Health State Utility Values in Children and Adolescents with Disabilities: A Systematic Review

TL;DR

This review summarizes methods for measuring health utility values in children and adolescents with disabilities, highlighting issues with current approaches and the need for better tools.

Contribution

The study provides a systematic review of methods for health state utility values in children with disabilities, emphasizing methodological limitations and the need for validated instruments.

Findings

Only 2 of 31 studies used direct methods for measuring health state utility values.

Generic measures like EuroQol 5 Dimensions were commonly used but showed mixed psychometric properties.

Excessive reliance on proxy respondents and inconsistent reporting were identified as major issues.

Abstract

To (1) provide a comprehensive summary of the methods used to obtain health state utility values (HSUVs) from children and adolescents with disabilities (CAD), (2) describe the administration and psychometric properties of these methods in children and adolescents with disabilities, and (3) report summary statistics for HSUVs obtained from each method.

English-language studies from MEDLINE (via PubMed), PsychInfo, Scopus, CINAHL Plus, EconLit, and Embase were searched from inception to November 2024. Two reviewers independently screened titles, abstracts, and full texts. Studies were included if they used direct or indirect methods to measure HSUVs, reported utilities and/or psychometric properties of these measures, and involved CAD aged 0-19 years. Two reviewers independently extracted study details including sample descriptors, instruments used, and summary statistics. Studies quality was assessed using a novel tool derived from 3 validated checklists.

Of the 3541 screened articles, 31 met inclusion criteria. Only 2 studies used direct methods, such as time trade-off, visual analog scale, and standard gamble, whereas 29 employed generic measures (eg, EuroQol 5 Dimensions, Health Utilities Index 3) with diverse preference elicitation methods. Excessive dependence on proxy respondents was noted, and psychometric properties of generic measures were mixed.

Inconsistent HSUVs reporting and limited data availability are common. Reported HSUV summary statistics may be inaccurate if methodologies are unsuitable for the population. This review emphasizes the need for validated instruments to assess HSUVs in CAD.

Figures