A investigation of suitability of EQ-5D-Y proxy version to assess health of children aged from 4 years

Rationale: The EQ-5D-Y was developed and validated on youths aged from 8 years and considered able to self-complete it. A proxy version of the EQ-5D-Y was then tested on children of the same age. Some children-specific questionnaires allow to assess QoL from the age of 2 or 4 years, reported by proxy responders. The suitability and performance of EQ-5D-Y applied on children younger than 8 was not tested. Objective: to test if EQ-5D-Y proxy could be used to assess QoL of children aged from 4 to 7 years. Methods: To reach our objective, the target population was sampled from schools and kindergartens. For stake of comparison, we also analysed data on 8-11 years old schoolchildren and on children affected with severe diseases. The “third person” version of the EQ-5D-Y proxy was adopted: the responder was asked to report what the child would report. After completing the EQ-5D-Y, the responder was asked to specify, for each profile domain, the reason of his answer. We analysed the EQ-5D-Y feasibility and perceived appropriateness (with a scale rated 0=not appropriate to 10=very appropriate) on target children, the construct (with the PedsQL as a standard tool) and known–group validity, and investigate the reasons why difficulties were reported. Results: 295 valid children from kindergartens or schools. One-hundred-twenty severely ill patients with the same age were involved from hospitals. Among schoolchildren, 94% (10 y.o.) to 100% (11 y.o.) had no difficulties in walking about. Children with no difficulties with self-care were 34-42% among those aged 4-5 and 73-100% among those aged 6-11 years. Doing usual activities was not a problem for 74% (5 y.o.) to 90% (11 y.o.) of children. Children with no pain/discomfort were 50% (5 y.o.) to 80% (7 y.o.). Absence of unhappiness/worry/sadness was reported for 30% (11 y.o.) to 85% (4 y.o.) of children. The questionnaire was completed with no (profile) or 0.7% (VAS) missing answers. Median appropriateness was 7-8. Construct validity was fair for usual activities, good for the other 4 domains. Construct validity was better as regards the ill children group. Known-group validity was good. Among responders specifying the reasons of difficulties reported on schoolchildren, 46.2% (mobility), 59.7% (self-care) and 29.2% (usual activities) reported that the domain was not suitable to the child’s age. This type of explanation was more frequently reported about children aged 4-5 years. In contrast, having pain/discomfort or unhappiness/worry/sadness were not considered unsuitable to the children’s young age. In contrast, apart from few cases, reasons of difficulties reported on ill children were attributed to their diseases or treatment. Discussion: EQ-5D-Y proxy version shows to be potentially a valid instrument for the assessment of QoL of children from 4 years. However, in its current version, some domains are sometimes perceived as not suitable to young children. Accordingly, we consider worth discussing whether and how to update, yet keeping its current structure, the recently developed EQ-5D-Y proxy version, in order to optimize its suitability for the assessment of QoL of all children aged from 4 years.