The performance of the EQ-5D-Y-5L compared to the EQ-5D-Y-3L in children and adolescents with cerebral palsy (CP).

Objectives: The aim of this study is to compare the performance and validity of the EQ-5D-Y-3L (Y-3L) and EQ-5D-Y-5L (Y-5L) in South African children and adolescents with cerebral palsy (CP).

Methods: Children/adolescents with CP and those from the general population completed the Y-5L, Y-3L, and PedsQL. Physiotherapists at the school classified participants' functional ability on the Gross Motor Functioning Classification System (GMFCS).

Results: Fifty-one children/adolescents completed the measures. The ceiling effect had a 44% relative reduction for and floor effects decreased across all dimensions except for when moving from the Y-3L to Y-5L. Informativity of dimensions improved on average by 0.27 on the Y-5L with similar evenness. There was a range of 6-16% inconsistent responses when moving from the Y-3L to the Y-5L. Convergent validity was strong on paired Y-3L and Y-5L dimensions: Kendall's Tau B (range 0.53 - 0.85) and Gamma (range 0.79 - 0.99). There was significant moderate association between Y-3L and Y-5L with similar items on the PedsQL. The physical dimensions of and were significantly associated with GMFCS with those having less independent mobility reporting more severe problems on dimension scores.

Conclusion: The Y-5L showed a notable reduction in ceiling and floor effects, improved discriminatory power, higher criterion validity with the GMFCS and similar concurrent validity with the PedsQL as the Y-3L. It is recommended that the Y-5L is further tested for reliability and responsiveness in this population group so that its utility for detecting change in clinical trials or as a routine outcome measure can be determined.