Development and validation of a nomogram for predicting intellectual disability in children with cerebral palsy

• Junying Yuan ^[1] ; Gailing Wang ^[1] ; Mengyue Li ^[1] ; Lingling Zhang ^[1] ; Longyuan He ^[1] ; Yiran Xu ^[1] ; Dengna Zhu ^[1] ; Zhen Yang ^[1] ; Wending Xin ^[1] ; Erliang Sun ^[1] ; Wei Zhang ^[1] ; Li Li ^[1] ; Xiaoli Zhang ^[1] ; Changlian Zhu ^[1]
  1. [1] Third Affiliated Hospital of Zhengzhou University

     Third Affiliated Hospital of Zhengzhou University

     China

     
• Localización: International journal of clinical and health psychology, ISSN 1697-2600, Vol. 24, Nº. 3, 2024, págs. 181-190
• Idioma: inglés
• Enlaces
  • Texto completo
• Resumen

  • Objective Intellectual disability (ID) is a prevalent comorbidity in children with cerebral palsy (CP), presenting significant challenges to individuals, families and society. This study aims to develop a predictive model to assess the risk of ID in children with CP.

    Methods We analyzed data from 885 children diagnosed with CP, among whom 377 had ID. Using least absolute shrinkage and selection operator regression, along with univariate and multivariate logistic regression, we identified key predictors for ID. Model performance was evaluated through receiver operating characteristic curves, calibration plots, and decision curve analysis (DCA). Bootstrapping validation was also employed.

    Results The predictive nomogram included variables such as preterm birth, CP subtypes, Gross Motor Function Classification System level, MRI classification category, epilepsy status and hearing loss. The model demonstrated strong discrimination with an area under the receiver operating characteristic curve (AUC) of 0.781 (95% CI: 0.7504-0.8116) and a bootstrapped AUC of 0.7624 (95% CI: 0.7216-0.8032). Calibration plots and the Hosmer-Lemeshow test indicated a good fit (χ2= 7.9061, p = 0.4427). DCA confirmed the model's clinical utility. The cases were randomly divided into test group and validation group at a 7:3 ratio, demonstrating strong discrimination, good fit and clinical utility; similar results were found when stratified by sex.

    Conclusions This predictive model effectively identifies children with CP at a high risk for ID, facilitating early intervention strategies. Stratified risk categories provide precise guidance for clinical management, aiming to optimize outcomes for children with CP by leveraging neuroplasticity during early childhood.