Related Experiment Video
Updated: Sep 25, 2025
Author Spotlight: A Novel 3D-Printed Titanium Implant for Minimally Invasive Treatment of Hip Dysplasia in Young Dogs
Published on: April 19, 2024
The CPUP Hip Score predicts displacement of the hip in children with cerebral palsy
Mark S Gaston1, Sarah J Wordie2, Philippe Wagner3
1Royal Hospital for Sick Children, Edinburgh, UK.
Insights
The Cerebral Palsy (CP) UP Hip Score effectively identifies children with cerebral palsy at risk for hip displacement. While accurate in discrimination, it overestimated risks, suggesting potential need for recalibration in Scottish populations.
Area of Science:
- Orthopedics
- Pediatric Medicine
- Rehabilitation Medicine
Background:
- The Uppföljningsprogram för cerebral pares (CPUP) Hip Score is a tool developed in Sweden to assess hip displacement risk in children with cerebral palsy (CP).
- Its validity and calibration have not been previously confirmed in populations outside of Sweden.
- Hip displacement is a significant concern in children with CP, particularly those with higher Gross Motor Function Classification System (GMFCS) levels.
Purpose of the Study:
- To evaluate the calibration and discriminatory accuracy of the CPUP Hip Score in a Scottish cohort of children with CP.
- To determine if the score's performance in predicting hip displacement risk is consistent across different populations.
- To assess the potential need for score recalibration for clinical use in Scotland.
Main Methods:
- A population-based study utilizing data from the Cerebral Palsy Integrated Pathway Scotland.
- Inclusion criteria focused on children with CP classified in GMFCS levels III to V.
- Hip displacement was defined as a migration percentage (MP) exceeding 40%. Calibration was assessed using calibration slope and Kaplan-Meier curves; discriminatory accuracy was measured by the time-dependent area under the receiver operating characteristic curve (AUC).
Main Results:
- The CPUP Hip Score demonstrated high discriminatory accuracy in the Scottish cohort (AUC 0.78).
- However, the score's calibration was insufficient, with an overestimation of absolute risks for hip displacement (slope 0.48).
- Re-estimation of score weights, using original predictive variables, significantly improved discriminatory accuracy (AUC 0.85).
Conclusions:
- The CPUP Hip Score possesses a strong capacity to differentiate risk levels for hip displacement in children with CP.
- Overestimation of absolute risks in the Scottish cohort suggests potential differences in patient registration between Swedish and Scottish programs.
- While promising, recalibration of the score's weights may be necessary for reliable clinical application in Scotland.
Aims:
The Uppföljningsprogram för cerebral pares (CPUP) Hip Score distinguishes between children with cerebral palsy (CP) at different levels of risk for displacement of the hip. The score was constructed using data from Swedish children with CP, but has not been confirmed in any other population. The aim of this study was to determine the calibration and discriminatory accuracy of this score in children with CP in Scotland.
Methods:
This was a total population-based study of children registered with the Cerebral Palsy Integrated Pathway Scotland. Displacement of the hip was defined as a migration percentage (MP) of > 40%. Inclusion criteria were children in Gross Motor Function Classification System (GMFCS) levels III to V. The calibration slope was estimated and Kaplan-Meier curves produced for five strata of CPUP scores to compare the observed with the predicted risk of displacement of the hip at five years. For discriminatory accuracy, the time-dependent area under the receiver operating characteristic curve (AUC) was estimated. In order to analyze differences in the performance of the score between cohorts, score weights, and subsequently the AUC, were re-estimated using the variables of the original score: the child's age at the first examination, GMFCS level, head shaft angle, and MP of the worst hip in a logistic regression with imputation of outcomes for those with incomplete follow-up.
Results:
The discriminatory accuracy of the score in the new population of 367 children was high (AUC 0.78 (95% confidence interval (CI) 0.71 to 0.86)). The calibration of the score was insufficient (slope 0.48 (95% CI 0.31 to 0.65)), and the absolute risks of displacement of the hip in this population were overestimated. The AUC increased with re-estimated weights (0.85 (95% CI 0.79 to 0.91)).
Conclusion:
The CPUP Hip Score had a high ability to discriminate between children at different levels of risk for displacement of the hip. The score overestimated the absolute risks of displacement in this population, which may have resulted from differences in the way children were initially registered in the two programmes. The results are promising, but the score weights may need re-estimation before its clinical application in Scotland. Cite this article: Bone Joint J 2022;104-B(5):640-644.
