Patterns of hip migration in non-ambulant children with cerebral palsy: A prospective cohort study.

• The prevalence of hip migration requiring surgery is 24.3% in children with cerebral palsy. • The most advanced cases are not for the oldest children. • Hips of most children remain stable over time (up to 5 years). • Pelvic obliquity and hip migration are probably linked. • Three trajectories of hip migration may be distinguished. In children with cerebral palsy (CP), we have little information on when hip migration (HM) starts, what causes hip displacement, how HM changes over time, and how to halt this migration to avoid surgery. We aimed to estimate the prevalence of HM percentage (HMP) >4 0% in a homogeneous population of non-ambulant children with CP and model the changes in HMP over a 2.6-year mean follow-up. From September 2009 to September 2015, this observational, prospective, multicenter cohort study recruited 235 children from 51 centers who were 3 to 10 years old and had levels IV and V of the Gross Motor Function Classification System for CP. The outcomes were yearly HMP measurements by the Reimers index. Only children with at least one hip with HMP ≤ 40% at baseline were included in trajectory modeling. Comparisons of chidren's characteristics between trajectory groups were adjusted by the false discovery rate method. The prevalence of children with at least one hip with HMP > 40% was estimated at 24.3% (95% confidence interval 18.6–30.0). Pelvic obliquity was observed in 51.4% and 24.4% of children with asymmetric and symmetric HMP (P = 0.002). The trajectory modelling identified 3 types of MP changes over time. Many children (67.4% and 79.3% for the right and left hip) could be assigned to the "stable" trajectory group. In non-ambulant children with CP, the prevalence of HM requiring surgery is low and most hips remain practically stable over time. [ABSTRACT FROM AUTHOR]