Stability of the Gross Motor Function Classification System over time in children with cerebral palsy

Aim: To assess the stability of the Gross Motor Functional Classification System (GMFCS) in children with cerebral palsy (CP) from time of preliminary diagnosis (similar to 2 years of age) to time of diagnosis (similar to 5 years of age), and to examine factors associated with reclassification. Method: We conducted a longitudinal study using a sample from the Canadian CP Registry. Stability was analysed by using the percentage of agreement between time-points and a weighted prevalence and bias adjusted kappa statistic. Univariate and multivariate logistic regressions were performed to identify variables associated with reclassification. Results: The study included 1670 children (857 males, 713 females) with a mean age of 11 years 4 months (SD 4 years, range 3 years 5 months-20 years 1 month) at time of data extraction (3rd September 2019), of which 1435 (85.9%) maintained a stable GMFCS, with a weighted kappa of 0.91 (95% confidence interval 0.89-0.92). Univariate logistic regression showed that initial GMFCS level, CP subtype, and the presence of cognitive impairment were associated with the likelihood of change in the GMFCS level (p <0.1). In the multivariate analysis, however, the likelihood was associated with initial GMFCS level only (odds ratio 7.10-8.88, p <0.00). Interpretation: The GMFCS has good stability in early childhood. For the majority of children, it is predictive of their long-term motor function.