The EQ-5D-Y for personal interview: development and preliminary psychometric testing with children aged 6 to 8 years

Rationale The availability of EQ-5D-Y in a format suitable to be used in face-to-face interviews could be beneficial to extend its applicability to children who are already able to report on their HRQL, but are too young or too ill, to read and complete the EQ-5D-Y in autonomy. Objective The objective of the present research was two-fold: 1) to develop a version of the EQ-5D-Y suitable to be completed by interviewing the youth; 2) to test psychometric properties of the instrument in children aged 6 to 8, i.e., younger or at the lower burden of applicability of the self-administered EQ-5D-Y. Methods The EQ-5D-Y was slightly modified by adding instructions for the interviewer and a supplementary A4-sheet with three smiles associated with the 3 severity levels for the profile. The interviewer reads the profile and by showing the smiles to the child asks him/her to specify his/her own response. Then the VAS is shown by requesting to indicate the point of the scale to mark. A sample of parents and their children aged 6-8 years received a booklet including the EQ-5D-Y for interview, the PedsQL for interview, and a set of feasibility checks (self-completed by the parent). To test external test-retest reliability and validity, within 7 days, children were re-interviewed by a psychologist, who asked them to answer again the EQ-5D-Y questions and to specify the reasons of their answers. Results Sixty children (63.3% females; 22 aged 6, 13 aged 7, 25 aged 8 years) and their parents participated. Moderate difficulties were reported by 3.3% of the children with Mobility, 25% with Self-Care, 13.3% with Daily Activities; 31.1% reported some Pain/Discomfort, and 11.7% some Worry/Sadness/Unhappiness. One child reported severe difficulties with Self-Care and one with Daily Activities. Content analysis of open answers indicates that no child reported difficulties with Self-care for health-related problems, but rather because of age (not being capable to wash/dress in autonomy), or lack of will/pleasure in doing such activities. Seven children also justified difficulties with usual activities with reasons not related to health. The VAS (mean+SD=91.87+11.88, median=100 range: 55-100) was correctly completed by 95% of respondents (1 incorrect completion, 2 missing cases). Feasibility checks (10-points Likert scales 10=completely agree) indicate that parents rated the instrument as easy and fast to complete, and the smiles’ sheet as helpful to the children (all median values=9.00). Convergent validity, compared to the PedsQL for interview, was satisfactory as to the VAS and to Mobility and Pain/Discomfort domains, but not for domains 2, 3, and 5. As to test-retest reliability, the median difference between the VAS scores was 0 (25th/75th percentiles=+5.0). Percentages of agreement with the profile domains ranged between 76.6% (Self-care) and 100% (Mobility). Discussion The proposed EQ-5D-Y for interview appears to be feasible and easy to administer. However, applicability to 6-8 years old children is not supported, mainly because of weaknesses in domains 2 and (partly) 3 and 5, as they are presently formulated. Future advancements in instrument development need to be devoted to overcome these limits.