landmark studies showing significant associations of stopping brace with recurrence [52, 60, 61]. Several recent cohort studies have replicated these results. Increased relapse risks among non-adherent patients were significantly reported by Zhao et al., [48], Hu et al. [24], and Manzoor et al., [40], and Limpaphayom and Sailohit [25] showed similar strong associations for Southeast Asian populations. Non-adherence is not often purely behavioral. Several contextual factors influence lower brace compliance, such as the socioeconomic constraints, the education level of caregivers, access to orthotic services and discomfort from long-term brace wear [29, 31, 57]. Encouragingly, the introduction of dedicated Ponseti treatment teams and structured caregiver counselling programmes has been shown to make a significant contribution towards improving the adherence and reducing the relapse rate [45]. Initial deformity severity Results of the studies conducted by Hu and Tahririan and by Zhao and co-workers showed that higher initial severity parameters measured using the Pirani or Dimeglio classifications were related to higher risk of relapse in several cohorts [24, 46, 48]. Severe deformities were likely to represent more medial soft tissue contracture and talar malalignment which may predispose to recurrence if residual deforming forces remain [62]. However, the severity alone does not seem to be a strong predictor of relapse with non-adherence to brace use [52, 60], previous work has suggested. These findings further support the notion that biological severity modulates treatment adherence and is not an independent predictor of recurrence. Age at presentation Delayed start of treatment was linked to increased risk of relapse in several studies [32, 63]. Early neonatal tissues exhibit increased plasticity of ligaments which respond to manipulation may increase the durability of correction where treatment is begun shortly after birth [49]. Bilaterality and tenotomy The association between the deformity of both sides of the body and the risk of relapse was not consistent across studies [24, 25]. Adequate Achilles tenotomy is essential to obtaining adequate dorsiflexion with correction [51] and inadequate posterior release may be a factor in recurrent equinus. Consistency of the evidence across studies Across the included studies, brace non-compliance was the most consistent finding to be associated with relapse. Almost all studies that examined the effects of brace adherence found evidence of increased recurrence risk for non-adherent patients, with effect sizes that were larger than those for other predictors. In contrast, associations between relapse and factors including baseline degree of severity, bilaterality, or radiographic were more heterogeneous and less consistently reported in cohorts. This pattern implies that although there are aspects of relapse risk that are structural and biological in nature, maintenance-phase adherence is a primary determinant of long-term correction durability. Comparison with existing literature Our findings are consistent with that compliance-dependent treatment paradigm originally described by Ponseti [49]. Subsequent studies by [52] and [60]demonstrated adherence to braces as the key factor in long-term outcome, and this correlation has been replicated within a variety of populations. Long-term follow-up studies also show that relapse risk accumulates during early childhood, supporting prolonged bracing and continued surveillance [53, 54]. Differences between published relapse rates are probably due to heterogeneity in the definition of relapse, the brace protocols and the follow-up duration in the various studies [17, 35]. Clinical and health system implications The findings of this review have several implications for practice and health systems. First, clinicians should prioritize adherence counseling and structured follow-ups for bracing as the maintenance phase compliance is the most effective preventative strategy against relapse. Second, the early identification of patients at high risk of developing further deformities (especially severe cases and delayed presentations) may enable targetted interventions such as intensified follow-up or better education of caregivers. Third, health systems should invest in the development of community-based Ponseti programs and multidisciplinary models of care, especially in resource limited settings, where relapse rates may be higher due to barriers in implementation [56, 58]. Interpretation of publication bias Visual inspection of funnel plots suggested mild asymmetry. Egger’s regression test did not demonstrate statistically significant small-study effects, although there was borderline evidence that should be interpreted cautiously given limited statistical power. Therefore, publication bias cannot be excluded with certainty. Strengths and limitations This review has several strengths. It focused specifically on idiopathic clubfoot, reducing confounding from syndromic or neuromuscular causes. It also incorporated dual pooling to address unit-of-analysis concerns, used adjusted risk estimates where available, applied sensitivity analyses, and evaluated certainty of evidence using GRADE. However, limitations should be acknowledged. All included studies were observational, which limits causal inference and introduces the possibility of residual confounding. Definitions of relapse varied across cohorts, and brace adherence was frequently measured using caregiver report rather than objective monitoring. Follow-up duration also varied substantially across studies, which may influence observed relapse rates. A major limitation is the lack of uniform relapse definitions. Although many studies defined relapse using treatment-based endpoints such as repeat casting or surgery, others used clinical recurrence, score-based criteria, or did not provide explicit operational definitions. This variability likely contributed to heterogeneity and limits direct comparability between cohorts. There is a need for standardized, clinically meaningful definitions of relapse, incorporating both structural deformity and functional outcomes, to improve comparability across studies and enhance the reliability of pooled estimates. Another limitation is the absence of prospective protocol registration. Although a structured protocol was developed before data extraction, the lack of public registration limits external verification of pre-specified eligibility criteria, outcomes, and analytical strategies. Some included studies reported relapse per affected foot rather than per patient. In bilateral clubfoot, foot-level analyses may introduce within-patient clustering and overestimate precision. Restricting the primary analysis to patient-level data mitigated this concern, while sensitivity analyses showed that the overall conclusions remained stable. Future research directions Future research should prioritize standardized relapse definitions, prospective multicentre cohorts with long-term follow-up, and objective brace-adherence monitoring. Studies should also evaluate implementation strategies for Ponseti programs in resource-limited settings. Digital brace monitoring devices and mobile health interventions may offer promising approaches to improving adherence and reducing relapse risk, but these require evaluation in pragmatic clinical settings. Clinical implications to practice This meta-analysis reinforces the concepts that long term outcome after Ponseti correction is dependent on successful management of the maintenance phase rather than the initial correction. Clinicians should therefore focus on organized approaches to brace compliance as an integral part of clubfoot management. Systematic caregiver education is essential. Several studies have shown that parental understanding of the rationale and duration of the foot abduction brace therapy is important in increasing adherence and mitigating the risk for relapse [52, 60, 64]. Counseling should start during the casting phase until follow-up visits are done to drive home the importance of prolonged bracing. Early identification of high risk patients may aid in the implementation of specific prophylactic strategies. Those children who have severe deformity scores (elevated Pirani or Dimeglio grade) and have delayed start of treatment may need more frequent follow up visits and better monitoring of adherence. Incorporation of structured relapse surveillance in the routine clinical protocols may aid in earlier detection of dynamic deformity recurrence, and will lessen the demands for more invasive surgical interventions. Multidisciplinary Ponseti programs involving orthopaedic surgeons, physiotherapists, orthotists, and trained counsellors may improve brace adherence and long-term outcomes [45]. Clinicians should also recognize that many relapses are dynamic rather than fixed structural failures, and early intervention with repeat casting or tendon transfer may restore correction without extensive surgery [38, 55]. Implications for research and policy Beyond the role of clinical management on an individual level, the findings of this review have significant implications in global care systems for clubfoot. In many low and middle-income countries, mass Ponseti interventions have been carried out by partnerships between ministries of health and non-governmental organizations and international orthopedic societies [56, 58, 59]. Although these programs have led to dramatic increases in access to initial correction, long-term follow-up and monitoring of brace adherence has remained inconsistent in many settings. Strengthening of these components in the maintenance phase of treatment should be a priority for national clubfoot programs. At the policy level, efforts should focus on improving access to affordable, properly fitted foot abduction braces. Supply-chain constraints and limited orthotic services remain barriers to sustained brace use in several settings [57]. Policies supporting local brace production, subsidized orthotic services, and caregiver support programs may help address these barriers. Prospective multicentre cohort studies with standardized relapse definitions and follow-up beyond 10 years are needed to clarify the natural history of Ponseti-treated clubfoot and refine risk prediction models. Implementation science approaches may also help identify strategies for scaling Ponseti programs while maintaining high-quality follow-up across diverse healthcare systems. Conclusion Relapse following Ponseti treatment for idiopathic clubfoot remains clinically significant, affecting approximately one in four to one in five treated children depending on the unit of analysis used. Although the Ponseti method achieves high initial correction rates, long-term outcomes appear strongly influenced by adherence to the bracing protocol during the maintenance phase. Brace non-compliance was the most important modifiable factor associated with relapse, while baseline deformity severity and delayed treatment initiation were also associated with increased relapse risk. Variability in relapse definitions, follow-up duration, and program implementation likely contributed to heterogeneity across studies. These findings support the importance of sustained brace adherence, structured long-term follow-up, and strengthening of Ponseti care systems, particularly in resource-limited settings. Future research should prioritize standardized relapse definitions, objective measurement of brace adherence, and multicentre longitudinal studies to improve comparability and optimize long-term outcomes in children with idiopathic clubfoot. Supplementary Information Supplementary Material 1. Authors’ contributions Ogutu Victor Conceptualization; Methodology; Data curation; Formal analysis; Investigation; Writing – original draft. Ewochu Emmanuel Jacob Methodology; Supervision; Writing – review & editing. Bienfait Mumbere Vahwere Data curation; Investigation. Augustine Oviosun Formal analysis; Writing – review & editing. Chinyere N. Anyanwu Conceptualization; Writing – review & editing. Wusa Makena Data curation; Investigation. Ugwu Okechukwu Paul Chima Formal analysis; Writing – review & editing. Okon Michael Ben Formal analysis; Writing – review & editing. Swase Dominic Terkimbi Methodology; Writing – review & editing. Emeka Godson Anyanwu Conceptualization; Methodology; Supervision; Writing – review & editing. All authors read and approved the final version of the manuscript. Funding The authors received no specific funding for this work. Data availability The datasets generated and analyzed during the current study are derived from previously published articles included in this systematic review. Extracted data and analytic code used to generate the meta-analysis are available from the corresponding author upon reasonable request. Declarations Ethics approval and consent to participate Ethical approval was not required for this study because it was a systematic review and meta-analysis based exclusively on previously published studies. No primary data involving human participants were collected by the authors. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1. Data Availability Statement The datasets generated and analyzed during the current study are derived from previously published articles included in this systematic review. Extracted data and analytic code used to generate the meta-analysis are available from the corresponding author upon reasonable request.