Pectus carinatum (PC) is a common chest wall deformity, but there is currently a lack of predictive models and tools for forecasting deformity improvement with compressive orthotic bracing (COB) therapy. This study aimed to identify key factors influencing treatment outcomes, and to develop and integrate a predictive efficacy scoring tool with a clinical decision pathway to provide evidence-based treatment recommendations for pediatric PC patients.

In this retrospective study, 182 pediatric PC patients evaluated and followed using three-dimensional scanning were enrolled. A multiple linear regression model was developed to create and validate a predictive efficacy scoring tool, which subsequently informed the establishment of a treatment decision pathway.

The predictive model identified smaller initial external thoracic width (p < 0.001, β = − 0.644), older age at treatment initiation (p = 0.04, β = 0.271), good compliance (p = 0.008, β = 0.188), and favorable chest wall morphology (p = 0.033, β = 0.152) as independent predictors of deformity improvement. The model was optimized into a predictive efficacy scoring tool (Score = Expected %dEHI + 4.761). Comparison between the good response (score > 0) and poor response (score ≤ 0) groups confirmed the tool's excellent discriminative ability (8.348 vs. 15.320, p < 0.001).

Treatment recommendations were derived by integrating the scoring tool with the decision pathway: children with score > 0 are strongly recommended to continue bracing; those with score ≤ 0 due to poor compliance should receive intensified supervision, while those with score ≤ 0 due to other factors should be counseled on possible surgery; children under 4 years require personalized recommendations integrating the score, symptoms, and parental preferences. However, as a single-center retrospective study without a control group and with subjective compliance assessment, future prospective multicenter studies are required to validate the model's generalizability.

H.W.: conceptualization, data curation, investigation, methodology, writing—original draft, writing—review and editing; Y.H.: resources, software; M.L.: formal analysis; J.L.: supervision; L.L.: supervision, validation; J.Z.: validation; Q.Y.: project administration; Z.L.: visualization; J.L.: validation.

Received: 04 September 2025

Accepted: 07 January 2026

Accepted Manuscript online:
09 January 2026

Article published online:
21 January 2026

© 2026. Thieme. All rights reserved.

Georg Thieme Verlag KG
Oswald-Hesse-Straße 50, 70469 Stuttgart, Germany