Abstract
Background
The EQ-5D instruments have been widely used to measure health-related quality of life (HRQoL) in child and adolescent populations, especially the EQ-5D-Y-3L and EQ-5D-Y-5L (beta version). Although not specifically designed for younger users, the adult versions (EQ-5D-3L and EQ-5D-5L) are also used in these populations. While the measurement properties of these instruments have been evaluated in children and adolescents, no systematic review to date has employed a rigorous method to assess risk of bias. Additionally, quality criteria for good measurement properties and certainty of evidence have not been thoroughly evaluated. The aim of this study was to summarize and critically appraise the evidence on the measurement properties of all EQ-5D instruments in children and adolescents.
Methods
We conducted electronic searches on MEDLINE, EMBASE, CINAHL, EconLit, National Health Service Economic Evaluation Database (NHS-EED), and Health Technology Assessment (HTA) databases up to May 2024. We included studies measuring HRQoL using either the self-reported or proxy-reported version of the EQ-5D instruments—EQ-5D-3L, EQ-5D-5L, EQ-5D-Y-3L, and EQ-5D-Y-5L—using the descriptive system, visual analogue scale, and/or utility score in children and adolescents up to 19 years of age, and that tested at least one measurement property (e.g., reliability). The COnsensus-based Standards for the selection of health Measurement INstruments (COSMIN) methodology was followed to assess risk of bias, to score results for measurement properties, and to perform an evidence synthesis using a modified Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) approach.
Results
From 3586 records identified through the search, 65 studies were included in this systematic review. We found moderate certainty of evidence of sufficient comprehensibility and comprehensiveness of the EQ-5D-Y-3L and EQ-5D-Y-5L. Furthermore, we found very low certainty of evidence of inconsistent relevance for the EQ-5D-Y-3L, whereas the EQ-5D-Y-5L had sufficient relevance. Almost all the measurement properties (reliability, hypothesis testing for construct validity, and responsiveness) considering all the EQ-5D versions ranged from moderate certainty of evidence of insufficient results to very low certainty of evidence of insufficient results.
Conclusion
There is moderate certainty of evidence that the EQ-5D-Y-3L and EQ-5D-Y-5L have sufficient content validity. Both instruments can be recommended to measure HRQoL in children and adolescents aged 8–15 years. However, most of the measurement properties across all EQ-5D versions showed insufficient results, with certainty of evidence ranging from moderate to very low due to inconsistency and doubtful to inadequate risk of bias. Therefore, further research is needed to improve the methodological quality of studies on EQ-5D instruments for children and adolescents.
Systematic Review Registration
International prospective register of systematic reviews (PROSPERO): CRD42020218382 and Open Science Framework: https://osf.io/r8kt9/.
Supplementary Information
The online version contains supplementary material available at 10.1007/s40258-025-00953-0.
Key Points for Decision Makers
| The EQ-5D-Y- 3L and EQ-5D-Y-5L are commonly used worldwide to measure health-related quality of life in children and adolescents. |
| The EQ-5D-Y-3L and EQ-5D-Y-5L showed moderate certainty of evidence for content validity whereas the EQ-5D-Y-3L had higher certainty of evidence for other measurement properties compared with other versions. |
| The risk of bias of most of the studies of all EQ-5D versions in children and adolescents ranged from doubtful to inadequate quality. |
Introduction
There has been growing interest in the development of health-related quality-of-life (HRQoL) instruments in children and adolescents over the last decades [1, 2]. Although these instruments (i.e., generic and disease-specific) measure self-perceived health status, they can also provide the value of health status, known as utility values (score). These values are essential to facilitate the calculation of quality-adjusted life-years (QALYs) in economic evaluations, such as cost-utility analysis [3, 4]. The Child Health Utility 9-Dimension (CHU9D) [5], the Health Utilities Index Mark 2 and 3 (HUI2 and HUI3) [6], and the EQ-5D-Y-3L [7] are examples of generic measures, whereas the Paediatric Asthma Quality of Life Questionnaire (PAQLQ), the Escala de Calidade de Vida del Niño con Epilepsia (CAVE), and the Pediatric Asthma Health Outcome Measure (PAHOM) are examples of disease-specific measures used to obtain the utility score [8].
Among these instruments, the HUI3, CHU9D, and EQ-5D-Y-3L are the most widely used in studies assessing measurement properties in youth populations, largely due to their ability to provide comprehensive utility values for economic evaluations and their acceptance in health technology assessments worldwide [9, 10]. The EQ-5D family instruments are particularly preferred for economic evaluations, and have been recommended by national health technology assessment guidelines worldwide [11]. Furthermore, both the adult and child versions of the EQ-5D instruments have been extensively used as a patient-reported outcome measure (PROM) to evaluate HRQoL. Their measurement properties have been tested among children and adolescents across various countries [12–15].
Despite the existence of previous systematic reviews [9, 10, 16–19] summarizing the measurement properties of EQ-5D instruments in children and adolescents, no previous review has used a robust methodological approach to assess risk of bias. Additionally, quality criteria for good measurement properties and certainty of evidence evaluation have not been thoroughly addressed. Such assessment is needed to determine the reliability and validity of the results of the included studies [20]. Applying the quality criteria provides a broad perspective of the quality of the measurement properties [21].
While primary studies with good results may not necessarily be methodologically sound, combining risk of bias assessment with quality criteria for good measurement properties provides a comprehensive evaluation of the evidence [21]. Assessing certainty of evidence is fundamental for summarizing the available evidence, helping clinicians, researchers, and policymakers to choose the best instruments to ensure healthcare decisions are well informed, consistent, and based on the best available evidence [22]. In addition, content validity is particularly important for any instrument and must be assessed rigorously [23].
This review is innovative in considering all versions of the EQ-5D family instruments, incorporating the three ways of interpretation (e.g., descriptive system, visual analogue scale of the EQ-5D (EQ VAS) and utility score) and their modes of administration (e.g., self- and proxy-reported) is innovative. To date, no previous systematic reviews have considered the combination of these important methodological key aspects and there is still a gap in the literature in understanding the measurement properties of the EQ-5D instruments when considering the population of children and adolescents. Thus, this systematic review aimed to critically appraise and summarize the evidence on the measurement properties of the EQ-5D instruments (self-reported version and proxy versions) in children and adolescents.
Methods
Registration and Protocol
The protocol has been previously registered on the international prospective register of systematic reviews (PROSPERO: CRD42020218382) and published on the Open Science Framework: https://osf.io/r8kt9/. The protocol of this systematic review was published previously, and more details of the methodology can be found elsewhere [24]. This systematic review was reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA-COSMIN (COnsensus-based Standards for the selection of health Measurement Instruments)) of the Outcome Measurement Instrument 2024 [25, 26] (see Online Supplementary Material (OSM) 1). Also, we followed the methodology of the COSMIN initiative for systematic reviews of patient-reported outcome measures (PROMs) [22].
Instruments of Interest
The instruments of interest are the EQ-5D instruments used in a youth population, including adult and children’s versions: EQ-5D-3L, EQ-5D-5L, EQ-5D-Y-3L, and EQ-5D-Y-5L (beta version) [7, 27–29]. The adult versions (EQ-5D-3L and EQ-5D-5L) seem to be comparable to the youth versions as they cover the same health dimensions in the descriptive system. We evaluated the EQ-5D instruments in all available languages and different modes of administration [4]. The adult versions were also considered as recent studies have tested their measurement properties in children and adolescents, even when the children’s versions were already available in the literature [12, 30–37]. In addition, depending on the study population (i.e., both adults and adolescents or children and adolescents up to 18 years), researchers may choose to use the adult versions.
EQ-5D-3L and EQ-5D-5L
The EQ-5D-3L [38] was developed for the adult population and measures HRQoL in five dimensions: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. Each dimension has three levels of severity: ‘no problems,’ ‘some problems,’ and ‘a lot of problems.’ The EQ-5D-5L[28] was developed for the adult population with the aim of also improving the sensitivity of the instrument and to reduce ceiling effects when compared to the EQ-5D-3L. This instrument also measures HRQoL in the same five dimensions as the EQ-5D-3L; however, each dimension has five levels of severity: ‘no problems,’ ‘slight problems,’ ‘moderate problems,’ ‘severe problems,’ and ‘extreme problems.’ Both instruments can provide a health state of a descriptive system (e.g., 11111 – meaning no problems in all dimensions), which can be converted to an index value of the general population of a country, ranging from 0 ‘death’ or ‘worst health state’ to 1 ‘full health’ or ‘best health state’); however, negative values are also possible, which represent a health state ‘worse than dead.’ Both instruments also include a visual analogue scale (EQ VAS) ranging from 0 to 100, in which 0 represents ‘the worse health-perceived status’ and 100 represents ‘the best health-perceived status.’
EQ-5D-Y-3L and EQ-5D-Y-5L
The EQ-5D-Y-3L [7] and EQ-5D-Y-5L [29] were created based on the EQ-5D-3L and EQ-5D-5L, respectively. Both instruments have a self-reported version (answered by the children and adolescents) and the proxy versions (reported by parents, caregivers, or health professionals when children are unable to report their HRQoL). Both youth versions have been developed specifically for children and adolescents aged 8–15 years. The self-reported version is recommended for children and adolescents aged 8–15 years, while the proxy-reported versions are recommended for children aged 4–7 years [39]. Both versions (EQ-5D-Y-3L and EQ-5D-Y-5L) differ from the adult versions (EQ-5D-3L and EQ-5D-5L), mainly in terms of the language to allow better understanding for children and adolescents (i.e., the ‘mobility’ dimension is described as ‘walking about’) [40]. However, both versions have the same structure (i.e., dimensions and levels of severity of descriptive system and EQ VAS) of the original versions of the instrument. However, the health states of the EQ-5D-Y-3L’s descriptive system specifically can be converted to an index value of the general population for Japan, Slovenia, The Netherlands, Germany, China, Indonesia, Spain, Belgium, and Hungary, as these are the countries with valuation studies available [41–49]. It is important to note that the EQ-5D-Y-5L is still in process of development and the studies conducted to date have being using a beta version of the instrument, and so there is no value-set specifically developed for the EQ-5D-Y-5L [40].
Eligibility Criteria
Studies that used one or more EQ-5D instruments in children and adolescents up to 19 years of age (or at least 80% of the sample with a mean age up to 19 years). We decided to be inclusive and cover most studies in this population because the definition of adolescents can vary from one country to another.
Studies using the self-reported or the proxy versions of the instruments and testing at least one measurement property (as defined by the COSMIN taxonomy) in the study [50]. According to the COSMIN guideline, we provided a list of the measurement properties of interest, their definitions (taxonomy), and the quality criteria used for each measurement property (Table 1).
There were no restrictions on the type of setting or health conditions of the population or language, and we only included full-text articles published in peer-review journals.
Table 1.
List of the domains and measurement properties, definition, and quality criteria according to the COSMIN guideline
| Domain | Measurement property | Definition | Quality criteria |
|---|---|---|---|
| Validity | The degree to which a PROM measures the construct(s) it purports to measure | ||
| Content validity | The degree to which the content of a PROM is an adequate reflection of the construct to be measured | Not applicable | |
| Construct validity (hypothesis testing) | The degree to which the scores of a PROM are consistent with hypotheses (for instance with regard to internal relationships, relationships to scores of other instruments, or differences between relevant groups) based on the assumption that the PROM validly measures the construct to be measured |
(+) The result is in accordance with the hypothesis (?) No hypothesis defined (by the review team) (−) The result is not in accordance with the hypothesis |
|
| Criterion validity | The degree to which the scores of a PROM are an adequate reflection of a ‘gold standard’ |
(+) Correlation with gold standard ≥ 0.70 OR AUC ≥ 0.70 (?) Not all information for ‘+’ reported (−) Correlation with gold standard < 0.70 OR AUC < 0.70 |
|
| Reliability | The degree to which the measurement is free from measurement error | ||
| Reliability | The proportion of the total variance in the measurements which is due to ‘true’† differences between patients |
(+) ICC or weighted Kappa ≥ 0.70 (?) ICC or weighted Kappa not reported (−) ICC or weighted Kappa < 0.70 |
|
| Measurement error | The systematic and random error of a patient’s score that is not attributed to true changes in the construct to be measured |
(+) SDC or LoA < MIC (?) MIC not defined (−) SDC or LoA > MIC |
|
| Responsiveness | The ability of a PROM to detect change over time in the construct to be measured |
(+) The result is in accordance with the hypothesis7 OR AUC ≥ 0.70 (?) No hypothesis defined (by the review team) (−) The result is not in accordance with the hypothesis7 OR AUC < 0.70 |
|
| Responsiveness | |||
| Interpretabilitya | Interpretability is the degree to which one can assign qualitative meaning—that is, clinical or commonly understood connotations—to a PROM’s quantitative scores or change in scores | Not applicable |
Definitions and quality criteria are in accordance to the COSMIN guideline (Prinsen, 2018 Quality of Life Research)
COSMIN COnsensus‐based Standards for the selection of health Measurement Instruments, PROM patient-reported outcome measure, AUC area under the curve, ICC intraclass correlation coefficient, SDC standard error of measurement, LoA limits of agreement, MIC minimal important change
(+): sufficient; (−): insufficient; (?): indeterminate
aInterpretability is not considered a measurement property, but an important characteristic of a measurement instrument
Literature Search
The following databases were searched: MEDLINE, EMBASE, CINAHL, EconLit, National Health Service Economic Evaluation Database (NHS-EED), and Health Technology Assessment (HTA) up to May 2024. To retrieve other potentially eligible studies, we conducted manual searches on the EuroQol Research Foundation website [40] and we screened the reference lists of the included studies. Moreover, we also searched for studies meeting our inclusion criteria among the studies included in some previous systematic reviews on the EQ-5D [9, 10, 16–19].
Search Strategy
The search strategy of this systematic review was created based on three key elements, according to COSMIN recommendations [22]: (1) ‘children and adolescents’; (2) ‘EQ-5D’; and (3) a sensitive filter to identify studies on measurement properties [51]. The detailed search strategy can be found in OSM 2.
Study Selection
The results from the searches were exported to the EndNote Software version X9 for removing duplicates [52]. Two review authors (CMES and VSS) independently screened all titles and abstracts, and the full texts for potentially eligible studies. Any disagreement was discussed and, if consensus could not be reached, a third review author (TPY or GCM) made the final decision.
Data Extraction, Data Items, and Outcomes
Two authors (CMES and VSS) independently extracted the data for each eligible study. Disagreements between the reviewers were discussed and a third reviewer (TPY or GCM) arbitrated lack of consensus. If necessary, we contacted the authors from the included studies, which has been done three times via email over a 1-month period to ask for additional information not reported in the article. Two authors (TPY and GCM) with experience in conducting systematic reviews developed a data-extraction form [22]. The main outcome of interest was any measurement properties of the EQ-5D instruments (self-reported or proxy versions) used to measure HRQoL in children and adolescents. We extracted the following data:
EQ-5D version (e.g., EQ-5D-Y-3L, EQ-5D-Y-5L, EQ-5D-3L, EQ-5D-5L);
Scores used (e.g., descriptive system, EQ VAS, and/or utility score);
Reference, year;
Language, country;
Population characteristics (e.g., sample size, age range, mean age, disease);
Context of use (e.g., primary care);
Mode of administration (e.g., proxy-version 1) and delivery mode (e.g., paper and pencil);
Number and/or percentage of missing data and response rates;
Measurement properties of interest according to the COSMIN taxonomy [50];
Results of measurement properties (e.g., values of interclass correlation coefficient);
Feasibility and interpretability data (e.g., ceiling and floor effects by dimension);
Recall period reported.
We extracted data from primary studies that assessed reliability using different statistical methods (e.g., Gwet’s AC method). We also extracted data from studies that used alternative methods to generate a continuous value when testing measurement properties (such as the level-sum score and/or the adult value set in youth population), if the study included at least one of the three official EQ-5D interpretation methods: the descriptive system, EQ VAS, and/or utility score, as defined by the EQ-5D-Y-3L value set [39]. In addition, we did not use any special handling for different approaches to be consistent with our methodology and to facilitate the interpretation of our findings. Although there is no established ‘gold standard’ for measuring HRQoL, we considered the criterion validity in our review when studies correlated the generic measure of the EQ-5D instruments with a gold standard for a specific disease, thus categorizing it as criterion validity in this specific context.
Evaluation of Measurement Properties and Specific Assessment of Content Validity
The evaluation of measurement properties was performed following the three main components [20, 22, 23]: (1) risk of bias of individual studies; (2) quality criteria for (in)sufficient measurement properties; and (3) Grading of Recommendations Assessment, Development, and Evaluation (GRADE) assessment to rate the certainty of evidence for each measurement property. All three components were assessed by two authors (CMES and VSS). Disagreements between the reviewers were arbitrated by a third reviewer (TPY or GCM).
Risk of Bias of Individual Studies
The risk of bias assessment of the included studies was performed using the COSMIN Risk of Bias checklist for PROMs [20]. The measurement property called content validity was scored separately to the other measurement properties (i.e., reliability, hypothesis testing for construct validity, and responsiveness) [22]. To assess the content validity, we followed the COSMIN guidelines specifically for assessing the content validity of PROMs; more details can be found in our published protocol [24, 53]. Each standard within a COSMIN box was rated in a four-point rating system as ‘very good,’ ‘adequate,’ ‘doubtful,’ or ‘inadequate’ quality [20]. To determine an overall score for each measurement property, the lowest rating of any standard in the correspondent box was taken into consideration as ‘worse score counts’ [54]. For the hypothesis testing for construct validity (box 9) and responsiveness (box 10), we considered separately the overall rating in two and four parts, i.e.: box 9 – (a) the hypothesis testing for construct validity and (b) the known-group validity; and box 10 – (a) comparison to a gold standard, (b) comparison with other outcome measures instruments, (c) comparison between subgroups, and (d) before and after intervention. As there were studies that only assessed known-groups validity without evaluating hypothesis testing for construct validity compared to other instruments, for instance, we made the decision to present them as separate categories to ensure transparency for those studies that assessed the known-groups validity isolated in the risk of bias assessment. As an article could test two or more measurement properties (i.e., test-retest reliability [box 6] and hypothesis testing for construct validity [box 9]) but also consider different health conditions data from different countries, we considered two or more studies within one article.
Risk of Bias of the Generic Preference-Based Measures
Conventional psychometric tests of validity in the generic preference-based measures for use in economic evaluation could be considered inappropriate [55]. Thus, we complemented our methodological quality assessment using a checklist to describe the practicality, reliability, and validity for judging the appropriateness of the EQ-5D of the included studies, which incorporates economists’ notion of preferences. This checklist is composed of practicality (three items), reliability (four items), and validity, which is subdivided into three parts: description (three items), valuation (four items), and empirical (one item) [55]. If one study tested reliability, then we completed the four items of reliability on this checklist. More detailed information of this checklist can be found elsewhere [55]. We decided to complement our assessment of methodological quality using the checklist for generic preference-based measures. This decision was made to avoid critiques on the use of the COSMIN Risk of Bias Checklist for the EQ-5D instruments.
Quality Criteria for Measurement Properties
For each measurement property tested in each included study, a quality criterion was defined as: sufficient (+), insufficient (−), or indeterminate (?). We interpreted the quality criteria on the reliability measurement property as follows: sufficient (+) for intraclass correlation coefficient (ICC) or weighted Kappa ≥ 0.70; indeterminate (?) for ICC or weighted Kappa not reported; insufficient (−) for ICC or weighted Kappa < 0.70. More information of the quality criteria for other measurement properties can be found in Table 1 [22].
Evidence Synthesis
Similar to the GRADE approach for systematic reviews of intervention studies [56], the certainty of evidence was graded as ‘high,’ ‘moderate,’ ‘low,’ or ‘very low’ evidence. We separated the analysis of the instruments according to the measurement properties presented for each part of the instruments (i.e., descriptive system, EQ VAS, and utility score). A rating system was adopted specifically to summarize the certainty of evidence for content validity [23]. This rating system was divided into three steps according to the COSMIN user manual presented elsewhere [23].
Regarding the other measurement properties, we used the criteria for good measurement properties to determine if the results of the included studies were consistent or inconsistent [22, 57]. Studies were considered consistent if they showed at least 75% of the results with the same rating of quality criteria (i.e., sufficient [+], insufficient [−], or indeterminate [?]). If the studies displayed clinical and methodological homogeneity, the results of some measurement property (i.e., coefficient correlation, ICC) were pooled in a random-effects meta-analysis. The pooled estimates were assessed by calculating weighted correlation coefficients and 95% confidence intervals [22]. The results were shown by range (lowest and highest) of values or percentage of confirmed hypotheses (when construct validity) [22]. Inconsistency was assessed considering if less than 75% of the results had the same rating of quality criteria (i.e., sufficient [+], insufficient [−], or indeterminate [?]) and if there was a range of health conditions that we could not group. If the studies were inconsistent, we explained the reasons for inconsistency and summarized per subgroup. If no explanation was found, the overall rating was classified as inconsistent (±). If not enough information was available, the overall rating was classified as indeterminate (?) [22]
We formulated a priori hypotheses to assess hypothesis testing for construct validity and responsiveness, in line with the COSMIN guideline [22]. Our hypotheses were formulated for the EQ-5D instruments using the three modes of interpretation according to the User Guide (i.e., descriptive system, EQ VAS and utility score) [39], comparing them to other similar instruments (OSM 3): We have considered that similar constructs should correlate positively or have negative values ≥ 0.50; related, but dissimilar constructs should correlate positively or have negative values between 0.30 and 0.50; and unrelated constructs should correlate positively or have negative values < 0.30. For responsiveness, meaningful changes using the standardized effect size or response means between relevant (sub)groups were considered. In addition, the detailed spreadsheet subdivided into the descriptive system, EQ VAS, and utility score for reliability and measurement error, hypothesis testing for construct validity (with correlation values from primary studies compared to each comparator instrument) and responsiveness (with standardized effect size, response means, and other information) with regard to the review team’s predefined hypotheses, is available in OSM 4. For the other measurement properties (excluding content validity), the certainty of evidence downgrading was applied depending on the presence of four factors, in line with COSMIN guidance [22, 57].
Data Synthesis
We synthesized the final score of the risk of bias (i.e., very good, adequate, doubtful, and inadequate) according to all measurement properties for each version of the EQ-5D (e.g., EQ-5D-Y-3L self-reported) according to the scores used (e.g., descriptive system, EQ VAS, and/or utility score). As one article could evaluate two or more measurement properties (i.e., test-retest reliability and hypothesis testing for construct validity), health conditions, and even countries, we considered two or more ‘studies’ within one article. Then, the number of ‘study outputs’ was used to refer to ‘individual studies,’ while the total number of ‘articles assessed’ was used to refer to ‘articles.’ We pooled the reliability results from studies with similar characteristics, such as health condition, test-retest interval, instrument versions (i.e., EQ-5D-Y-3L), and mode of administration (i.e., self-reported), presenting the range of ICC along with the sample size and number of articles assessed. For hypothesis testing for construct validity, studies were pooled using the same criteria, and we evaluated whether their hypotheses aligned with our a priori hypotheses, as recommended by COSMIN. We presented the overall quality criteria for good measurement properties (i.e., sufficient [+], insufficient [−], inconsistent [±] or indeterminate [?]) with certainty of evidence (i.e., high, moderate, low, and very low). The results of evidence synthesis were also categorized by measurement property for each version of the instrument, according to the scores used by original study.
Results
Study Selection
We identified 3586 records in our electronic search. We screened 2766 titles and abstracts after duplicate removal; 164 articles were screened for full-text assessment of which 99 articles were excluded, and 65 articles [7, 12–15, 29–37, 58–108] were included in this review. Details of study selection are presented in Fig. 1. Among the 99 excluded articles, 12 met the inclusion criteria but were excluded as they were qualitative studies (n = 6), cross-cultural adaptation studies (n = 2), or those that tested other measurements (n = 4). We provide a table with the specific reasons for exclusion in OSM 5.
Fig. 1.
Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 flow diagram for the systematic review, which included searches of databases, registers, and other sources. HTA Health Technology Assessment, NHS-EED National Health Service Economic Evaluation Database, NA not applicable. *We reported the number of records identified from all databases, each database individually, and manual searches
Study Characteristics
Among the 65 articles included, ten examined the measurement properties of the EQ-5D-3L [13, 30–32, 34, 36, 69, 70, 77, 78], five evaluated the EQ-5D- 5L [12, 33, 35, 37, 106], 30 focused on the EQ-5D-Y-3L [7, 14, 59–68, 71–76, 79–81, 83, 87, 95, 97, 98, 100, 102, 103, 105], and three investigated the EQ-5D-Y-5L [29, 58, 96]. Additionally, 17 articles investigated both the EQ-5D-Y-3L and EQ-5D-Y-5L simultaneously [15, 82, 84–86, 88–94, 99, 101, 104, 107, 108]. Two articles provided information on PROM development and content validity (one article for the EQ-5D-Y-3L [7] and one for the EQ-5D-Y-5L [29]). The characteristics of the included studies and the measurement properties assessed for each study by the descriptive system, EQ VAS and utility score are presented in OSM 6 and 7, respectively. The characteristics and risk of bias assessment on development studies of the EQ-5D-Y-3L and EQ-5D-Y-5L instruments are presented in Table 2. The data extraction spreadsheets subdivided into the descriptive system, EQ VAS, and utility score for each measurement property across all included studies are available in OSM 4.
Table 2.
Characteristics and risk of bias assessment on the studies development of the EQ-5D-Y-3L and EQ-5D-Y-5L
| PROM | References | Country of development | Construct definition | Target population for that the PROM was developed | Intended context of use | Concept elicitation study | Cognitive interview study? | Overall development study quality (risk of bias) | |
|---|---|---|---|---|---|---|---|---|---|
| Quality (risk of bias) | Patients involved? | ||||||||
| EQ-5D-Y-3L | Wille, 2010 | Germany, Italy, Spain, and Sweden | “The concept of HRQoL focuses on the viewpoint of the individual and is factored into different areas of health assessment, such as the impact of disease on daily living and functioning or health care utilization” | Healthy and chronically ill children and adolescents aged 8–18 years old | “It has been utilized internationally in many settings, such as clinical trials and population surveys. Its dissemination is supported by its concise contents, its applicability to a wide range of health conditions and its twofold reporting format, both as a descriptive profile and as an index which can be interpreted as a preference value. Although the literature emphasizes the use of EQ-5D in economic evaluations, it can nevertheless also be used as a measure of health status in general (non-economic) health assessments.” | Doubtful | Yes | Yes | Doubtful |
| EQ-5D-Y-5L | Kreimeier, 2019 | Germany, Spain, Sweden, and United Kingdom | Not reported; assumed same of EQ-5D-Y-3L | Healthy and chronically ill children and adolescents aged 8–15 years old | Not reported; we assumed same of EQ-5D-Y-3L | Doubtful | Yes | Yes | Doubtful |
COSMIN risk of bias: very good, adequate, doubtful, or inadequate quality. Development studies of the EQ-5D-3L and EQ-5D-5L (adult versions) instruments are not presented in this table because they were already assessed in another review
PROM patient-reported outcome measure, HRQoL health-related quality of life
Risk of Bias of Individual Studies Using the COSMIN Approach
PROM Development and Content Validity
No studies on PROM’s development and content validity of the adult versions (EQ-5D-3L and EQ-5D-5L) were retrieved in children and adolescents. Details of the EQ-5D-3L and EQ-5D-5L development studies in adults can be found elsewhere [109]. Two articles [7, 29] included information regarding the development of the instruments and their content validity (i.e., comprehensiveness and comprehensibility) (see Table 2). The risk of bias for the design PROM, cognitive interview study, and development for both instruments was rated as ‘doubtful.’ The studies assessing content validity focused on domains of comprehensiveness and comprehensibility collecting data exclusively from target population. Details on the risk of bias assessment for the development and the content validity of the EQ-5D-Y-3L and EQ-5D-Y-5L are described in OSM 8 and 9.
Other Measurement Properties
Descriptive System
COSMIN risk of bias ratings for measurement properties using the descriptive system are presented in Fig. 2. Test-retest reliability was evaluated in 34 study outputs from 22 articles [12, 14, 15, 37, 62, 65, 66, 75, 76, 79, 81, 82, 84–86, 89, 90, 93, 95, 96, 104, 107], with final ratings ranging from ‘adequate’ to ‘inadequate.’ Measurement error was assessed in 23 study outputs from 14 articles [12, 14, 15, 62, 65, 66, 75, 79, 82, 84–86, 89, 90] also with ratings from ‘adequate’ to ‘inadequate.’ There were 64 study outputs in 36 articles assessed [13, 14, 30, 36, 58–60, 62–66, 71, 74, 75, 78, 79, 83–85, 87, 88, 91–99, 101, 103, 105–107] that evaluated the hypothesis testing for construct validity (convergent validity) with ratings ranging from ‘adequate’ to ‘inadequate.’ Five study outputs in three articles [84, 85, 98] assessed responsiveness (one comparing instruments and four pre- and post-interventions) using the descriptive system, and the final rating was rated as ‘inadequate.’
Fig. 2.
COSMIN (COnsensus-based Standards for the selection of health Measurement Instruments) risk of bias: review authors'judgements about the final rating of each measurement property presented as percentages across all included studies using the descriptive system. COSMIN risk of bias: very good, adequate, doubtful or inadequate quality. Each measurement property was presented with the total number of studies (n) assessed. We considered criterion validity when studies correlated the EQ-5D instruments with a recognized disease-specific gold standard
EQ VAS
COSMIN risk of bias assessment for the final rating of each measurement property using the EQ VAS is presented in Fig. 3. Overall, 22 study outputs in 18 articles [14, 35, 37, 65, 66, 71, 75, 76, 78, 79, 81, 86, 87, 90, 93, 95, 96, 107] were found to assess the test-retest reliability, and the final rating ranged from ‘adequate’ to ‘inadequate.’ There were 44 study outputs in 27 articles assessed [13, 14, 35, 36, 58–60, 64, 65, 69–71, 76, 78, 80, 81, 87, 88, 91, 92, 94, 96–98, 103, 105, 106] that evaluated the hypothesis testing for construct validity (convergent validity), and the final rating ranged from ‘adequate’ to ‘inadequate.’ Overall, nine study outputs in eight articles [31, 35, 69, 71, 81, 87, 98, 108] evaluated responsiveness (one comparing between subgroups, one comparing between instruments and seven pre- and post-interventions), and the final rating ranged from ‘adequate’ to ‘inadequate.’
Fig. 3.
COSMIN (COnsensus-based Standards for the selection of health Measurement Instruments) risk of bias: review authors'judgements about the final rating of each measurement property presented as percentages across all included studies for the EQ VAS. COSMIN risk of bias: very good, adequate, doubtful, or inadequate quality. EQ VAS the visual analogue scale of the EQ-5D. Each measurement property was presented with the total number of studies (n) assessed. We considered criterion validity when studies correlated the EQ-5D instruments with a recognized disease-specific gold standard
Utility Score
COSMIN risk of bias assessment for the final rating of each measurement property using the utility score is presented in Fig. 4. Ten studies [12, 35, 37, 71, 76, 78, 87, 102, 104, 107] evaluated test-retest reliability using the utility score with ratings ranging from ‘adequate’ to ‘inadequate.’ There were 36 study outputs in 20 articles assessed [12, 13, 30, 32, 35, 70, 71, 74, 76–78, 80, 81, 87, 97, 100–103, 106] that evaluated the hypothesis testing for construct validity (convergent validity), and the final rating ranged from ‘adequate’ to ‘inadequate.’ For responsiveness, one study [34] was found that compared to other instruments and four study outputs in three articles [15, 33, 87] compared between subgroups, and all these studies were rated as ‘inadequate.’ In addition, 11 study outputs in ten articles [15, 31–33, 35, 71, 81, 102, 104, 108] evaluated the responsiveness comparing pre- and post-intervention, and the final rating ranged from ‘adequate’ to ‘inadequate.’ Details of the risk of bias assessment of each study and version of EQ-5D on measurement properties using the descriptive system, EQ VAS, and utility score can be found in OSM 10, 11, and 12, respectively.
Fig. 4.
COSMIN (COnsensus-based Standards for the selection of health Measurement Instruments) risk of bias: review authors'judgments about the final rating of each measurement property presented as percentages across all included studies for the utility score. COSMIN risk of bias: very good, adequate, doubtful, or inadequate quality. Each measurement property was presented with the total number of studies (n) assessed
Risk of Bias of the Generic Preference-Based Measures
Overall, the EQ-5D instruments had an adequate feasibility with missing values lower than 5% and easy application in less than 5 min in children and adolescents. Also, all the instruments had a ceiling effect for dimensions of the descriptive system, as well the best health profile, i.e., ‘11111.’ However, the versions with five levels (EQ-5D-5L and EQ-5D-Y-5L) tended to have lower ceiling effects. Details of methodological quality assessment using the checklist of the generic preference-based measures, the results related to the interpretability of the studies (ceiling and floor effects), and the feasibility of the EQ-5D instruments are available in OSM 13.
Evidence Synthesis
EQ-5D-3L—Self-Reported Version
Content Validity
There is very low certainty of evidence of sufficient relevance and comprehensibility and very low certainty of evidence of insufficient comprehensiveness for the EQ-5D-3L (Table 3) according to previous findings [109]. Details of the development of the EQ-5D-3L and of the content validity of this instrument can be found elsewhere [109].
Table 3.
Certainty of evidence for the measurement properties of each adult EQ-5D version
| PROM | EQ-5D-3L Self-reported version |
EQ-5D-3L Proxy-reported version |
EQ-5D-5L Self-reported version |
EQ-5D-5L Proxy-reported version |
||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Overall rating | Certainty of evidence | Articles (n) | Overall rating | Certainty of evidence | Articles (n) | Overall rating | Certainty of evidence | Articles (n) | Overall rating | Certainty of evidence | Articles (n) | |
| +/−/? | High, moderate, low, very low | +/−/? | High, moderate, low, very low | +/−/? | High, moderate, low, very low | +/−/? | High, moderate, low, very low | |||||
| Descriptive system | ||||||||||||
| Content validity | ||||||||||||
| Relevance | + | Very low | 1 (592)d | + | Very low | 1 (77)d | ||||||
| Comprehensiveness | − | Very low | 1 (592)d | + | Very low | 1 (77)d | ||||||
| Comprehensibility | + | Very low | 1 (592)d | + | Very low | 1 (77)d | ||||||
| Reliability | − | Very lowʀʀʀ¥ |
1 (106) |
|||||||||
| Construct validity | − | Lowʀ¥ | 1 (163)b | − | Lowʀ¥ | 4 (1437)a,b | ± | Not rated | 1 (849)e | |||
| EQ VAS | ||||||||||||
| Reliability | − | Low±¥ | 1 (139)a,b | − | Very low±±≠¥ | 1 (84)b | + | Lowʀ¥ | 2 (505)a | |||
| Construct validity | − | Very lowʀʀ±±¥ | 2 (212)b | − | Very lowʀʀ¥ | 4 (769)b | − | Very lowʀʀʀ¥≠ | 1 (87) | |||
| Responsiveness | + | Very lowʀʀ¥ | 2 (299) | + | Very lowʀʀʀ¥≠≠ | 1 (21)a | ± | Not rated | 1 (849)e | |||
| Utility score | ||||||||||||
| Reliability | − | Moderate¥ | 1 (156)a,b | − | Low¥≠ | 1 (95)a,b | + | Lowʀ¥ | 3 (611) | |||
| Construct validity | − | Very lowʀʀ¥ | 4 (1287)a,b | − | Very lowʀʀ¥ | 3 (371)b | − | Very lowʀʀ¥ | 2 (314) | |||
| Responsiveness | + | Very lowʀʀ¥ | 3 (433)a | ± | Not rated | 2 (197)e | ± | Not rated | 1 (849)e | |||
Overall ratings: ‘+’ = sufficient; ‘-’ = insufficient; ‘?’ = indeterminate; ‘±’ = inconsistent. Certainty of evidence: high, moderate, low, very low. Reasons of downgrade: ‘ʀ’ downgrade due to the risk of bias; ‘±’ downgrade due to inconsistency; ‘¥’ downgrade due to the indirectness; ‘≠’ downgrade due to the imprecision. We carefully sum the participants who underwent a test and retest according to the studies’ description
PROM patient-reported outcome measure, n number of participants considered for each measurement property
aWe considered the smallest number in the sum of the total sample by measurement property when a study had a variation in the number of participants
bWhen an article presented two or more studies (subsamples) on a measurement property
cWe did not consider the same sample twice in the sum of the total sample by measurement property when a study had overlapping samples (same sample in different articles)
dContent validity of the EQ-5D-3L and EQ-5D-5L were based in a previous systematic review of the health-related quality of life of patients in low back pain (Chiarotto, 2018 Journal of Clinical Epidemiology)
eWe considered as inconsistent because half of the studies were rated as sufficient and the other one as insufficient with unexplained reasons
Other Measurement Properties
There is low certainty of evidence of insufficient hypothesis testing for construct validity for the self-reported version using the descriptive system (one article, n = 163). There is low certainty of evidence of insufficient reliability for the self-reported version of the EQ VAS (one article, n = 139). Regarding the hypothesis testing for construct validity, there is very low certainty of evidence of insufficient hypothesis testing for construct validity for the self-reported version of the EQ VAS (two articles, n = 212). For responsiveness, there is very low certainty of evidence for sufficient results (two articles, n = 299). There is moderate certainty of evidence of insufficient reliability for the self-reported version using the utility score (one article, n = 156). Regarding the hypothesis testing for construct validity, there is very low certainty of evidence of insufficient hypothesis testing for the construct validity (four articles, n = 1287). Furthermore, there is very low certainty of evidence of sufficient responsiveness (three articles, n = 433) (see Table 3).
EQ-5D-3L—Proxy-Reported Version
Other Measurement Properties
There is low certainty of evidence of insufficient hypothesis testing for construct validity for the proxy-reported version of the descriptive system (four articles, n = 1437). The summary results for a subgroup of children and adolescents with imperforate anus are available in OSM 15. There is very low certainty of evidence of insufficient reliability for the proxy-reported version using the EQ VAS (one article, n = 84). There is very low certainty of evidence of insufficient hypothesis testing for construct validity (four articles, n = 769) (Table 3). In addition, we have consistent findings for a subgroup of children with attention-deficit/hyperactivity disorder (n = 126) and with imperforate anus (n = 110). The summary results were of insufficient hypotheses testing for construct validity of the EQ-5D-3L proxy-reported compared to the Attention-Deficit/Hyperactivity Disorder Rating Scale-IV—Parent Version and Langemeijer Stool Questionnaire instruments, respectively (see Table 5). There is low certainty of evidence of insufficient reliability (one article, n = 95) and very low certainty of evidence of insufficient hypothesis testing for construct validity for the proxy-reported version (three articles, n = 371) using the utility score (Table 3).
Table 5.
Summarized results of studies on measurement properties using the EQ VAS
| PROM | Reference, year | Language, country | Condition | Reliability | Hypotheses testing of construct validity | ||||
|---|---|---|---|---|---|---|---|---|---|
| Sample size | Study quality | Result (rating) | Sample size | Study quality | Comparator instrument: result (rating) | ||||
|
EQ-5D-3L Proxy-reported |
Matza et al. [70] | English, United Kingdom | ADHD | 83 | Inadequate | ADHD-RS: Results in line with 3 hypo’s (3+) | |||
|
EQ-5D-3L Proxy-reported |
Matza et al. [70] | English, USA | ADHD | 43 | Inadequate | ADHD-RS: Results not in line with 3 hypo’s (3−) | |||
| Summary results (overall ratings) | n = 126 | ADHD-RS: 3+ and 3− overall (−) | |||||||
|
EQ-5D-3L Proxy-reported |
Stolk et al. [13] | Dutch, Netherlands | Imperforate anus (1–4 years) | 27 | Inadequate | LSQ- 7: Results not in line with 1 hypo (1−) | |||
|
EQ-5D-3L Proxy-reported |
Stolk et al. [13] | Dutch, Netherlands | Imperforate anus (5–10 years) | 47 | Inadequate | LSQ- 7: Results not in line with 1 hypo (1−) | |||
|
EQ-5D-3L Proxy-reported |
Stolk et al. [13] | Dutch, Netherlands | Imperforate anus (11–15 years) | 36 | Inadequate | LSQ- 7: Results in line with 1 hypo (1+) | |||
| Summary results (overall ratings) | n = 110 | LSQ- 7: 1+ and 2− overall (−) | |||||||
|
EQ-5D-Y-3L Self-reported |
Ravens-Sieberer et al. [14] | Italian, Italy | General population | 415 | Doubtful | ICC: 0.82 (+) | 415 | Inadequate | SRH: Results not in line with 1 hypo (1−) |
|
EQ-5D-Y-3L Self-reported |
Ravens-Sieberer et al. [14] | Spanish, Spain | General population | 973 | Doubtful | ICC: 0.83 (+) | 973 | Doubtful | KIDSCREEN- 27: Results in line with 2 hypo’s (2+) and not in line with 1 hypo (1−) |
| Inadequate |
SRH: Results not in line with 1 hypo (1−) Cantril-ladder: Results not in line with 1 hypo (1−) |
||||||||
|
EQ-5D-Y-3L Self-reported |
Ravens-Sieberer et al. [14] | German, Germany | General population | 756 | Doubtful | KIDSCREEN- 27: Results in line with 3 hypo’s (3+) | |||
| Inadequate |
SRH: Results in line with 1 hypo (1+) Cantril-ladder: Results in line with 1 hypo (1+) |
||||||||
|
EQ-5D-Y-3L Self-reported |
Ravens-Sieberer et al. [14] | Unclear, South Africa | General population | 258 | Doubtful | KIDSCREEN- 27: Results in line with 1 hypo (1+) and not in line with 2 hypo’s (2−) | |||
| Inadequate |
SRH: Results not in line with 1 hypo (1−) Cantril-ladder: Results not in line with 1 hypo (1−) |
||||||||
|
EQ-5D-Y-3L Self-reported |
Ravens-Sieberer et al. [14] | Sweden, Swedish | General population | 407 | Doubtful | KIDSCREEN- 27: Results in line with 3 hypo’s (3+) | |||
| Inadequate |
SRH: Results not in line with 1 hypo (1−) Cantril-ladder: Results in line with 1 hypo (1+) |
||||||||
|
EQ-5D-Y-3L Self-reported |
Gusi et al. [65] | Spanish, Spain | General population | 477 | Doubtful | KIDSCREEN- 27: Results in line with 3 hypo’s (3+) | |||
| Inadequate |
SRH: Results in line with 1 hypo (1+) Cantril-ladder: Results in line with 1 hypo (1+) |
||||||||
| Summary results (overall ratings) | 1.388 | 0.82 to 0.83 (+) | n = 2.871 |
KIDSCREEN- 27: 12+ and 3− overall (+) SRH: 2+ and 4− overall (−) Cantril-ladder: 3+ and 2− overall (−) |
|||||
|
EQ-5D-Y-3L Proxy-reported |
Gusi et al. [65] | Spanish, Spain | General population | 79 | Doubtful | ICC: 0.57 (−) | 442 | Inadequate |
KIDSCREEN- 27: Results not in line with 3 hypo’s (3−) SRH: Results not in line with 1 hypo (1−) Cantril-ladder: Results in line with 1 hypo (1+) |
|
EQ-5D-Y-3L Proxy-reported |
Gusi et al. [65] | Spanish, Spain | General population | 35 | Doubtful | ICC: 0.66 (–) | 266 | Inadequate |
KIDSCREEN- 27: Results in line with 1 hypo (1+) and not in line with 2 hypo’s (2−) SRH: Results not in line with 1 hypo (1−) Cantril-ladder: Results in line with 1 hypo (1+) |
| Summary results (overall ratings) | 114 | 0.57 to 0.66 (–) | n = 708 |
KIDSCREEN- 27: 1+ and 5− (−) SRH: 0+ and 2− (−) Cantril-ladder: 2+ and 0− overall (+) |
|||||
COSMIN risk of bias: very good, adequate, doubtful, or inadequate quality. Overall ratings: ‘+’ = sufficient; ‘-’ = insufficient; ‘?’ = indeterminate; ‘±’ = inconsistent. KIDSCREEN- 27: comparisons include the physical well-being, psychological well-being and KIDSCREEN- 10 index score
PROM(s) patient-reported outcome measure(s), EQ VAS visual analogue scale of the EQ-5D, n sample size, ADHD attention deficit/hyperactivity disorder, hypo(’s) hypothesis(es), ADHD-RS Attention-Deficit/Hyperactivity Disorder Rating Scale-IV—Parent Version, LSQ- 7 Langemeijer Stool Questionionaire, ICC intraclass correlation coefficient, SRH self-rated health item, Cantril-ladder Cantril’s Life-Satisfaction-Ladder
EQ-5D-5L—Self-Reported Version
Content Validity
There is very low certainty of evidence of sufficient relevance, comprehensibility, and comprehensiveness for the EQ-5D-5L (Table 3) according to previous findings [109]. Details of the development of the EQ-5D-5L and of the content validity of this instrument can be found elsewhere [109].
Other Measurement Properties
There is very low certainty of evidence of insufficient reliability for the self-reported version of the EQ-5D-5L using the descriptive system (one article, n = 106). There was no evidence for hypothesis testing for construct validity using the descriptive system as the known-groups validity was not considered in the quality criteria assessment, following the COSMIN recommendations [21, 22]. There is low certainty of evidence of sufficient reliability (two articles, n = 505), while very low certainty of evidence of sufficient responsiveness for the self-reported version of the EQ-5D- 5L using the EQ VAS (one article, n = 21). There is also very low certainty of evidence of insufficient hypothesis testing for construct validity using the EQ VAS (one article, n = 87). There is low certainty of evidence of sufficient reliability for the self-reported version of the EQ-5D-5L using the utility score (three articles, n = 611). There is also very low certainty of evidence of insufficient hypothesis testing for construct validity using the utility score (two articles, n = 314). We assumed the inconsistency responsiveness of the EQ-5D-5L using the utility score (two articles, n = 197) as there were only two articles in the assessment of the quality criteria, thus, we were not able to formulate an evidence synthesis of this measurement property due to the inconsistency across studies (Table 3).
EQ-5D-5L—Proxy-Reported Version
Other Measurement Properties
Also, we assumed the inconsistency for the EQ-5D- 5L proxy-reported version for the hypothesis testing for construct validity using the descriptive system, EQ VAS and utility score (one article, n = 849) due to the inconsistency across the studies (Table 3).
EQ-5D-Y-3L—Self-Reported Version
Content Validity
There is very low certainty of evidence of inconsistency relevance as we downgraded two levels for risk of bias and one level for inconsistency. Furthermore, there is moderate certainty of evidence of sufficient comprehensibility and comprehensiveness of the EQ-5D-Y-3L for children and adolescents, as we downgraded one level for risk of bias (Table 4).
Table 4.
Certainty of evidence for the measurement properties of each child EQ-5D version
| PROM | EQ-5D-Y-3L Self-reported version |
EQ-5D-Y-3L Proxy-reported version |
EQ-5D-Y-5L Self-reported version |
EQ-5D-Y-5L Proxy-reported version |
||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Overall rating | Certainty of evidence | Articles (n) | Overall rating | Certainty of evidence | Articles (n) | Overall rating | Certainty of evidence | Articles (n) | Overall rating | Certainty of evidence | Articles (n) | |
| +/−/? | high, moderate, low, very low | +/−/? | high, moderate, low, very low | +/−/? | high, moderate, low, very low | +/−/? | high, moderate, low, very low | |||||
| Descriptive system | ||||||||||||
| Content validity | ||||||||||||
| Relevance | ± | Very lowʀʀ± | 1 (103) | + | Lowʀʀ | 1 (120) | ||||||
| Comprehensiveness | + | Moderateʀ | 1 (103) | + | Moderateʀ | 1 (120) | ||||||
| Comprehensibility | + | Moderateʀ | 1 (103) | + | Moderateʀ | 1 (120) | ||||||
| Reliability | − | Moderateʀ | 13 (2738)b,c | − | Moderateʀ | 8 (1484)b | − | Lowʀ± | 7 (512)b,c | − | Low±± | 5 (252) |
| Criterion validity | − | Very lowʀʀʀ | 1 (118) | |||||||||
| Construct validity | − | Moderateʀ | 26 (16490)a,b,c | − | Lowʀʀ | 5 (1666)b | ± | Not rated |
10 (8380)d |
− | Moderateʀ | 2 (520) |
| Responsiveness | − | Very lowʀʀʀ≠ | 1 (83) | |||||||||
| EQ VAS | ||||||||||||
| Reliability | + | Moderateʀ | 9 (2512)b,c | − | Lowʀʀ | 5 (1309)b | − | Very lowʀ±± | 2 (945)b,c | − | Very lowʀ±± | 3 (339) |
| Criterion validity | − | Very lowʀʀʀ | 1 (118) | |||||||||
| Construct validity | − | Moderateʀ | 12 (5258)b,c | − | Lowʀʀ | 3 (1106)b | − | Moderateʀ | 5 (996)b | − | Very lowʀʀʀ | 1 (234) |
| Responsiveness | + | Lowʀʀ |
4 (361)b |
+ | Very lowʀ≠≠ | 1 (30) | + | Very lowʀ≠≠ | 1 (42) | |||
| Utility score | ||||||||||||
| Reliability | + | Moderateʀ |
4 (251) |
− | Lowʀ± | 2 (310)b | ||||||
| Construct validity | − | Low±± |
10 (3090)b |
− | Lowʀʀ | 1 (321)b | − | Lowʀʀ | 1 (298) | |||
| Responsiveness | + | Moderateʀ |
6 (571)b |
+ | Moderateʀ | 2 (128) | − | Very lowʀʀʀ | 1 (110) | |||
Overall ratings: ‘+’ = sufficient; ‘−’ = insufficient; ‘?’ = indeterminate; ‘±’ = inconsistent. Certainty of evidence: high, moderate, low, very low. Reasons of downgrade: ‘ʀ’ downgrade due to the risk of bias; ‘±’ downgrade due to inconsistency; ‘¥’ downgrade due to the indirectness; ‘≠’ downgrade due to the imprecision. We carefully sum the participants who underwent a test and retest according to the studies’ description. We considered as criterion validity when studies correlated the EQ-5D instruments with a recognized disease-specific gold standard
PROM patient-reported outcome measure, n number of participants considered for each measurement property
aWe considered the smallest number in the sum of the total sample by measurement property when a study had a variation in the number of participants
bWhen an article presented two or more studies (subsamples) on a measurement property
cWe did not consider the same sample twice in the sum of the total sample by measurement property when a study had overlapping samples (same sample in different articles)
dWe considered as inconsistent because half of the studies were rated as sufficient and the other one as insufficient with unexplained reasons
Other Measurement Properties
There is moderate certainty of evidence of insufficient reliability for the self-reported version of the EQ-5D-Y-3L using the descriptive system (13 articles, n = 2738). There is very low certainty of evidence of insufficient criterion validity (one article, n = 118) and moderate certainty of evidence of insufficient hypothesis testing for construct validity (26 articles, n = 16,490) (Table 4). The summary results of reliability and hypothesis testing for construct validity for a subgroup of children and adolescents from general population are available in OSM 14 and 15, respectively. Furthermore, there is very low certainty of evidence of insufficient responsiveness using the descriptive system (one article, n = 83).
There is moderate certainty of evidence of sufficient reliability for the self-reported version of the EQ-5D-Y-3L using the EQ VAS (nine articles, n = 2512). We combined two studies that used the EQ VAS and presented consistent findings for children and adolescents from the general population (n = 1388). The summary results of sufficient test-retest reliability of the EQ-5D-Y-3L self-reported through ICC ranged from 0.82 to 0.83 (see Table 5). There is very low certainty of evidence of insufficient criterion validity (one article, n = 118) and there is moderate certainty of evidence of insufficient hypothesis testing for construct validity (12 articles, n = 5258). We found consistent findings for the subgroup of children and adolescents from the general population (n = 2871) for sufficient hypothesis testing for construct validity of the EQ-5D-3L self-reported compared to KIDSCREEN- 27 as 80% of the results are in line with the hypotheses. However, we found consistent findings for insufficient hypotheses testing for construct validity when compared to Self-rated Health Item and Cantril’s Life-Satisfaction-Ladder instruments. In addition, there is low certainty of evidence of sufficient responsiveness (four articles, n = 361) (Table 4).
There is moderate certainty of evidence of sufficient reliability for the self-reported version of the EQ-5D-Y-3L using the utility score (four articles, n = 251). There is low certainty of evidence of insufficient hypothesis testing for construct validity (ten articles, n = 3090) and moderate certainty of evidence of sufficient responsiveness (six articles, n = 571) (Table 4).
EQ-5D-Y-3L—Proxy-Reported Version
Other Measurement Properties
There is moderate certainty of evidence of insufficient reliability (eight articles, n = 1484) and low certainty of evidence of insufficient hypothesis testing for construct validity (five articles, n = 1666) for the proxy-reported version of the EQ-5D-Y-3L using the descriptive system (Table 4). The summary results of reliability and hypothesis testing for construct validity for a subgroup of children and adolescents from the general population are available in OSM 14 and 15, respectively. Furthermore, there is no evidence for responsiveness because we found two articles in which the authors used a percentage of changes using the descriptive system in the analysis, and this was not considered in the quality criteria assessment.
There is low certainty of evidence of insufficient reliability for the proxy-reported version of the EQ-5D-Y-3L using the EQ VAS (five articles, n = 1309) (see Table 4). We combined two studies that used the EQ VAS and presented consistent findings for parents (mothers and fathers) of children and adolescents from the general population (n = 114). The summary results were of insufficient test-retest reliability of the EQ-5D-Y-3L proxy reported through ICC ranging from 0.57 to 0.66 (see Table 5). There is low certainty of evidence of insufficient hypothesis testing for construct validity using the EQ VAS (three articles, n = 1106). We found consistent findings for the subgroup of children and adolescents from the general population (n = 708) for insufficient hypothesis testing for the construct validity of the EQ-5D-3L proxy-reported compared to KIDSCREEN- 27 and Self-rated Health Item as 83 and 100% of the results are not in line with the hypotheses. However, we found consistent findings for sufficient hypothesis testing for construct validity when compared to Cantril’s Life-Satisfaction-Ladder instrument as 100% of the results are in line with the hypotheses (see Table 5).
There is low certainty of evidence of insufficient reliability (two articles, n = 310) and hypothesis testing for construct validity (one article, n = 321) for the proxy-reported version of the EQ-5D-Y-3L using the utility score (Table 4). Furthermore, there is moderate certainty of evidence of sufficient responsiveness (two articles, n = 128).
EQ-5D-Y-5L—Self-Reported Version
Content Validity
There is low certainty of evidence of sufficient relevance, as we downgraded two levels for risk of bias. Furthermore, there is moderate certainty of evidence of sufficient comprehensibility and comprehensiveness of the EQ-5D-Y-5L for children and adolescents, as we downgraded one level for the risk of bias (Table 4).
Other Measurement Properties
There is low certainty of evidence of insufficient reliability for the self-reported version of the EQ-5D-Y-5L using the descriptive system (seven articles, n = 512). For the hypothesis testing for construct validity using the descriptive system (ten articles, n = 8380), we assumed the inconsistency for the EQ-5D-Y-5L self-reported version due to the inconsistency across the studies (Table 4).
There is very low certainty of evidence of insufficient reliability (two articles, n = 945) and sufficient responsiveness (one article, n = 30) for the self-reported version of the EQ-5D-Y-5L using the EQ VAS. Also, there is moderate certainty of evidence of insufficient hypothesis testing (five articles, n = 996) for construct validity using the EQ VAS (Table 4).
There is low certainty of evidence of insufficient hypothesis testing for construct validity (one article, n = 298) and very low certainty of insufficient responsiveness (one article, n = 110) for the self-reported version of the EQ-5D-Y-5L using the utility score (Table 4).
EQ-5D-Y-5L—Proxy-Reported Version
Other Measurement Properties
There is low certainty of evidence of sufficient reliability for the proxy-reported version of the EQ-5D-Y-5L (five articles, n = 252). There is moderate certainty of evidence of insufficient hypothesis testing for construct validity (two articles, n = 520) using the descriptive system (Table 4).
There is very low certainty of evidence of insufficient reliability (three articles, n = 339) and hypothesis testing for construct validity (one article, n = 234) for the proxy-reported version of the EQ-5D-Y-5L using the EQ VAS. Furthermore, there is very low certainty of evidence of sufficient responsiveness (one article, n = 42) for the proxy-reported version of the EQ-5D-Y-5L using the EQ VAS (Table 4).
Value Set Used
Many studies used inadequate methods to estimate the value set for testing the measurement properties of the EQ-5D instrument in children and adolescents. Out of 65 studies, 29 studies (44.6%) estimated the utility score inadequately (e.g., summary score, value set from adults’ version (EQ-5D-3L), and other methods) [12, 13, 15, 30, 32–36, 62, 63, 70, 71, 74, 76–78, 80–83, 87, 99–101, 103, 104, 106, 107], and among these 29, four studies used the children's value set from other countries [83, 103, 104, 107]. Only two studies (3.1%) appropriately used the value set for children and adolescents available in their countries [97, 108]. Furthermore, it is important to note that 12 studies (18.5%) applied the level sum score of the descriptive system to test the measurement properties [88, 91–94, 98, 99, 102, 104, 105, 107, 108].
Discussion
Principal Findings
The descriptive system of the EQ-5D-Y-3L and EQ-5D-Y-5L had moderate certainty of evidence of sufficient comprehensibility and comprehensiveness. The descriptive system of the EQ-5D-Y-5L had low certainty of evidence of sufficient relevance, whereas the descriptive system of the EQ-5D-Y-3L had very low certainty of evidence of inconsistent relevance. The self-reported version of the EQ-5D-Y-3L showed better measurement properties considering the certainty of evidence and overall quality criteria for reliability and hypothesis testing for construct validity for the three ways of interpretation (descriptive system, EQ VAS, and utility score) compared to other versions. It is important to highlight that almost all the remaining measurement properties (reliability, criterion validity, hypothesis testing for construct validity (convergent validity), and responsiveness) when considering all the other EQ-5D versions have moderate certainty of evidence of insufficient results to very low certainty of evidence of insufficient results. In addition, despite high ceiling effects, the EQ-5D instruments had adequate feasibility to use in studies on measurement properties in children and adolescents.
Strengths and Weaknesses of the Study
In this systematic review, we used the certainty of evidence, the overall quality criteria for good measurement properties and the risk of bias assessment for each measurement property using the COSMIN methodology [22]. We also included all the instruments of the EQ-5D family used in children and adolescents (EQ-5D-3L, EQ-5D- 5L, EQ-5D-Y-3L and EQ-5D-Y-5L) including self- and proxy-reported versions and assessing the descriptive system, EQ VAS, and utility score [7, 27–29]. We believe this approach provides an overview beyond just the confidence using the GRADE, encompassing the quality of the measurement properties of the versions of the EQ-5D in children and adolescents. Moreover, we conducted a rigorous risk of bias assessment following the COSMIN guidelines, which has not been conducted in previous studies, or when conducted, authors have adapted the COSMIN guideline to fit their assessment of the EQ-5D [9, 10, 16–18].
Despite these strengths, we recognize that the COSMIN methodology was not fully developed for preference-based measures like the EQ-5D. Nevertheless, we used such methodology as the EQ-5D instruments are regularly used as PROMs in research and clinical practice [4, 110]. Also, we believe COSMIN is a robust methodology frequently used and recommend in the literature [20, 22, 23]. Adaptation and validation of the COSMIN methodology for generic preference-based measures may be considered by methodologists in the future. Unlike PROMs, there is not a systematic evaluation approach beyond COSMIN specifically for this type of instrument to assess the risk of bias, quality criteria for good measurement properties and certainty of evidence. We also understand that the cutoff of 0.70 for reliability might be high according to the quality criteria determined in the COSMIN guideline for generic preference-based measures as EQ-5D [21]. This threshold had been adopted as a criterion by the International Society for Quality of Life Research (ISOQOL) [111], previous systematic reviews [9, 10, 18, 19] and primary studies of the EQ-5D [14, 37, 79, 90, 95], nevertheless the interpretation of Landis and Koch guidelines can be an alternative for interpretation of test-retest reliability [112]. Furthermore, it is also important to highlight that the continuous values of the EQ-5D (i.e., EQ VAS and utility score) tend to frequently meet this cutoff in previous studies [14, 65, 71, 74, 87, 97]. Another possible limitation to be considered is that we did not assess the measurement properties of the other generic preference-based measures commonly used for the population of children and adolescents (e.g., HUI3, Quality of Well-Being Scale, the CHU9D, and the Adolescent Health Utility Measure) [2] to provide a comparison of all instruments. The COSMIN methodology could not provide certainty of evidence for generic preference-based measures because not all aspects of a PROM are relevant for them (i.e., internal consistency) and vice versa (distributional properties, informativity (Shannon’ indices) and inconsistency) [113, 114]. Although we provided a comprehensive data extraction spreadsheet for each measurement property, especially for hypothesis testing for construct validity and responsiveness, the use of generic hypotheses might be considered a limitation.
Since our review focused on generic instruments such as the EQ-5D and included children and adolescents with several health conditions, tailoring specific hypotheses for each health condition and comparator instruments would have not been feasible. Also, it would not be feasible as a team to have such expertise in numerous health conditions and instruments to establish specific cut-off points for responsiveness, given the heterogeneity of these conditions. We believe that a specific hypothesis approach would be more feasible in systematic reviews of a certain health condition with a more thorough knowledge of the most used instruments at hand. This would consequently lead to more precise hypotheses for construct validity and responsiveness.
Strengths and Weaknesses in Relation to Other Studies
Despite the existence of several previous systematic reviews that summarized the measurement properties of the EQ-5D instruments in children and adolescents, there are important flaws in the literature that have not been addressed by previous studies in the area [9, 10, 16–19]. A previous systematic review assessed the measurement properties of the EQ-5D instruments in children and adolescents [16]. The authors identified poor reporting of the measurement properties in the included studies and emphasized the need for conducting risk of bias assessments to improve further knowledge on EQ-5D in children and adolescents [16]. In our study, assessing the risk of bias using the COSMIN Risk of Bias Checklist [20], we found doubtful and inadequate quality for most measurement properties. On the other hand, another systematic review evaluated the measurement properties of the EQ-5D-Y-3L and reported high methodological quality among the included studies with 11.8 points on a scale of 14 points [18]. However, they used a modified tool designed for randomized and non-randomized studies of healthcare interventions to assess the risk of bias in the included studies which did not consider key aspects of measurement properties [18]. The assessment of risk of bias for example is a crucial factor to determine the reliability and validity of the results of the included studies and has not been considered in previous reviews [9, 10].
There are two systematic reviews that evaluated the measurement properties of the generic childhood preference-based measures (including the EQ-5D child versions) in children and adolescents [9, 10]. Although these systematic reviews provided the overview of the ratings for good measurement properties, they did not conduct a rigorous method to assess the risk of bias of the included studies and they also did not provide the certainty of evidence for each measurement property. The method used for applying the quality criteria for each measurement property is another example, as depending on the method used, we are uncertainty whether primary studies with good results could have been conducted appropriately or not. To address this issue, it is recommended that the assessment of risk of bias and the quality criteria for good measurement properties is conducted using the COSMIN Risk of Bias Checklist and the quality criteria of the COSMIN [20–22]. The assessment of the certainty of evidence is probably another flaw identified in the literature, and this is a fundamental characteristic of systematic reviews helping clinicians, researchers, and policymakers to choose the best instruments to ensure that healthcare decisions are well-informed, consistent, and based on the best available evidence.
Another systematic review aimed to summarize the evidence on the measurement properties of the EQ-5D-Y-3L and EQ-5D-Y-5L [18]. The authors found an ICC ranging from ‘excellent’ to ‘good’ for the EQ VAS of the EQ-5D-Y-3L for most subsamples and an inconsistent ICC ranging from ‘fair’ to ‘excellent’ considering the utility score of the EQ-5D-Y-3L [18]. On the other hand, our review found that there was moderate certainty of evidence of sufficient reliability and very low certainty of evidence of sufficient reliability of the self-reported version of the EQ-5D-Y-3L using the EQ VAS and utility score, respectively. The authors from the previous study also found inconsistent results for the reliability of the EQ-5D-Y-3L [18] and their findings are like our findings on moderate certainty of evidence of insufficient reliability for the self-reported version of the EQ-5D-Y-3L using the descriptive system. We combined some studies that had consistent findings according to Table 5 and OSM 14 and 15. The pooled data of the descriptive system results and the body of evidence of the EQ-5D-Y-3L favored this classification as moderate, but the certainty of evidence downgrade was done due to poor methodological quality. We believe though, that over the time, the certainty of evidence tends to improve with better studies being published. The authors found an adequate feasibility (proportion of missing values < 5%) for both the instruments and a ceiling effect (best health profile ‘11111’) was of 55% for the EQ-5D-Y-3L and 20% for the EQ-5D-Y-5L [18], but in this analysis with high heterogeneity [18]. We also collected feasibility and ceiling and floor effects but did not combine these results with the certainty of evidence. We believe it is important to include ceiling effects and missing data rates as it can complement our findings. According to COSMIN, feasibility and interpretability are not measurement properties as they do not refer to the quality of an instrument. However, both measures have important aspects in the selection process of an instrument, including the EQ-5D. Furthermore, we provided the details for these measures in OSM 13 and without specifying them in title and objectives, as they were not the main focus.
There are two systematic reviews that used the COSMIN methodology to evaluate generic childhood preference-based measures as EQ-5D in children and adolescents [17, 19]. However, one review from 2015 could not provide the certainty of evidence for each measurement property as the COSMIN methodology was still emerging at that time [17, 22]. The review from 2023 followed the COSMIN methodology (including risk of bias, quality criteria and, certainty of evidence), but they did not present the risk of bias assessment, the content validity according to the COSMIN. Furthermore, they included only studies using the descriptive system (excluding 15 studies) [19]. The assessment of the content validity (the most important measurement property of an instrument) is another problem identified in the literature. Even though an instrument demonstrates good measurement properties for reliability for example, it might not be sufficient to recommend its use if it fails to adequately assess the intended construct. Although there is no gold standard for measuring HRQoL, our systematic review included one study for the criterion validity as the authors correlated the EQ-5D-Y-3L with the estimated glomerular filtration rate, which is considered the gold standard for chronic kidney disease [66]. There is an absence in the literature of an understanding of how the measurement properties of all EQ-5D family instruments work for children and adolescents, considering their three ways of interpretation (e.g., descriptive system, EQ VAS, and utility score) and the different modes of delivery (e.g., self- and proxy-reported). No previous systematic reviews have considered those important methodological key aspects together. Finally, our systematic review aimed to address a significant research gap in the literature considering key methodological aspects including the assessment of the risk of bias, applied criteria for good measurement properties and certainty of evidence for each measurement property, as well the content validity of the EQ-5D instruments. This could possibly help clinicians, researchers, and policymakers to make better decisions on what to use among the EQ-5D family instruments and their versions (e.g., self-reported) in children and adolescents.
Implications for Clinicians and Policy Makers
The certainty of evidence of the EQ-5D-3L and EQ-5D- 5L has been assessed in a previous systematic review in adults with low back pain [109], and in this review, we have considered it as indirect evidence according to the COSMIN manual [53]. Besides the fact that there is evidence in the literature for the use of the EQ-5D adult versions in the children and adolescent populations, we believe those instruments can all be comparable as they cover the same health dimensions in the descriptive system. On the other hand, the EQ-5D-Y-3L and EQ-5D-Y-5L are specifically designed to be more age-appropriate for children and adolescents [110, 115]. According to our findings on content validity, the EQ-5D-Y-3L and EQ-5D-Y-5L should be recommended to measure HRQoL in children and adolescents aged 8–15 years. The EQ-5D-Y-3L showed moderate certainty of evidence for most measurement properties (i.e., reliability, hypothesis testing for construct validity, and responsiveness) in all its three ways of interpretation. Although some results indicated insufficient measurement properties, the EQ-5D-Y-3L demonstrated better results compared to other versions. The current evidence demonstrates that we are moderately confident regarding the other measurement properties, and that is why we recommend the EQ-5D-Y-3L to measure HRQoL in children and adolescents. Although there was moderate to very low certainty of evidence for most other measurement properties (reliability, hypothesis testing for construct validity, and responsiveness) for all the instruments evaluated in this review considering the descriptive system, EQ VAS, and utility score, we have moderately to very low confidence of these measurement properties. We believe that new studies can change the estimate of the measurement properties of the EQ-5D in children and adolescents [57]. In the risk of bias assessment, there was inadequate and doubtful quality for most measurement properties of the included studies (particularly due to the absence of a priori hypotheses), and these were the main reasons for downgrading the certainty of evidence on the use of the EQ-5D in children and adolescents. Therefore, the data extraction spreadsheet for measurement properties (available in OSM 4) could serve as the base for researchers to formulate their hypotheses before conducting a study on measurement properties, specifically when evaluating reliability, hypothesis testing for construct validity, and responsiveness.
The COSMIN Risk of Bias Checklist may not be designed to assess measurement properties in generic preference-based measures, such as the EQ-5D; however, we can assume that these measurement properties (i.e., test-retest reliability, hypothesis testing for construct validity, and responsiveness) are important for PROMs as well as for generic preference-based measures. Thus, researchers should consider using the items of the COSMIN Risk of Bias Checklist to improve the quality of the studies of measurement properties even on EQ-5D instruments, including in children and adolescents [20]. Furthermore, a checklist with relevant standards for studies of measurement properties, specific to generic preference-based measures, could be an option to be developed to facilitate well-conducted studies using similar methods and analysis to enable future comparisons [114].
Unanswered Questions and Future Research
Although there were studies that assessed the measurement properties of the EQ-5D-Y-3L using the utility score, only two studies (3.1%) correctly used the EQ-5D-Y-3L value set to estimate utility scores from the Chinese value set for children and adolescents according to the International Valuation Protocol of the EQ-5D-Y-3L [97, 108, 116]. Thus, future studies should assess the measurement properties of the EQ-5D-Y-3L and EQ-5D-Y-5L with the utility score adequately using the EQ-5D-Y-3L value set studies available [41–49]. As recommended by the EuroQol foundation, authors should not use alternative methods to generate continuous values as the level-sum score, summary score, or value set from other countries and adult versions (i.e., EQ-5D-3L or EQ-5D- 5L) [39, 110]. Future studies should improve reporting and methodological quality, especially using available tools such as the COSMIN Reporting Guideline and the COSMIN Risk of Bias Checklist [20, 117]. The development of a new checklist specifically designed to evaluate all measurement properties of generic preference-based measures would be beneficial for the scientific community. The new checklist could consider the criteria from existing tools for both PROMs and generic preference-based measures (i.e., content validity, test-retest reliability, hypothesis testing for construct validity, and responsiveness). This new checklist would provide a rating score for researchers to apply in those primary studies and/or to be used in systematic reviews in the field. Additionally, future studies should also consider the instruments that are currently being developed by the EuroQol Group, such as the EQ Health and Wellbeing instrument (EQ-HWB) and the EQ Toddler and Infant Populations (EQ-TIPS) [40]. The same recommendations on reporting and methodological quality should be addressed to ensure robust methods and, consequently, consistent findings. Future studies should also consider investigating the responsiveness of the EQ-5D-Y-3L and EQ-5D-Y-5L in children and adolescents with different conditions (i.e., diabetes, cerebral palsy, chronic pain). Further, it is important to consider head-to-head comparisons between the EQ-5D-Y-3L and EQ-5D-Y-5L to verify whether having five levels of severity is superior in the youth population as it is in the adult population (EQ-5D-3L and EQ-5D- 5L). Despite new studies of the EQ-5D-Y-3L and EQ-5D-Y-5L having been conducted in Africa (n = 12, 18.5%) and Oceania (n = 7, 10.8%), future studies should also consider countries on these continents, especially in North (n = 4, 6.2%), Central, and South America as well as low and middle-income countries, since most of the current included studies are from Europe (n = 22, 34%) and Asia (n = 20, 31%). Future systematic reviews of measurement properties (content validity, reliability, hypothesis testing for construct validity, and responsiveness) using the COSMIN methodology could also provide evidence synthesis of other generic preference-based measures (e.g., CHU9D, HUI2, HUI3) for the population of children and adolescents.
Conclusion
The findings of this systematic review support the use of the EQ-5D-Y-3L and EQ-5D-Y-5L instruments to measure HRQoL in children and adolescents due to sufficient results for content validity, especially the EQ-5D-Y-3L, which has better results across all measurement properties when compared to other versions. However, the certainty of evidence was from very low to moderate on the measurement properties across all versions of the EQ-5D instruments, where most results on reliability, hypothesis testing for construct validity, and responsiveness were insufficient and heterogeneous, largely due to inconsistency and risk of bias doubtful to inadequate in the downgrading for certainty of evidence. Therefore, further research is necessary to improve the methodological quality of the evidence and, consequently, to provide more trustworthy findings on the use of the EQ-5D for children and adolescents.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
Caique de Melo do Espirito Santo has received a PhD scholarship from the Sao Paulo Research Foundation (FAPESP) – process number 2021/08776 - 4; Verônica Souza Santos has received a PhD scholarship from the Sao Paulo Research Foundation (FAPESP) –process number 2019/12049 - 0; Dr. Tiê Parma Yamato has received a research grant from the Sao Paulo Research Foundation (FAPESP) – process number 2017/17484 - 1. The funding agency of this study had no influence in the design of the study and collection, analysis, and interpretation of data and writing of the manuscript.
Funding
Open Access funding enabled and organized by CAUL and its Member Institutions.
Declarations
Funding
The study has not received external funding from any funding agency in the public, commercial, or not-for-profit sectors.
Conflict of interest
The authors have no conflicts of interest to declare.
Data availability statements
All data included in this study are freely available upon request to the study authors.
Ethics approval
No ethics approval is required for this study.
Consent to participate
Not applicable.
Consent for publication
Not applicable.
Code availability
Not applicable.
Author contributions
CMES, VSS, GCM, and TPY developed the study design. CMES and VSS extracted data. CMES, VSS, GCM, and TPY analyzed and interpreted data for the work. CMES, VSS, AC, GCM, and TPY drafted the work or reviewed the content of the work. All authors read and approved the final version of the manuscript to be published.
References
- 1.Solans M, Pane S, Estrada MD, Serra-Sutton V, Berra S, Herdman M, et al. Health-related quality of life measurement in children and adolescents: a systematic review of generic and disease-specific instruments. Value Health. 2008;11:742–64. 10.1111/j.1524-4733.2007.00293.x. [DOI] [PubMed] [Google Scholar]
- 2.Chen G, Ratcliffe J. A review of the development and application of generic multi-attribute utility instruments for paediatric populations. Pharmacoeconomics. 2015;33:1013–28. [DOI] [PubMed] [Google Scholar]
- 3.Karimi M, Brazier J. Health, health-related quality of life, and quality of life: what is the difference? Pharmacoeconomics. 2016;34:645–9. [DOI] [PubMed] [Google Scholar]
- 4.Devlin NJ, Brooks R. EQ-5D and the EuroQol Group: past, present and future. Appl Health Econ Health Policy. 2017;15:127–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Stevens K. Developing a descriptive system for a new preference-based measure of health-related quality of life for children. Qual Life Res. 2009;18:1105–13. [DOI] [PubMed] [Google Scholar]
- 6.Feeny D, Furlong W, Torrance GW, Goldsmith CH, Zhu Z, DePauw S, et al. Multiattribute and single-attribute utility functions for the Health Utilities Index Mark 3 system. Med Care. 2002;40:113–28. [DOI] [PubMed] [Google Scholar]
- 7.Wille N, Badia X, Bonsel G, Burström K, Cavrini G, Devlin N, et al. Development of the EQ-5D-Y: a child-friendly version of the EQ-5D. Qual Life Res. 2010;19:875–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Thorrington D, Eames K. Measuring health utilities in children and adolescents: a systematic review of the literature. PLoS ONE. 2015;10:1–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Rowen D, Keetharuth AD, Poku E, Wong R, Pennington B, Wailoo A. A review of the psychometric performance of selected child and adolescent preference-based measures used to produce utilities for child and adolescent health. Value Health. 2021;24:443–60. 10.1016/j.jval.2020.09.012. [DOI] [PubMed] [Google Scholar]
- 10.Kwon J, Smith S, Raghunandan R, Howell M, Martin E, Kim S, et al. Systematic review of the psychometric performance of generic childhood multi-attribute utility instruments. Appl Health Econ Health Policy. 2023;21:559–84. 10.1007/s40258-023-00836-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kennedy-Martin M, Slaap B, Herdman M, van Reenen M, Kennedy-Martin T, Greiner W, et al. Which multi-attribute utility instruments are recommended for use in cost-utility analysis? A review of national health technology assessment (HTA) guidelines. Eur J Health Econ. 2020;21:1245–57. 10.1007/s10198-020-01195-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Cheung PWH, Wong CKH, Samartzis D, Luk KDK, Lam CLK, Cheung KMC, et al. Psychometric validation of the EuroQoL 5-dimension 5-level (EQ-5D-5L) in Chinese patients with adolescent idiopathic scoliosis. Scoliosis Spinal Disord. 2016. 10.1186/s13013-016-0083-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Stolk EA, Busschbach JJ, Vogels T. Performance of the EuroQol in children with imperforate anus. Quality Life Res. 2000;9:29–38. 10.1023/A:1008923029886. [DOI] [PubMed] [Google Scholar]
- 14.Ravens-Sieberer U, Wille N, Badia X, Bonsel G, Burström K, Cavrini G, et al. Feasibility, reliability, and validity of the EQ-5D-Y: results from a multinational study. Qual Life Res. 2010;19:887–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wong CKH, Cheung PWH, Luo N, Lin J, Cheung JPY. Responsiveness of EQ-5D Youth version 5-level (EQ-5D-5L-Y) and 3-level (EQ-5D-3L-Y) in patients with idiopathic scoliosis. Spine (Phila Pa 1976). 2019;44:1507–14. [DOI] [PubMed] [Google Scholar]
- 16.Noyes J, Edwards RT. EQ-5D for the assessment of health-related quality of life and resource allocation in children: a systematic methodological review. Value Health. 2011;14:1117–29. 10.1016/j.jval.2011.07.011. [DOI] [PubMed] [Google Scholar]
- 17.Janssens A, Rogers M, Thompson Coon J, Allen K, Green C, Jenkinson C, et al. A systematic review of generic multidimensional patient-reported outcome measures for children, part II: evaluation of psychometric performance of English-language versions in a general population. Value Health. 2015;18:334–45. [DOI] [PubMed] [Google Scholar]
- 18.Golicki D, Młyńczak K. Measurement properties of the EQ-5D-Y: a systematic review. Value Health. 2022;25(11):1910–21. [DOI] [PubMed]
- 19.Tan RLY, Soh SZY, Chen LA, Herdman M, Luo N. Psychometric properties of generic preference-weighted measures for children and adolescents: a systematic review. Pharmacoeconomics. 2023;41:155–74. [DOI] [PubMed] [Google Scholar]
- 20.Mokkink LB, de Vet HCW, Prinsen CAC, Patrick DL, Alonso J, Bouter LM, et al. COSMIN risk of bias checklist for systematic reviews of patient-reported outcome measures. Qual Life Res. 2018;27:1171–9. 10.1007/s11136-017-1765-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Terwee CB, Bot SDM, de Boer MR, van der Windt DAWM, Knol DL, Dekker J, et al. Quality criteria were proposed for measurement properties of health status questionnaires. J Clin Epidemiol. 2007;60:34–42. Available from: https://www.jclinepi.com/article/S0895-4356(06)00174-0/fulltext. [DOI] [PubMed]
- 22.Prinsen CAC, Mokkink LB, Bouter LM, Alonso J, Patrick DL, de Vet HCW, et al. COSMIN guideline for systematic reviews of patient-reported outcome measures. Qual Life Res. 2018;27:1147–57. 10.1007/s11136-018-1798-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Terwee CB, Prinsen CAC, Chiarotto A, Westerman MJ, Patrick DL, Alonso J, et al. COSMIN methodology for evaluating the content validity of patient-reported outcome measures: a Delphi study. Qual Life Res. 2018;27:1159–70. 10.1007/s11136-018-1829-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.do Espirito Santo C de M, Santos VS, Miyamoto GC, Chiarotto A, Santos M, Yamato TP. Measurement properties of the EQ-5D in children and adolescents: a systematic review protocol. Syst Rev. 2024;13(18):1–9. [DOI] [PMC free article] [PubMed]
- 25.Rethlefsen ML, Kirtley S, Waffenschmidt S, Ayala AP, Moher D, Page MJ, et al. PRISMA-S: an extension to the PRISMA statement for reporting literature searches in systematic reviews. Syst Rev. 2021;10(39):1–19. [DOI] [PMC free article] [PubMed]
- 26.Elsman EBM, Mokkink LB, Terwee CB, Beaton D, Gagnier JJ, Tricco AC, et al. Guideline for reporting systematic reviews of outcome measurement instruments (OMIs): PRISMA-COSMIN for OMIs 2024. J Clin Epidemiol. 2024;33:2029–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.EuroQol, Group. EuroQol—a new facility for the measurement of health-related quality of life. Health Policy (New York). 1990;16:199–208. [DOI] [PubMed] [Google Scholar]
- 28.Herdman M, Gudex C, Lloyd A, Janssen M, Kind P, Parkin D, et al. Development and preliminary testing of the new five-level version of EQ-5D (EQ-5D-5L). Qual Life Res. 2011;20:1727–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Kreimeier S, Åström M, Burström K, Egmar A-C, Gusi N, Herdman M, et al. EQ-5D-Y-5L: developing a revised EQ-5D-Y with increased response categories. Qual Life Res. 2019;28:1951–61. 10.1007/s11136-019-02115-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Bouwmans C, van der Kolk A, Oppe M, Schawo S, Stolk E, van Agthoven M, et al. Validity and responsiveness of the EQ-5D and the KIDSCREEN-10 in children with ADHD. Eur J Health Econ. 2014;15:967–77. [DOI] [PubMed] [Google Scholar]
- 31.Brussoni M, Kruse S, Walker K. Validity and reliability of the EQ-5D-3LTM among a paediatric injury population. Health Qual Life Outcomes. 2013;11:1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Byford S. The validity and responsiveness of the EQ-5D measure of health-related quality of life in an adolescent population with persistent major depression. J Ment Health. 2013;22:101–10. [DOI] [PubMed] [Google Scholar]
- 33.Cheung PWH, Wong CKH, Lau ST, Cheung JPY. Responsiveness of the EuroQoL 5-dimension (EQ-5D) in adolescent idiopathic scoliosis. Eur Spine J. 2018;27:278–85. [DOI] [PubMed] [Google Scholar]
- 34.Dickerson JF, Feeny DH, Clarke GN, MacMillan AL, Lynch FL. Evidence on the longitudinal construct validity of major generic and utility measures of health-related quality of life in teens with depression. Qual Life Res. 2018;27:447–54. 10.1007/s11136-017-1728-9. [DOI] [PubMed] [Google Scholar]
- 35.Dams J, Rimane E, Steil R, Renneberg B, Rosner R, König HH. Reliability, validity and responsiveness of the EQ-5D-5L in assessing and valuing health status in adolescents and young adults with posttraumatic stress disorder: a randomized controlled trail. Psychiatr Q. 2021;92:459–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Shafie AA, Chhabra IK, Hui Yi JW, Mohammed NS, Ibrahim HM. Validity of the Malay EQ-5D-3L in the Malaysian transfusion-dependent Thalassemia population. Value Health Reg Issues. 2021;24:47–56. [DOI] [PubMed] [Google Scholar]
- 37.Welie AG, Stolk E, Mukuria C, Belay YB, Krahn MD, Sander B, et al. Reliability and validity of using EQ-5D-5L among healthy and adolescents with major mental health disorders in Ethiopia. Eur J Health Econ. 2022;23:1105–19. 10.1007/s10198-021-01412-y. [DOI] [PubMed] [Google Scholar]
- 38.Brooks R. EuroQol: the current state of play. Health Policy (New York). 1996;37:53–72. [DOI] [PubMed] [Google Scholar]
- 39.EuroQol Research Foundation. EQ-5D-Y user guide [Internet]. 2020. Available from: https://euroqol.org/publications/user-guides/.
- 40.EuroQol Research Foundation. EuroQol Group [Internet]. 2024 [cited 2024 Jun 23]. Available from: https://euroqol.org.
- 41.Shiroiwa T, Ikeda S, Noto S, Fukuda T, Stolk E. Valuation survey of EQ-5D-Y based on the international common protocol: development of a value set in Japan. Med Decis Making. 2021;3:1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Prevolnik Rupel V, Ogorevc M, Greiner W, Kreimeier S, Ludwig K, Ramos-Goni JM. EQ-5D-Y value set for Slovenia. Pharmacoeconomics. 2021;39:463–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Roudijk B, Sajjad A, Essers B, Lipman S, Stalmeier P, Finch AP. A value set for the EQ-5D-Y-3L in the Netherlands. Pharmacoeconomics. 2022;40:193–203. 10.1007/s40273-022-01192-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Kreimeier S, Mott D, Ludwig K, Greiner W, Prevolnik Rupel V, Ramos-Goñi JM. EQ-5D-Y value set for Germany. Pharmacoeconomics. 2022;40:217–29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Yang Z, Jiang J, Wang P, Jin X, Wu J, Fang Y, et al. Estimating an EQ-5D-Y-3L value set for China. Pharmacoeconomics. 2022;40:147–55. 10.1007/s40273-022-01216-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Fitriana TS, Roudijk B, Purba FD, Busschbach JJV, Stolk E. Estimating an EQ-5D-Y-3L value set for Indonesia by mapping the DCE onto TTO values. Pharmacoeconomics. 2022;40:157–67. 10.1007/s40273-022-01210-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Ramos-Goñi JM, Oppe M, Estévez-Carrillo A, Rivero-Arias O, Wolfgang G, Simone K, et al. Accounting for unobservable preference heterogeneity and evaluating alternative anchoring approaches to estimate country-specific EQ-5D-Y value sets: a case study using Spanish preference data. Value Health. 2022;25:835–43. [DOI] [PubMed] [Google Scholar]
- 48.Rencz F, Ruzsa G, Bató A, Yang Z, Finch AP, Brodszky V. Value set for the EQ-5D-Y-3L in Hungary. Pharmacoeconomics. 2022;40:205–15. 10.1007/s40273-022-01190-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Dewilde S, Roudijk B, Tollenaar NH, Ramos-Goñi JM. An EQ-5D-Y-3L value set for Belgium. Pharmacoeconomics. 2022;40:169–80. 10.1007/s40273-022-01187-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Mokkink LB, Terwee CB, Patrick DL, Alonso J, Stratford PW, Knol DL, et al. The COSMIN study reached international consensus on taxonomy, terminology, and definitions of measurement properties for health-related patient-reported outcomes. J Clin Epidemiol. 2010;63:737–45. Available from: https://www.jclinepi.com/article/S0895-4356(10)00090-9/fulltext. [DOI] [PubMed]
- 51.Terwee CB, Jansma EP, Riphagen II, de Vet HCW. Development of a methodological PubMed search filter for finding studies on measurement properties of measurement instruments. Qual Life Res. 2009;18:1115–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Bramer WM, Milic J, Mast F. Reviewing retrieved references for inclusion in systematic reviews using endnote. Journal of the Medical Library Association. 2017;105:84–7. Available from: http://jmla.pitt.edu/ojs/jmla/article/view/111. [DOI] [PMC free article] [PubMed]
- 53.Terwee CB, Prinsen CAC, Chiarotto A, De Vet HCW, Bouter LM, Alonso J, et al. COSMIN methodology for assessing the content validity of PROMs—user manual. Amsterdam; 2018. Available from: https://www.cosmin.nl/wp-content/uploads/COSMIN-methodology-for-content-validity-user-manual-v1.pdf. [DOI] [PMC free article] [PubMed]
- 54.Terwee CB, Mokkink LB, Knol DL, Ostelo RWJG, Bouter LM, de Vet HCW. Rating the methodological quality in systematic reviews of studies on measurement properties: a scoring system for the COSMIN checklist. Qual Life Res. 2012;21:651–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Brazier J, Deverill M. A checklist for judging preference-based measures of health related quality of life: learning from psychometrics. Health Econ. 1999;8:41–51. [DOI] [PubMed] [Google Scholar]
- 56.Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-Coello P, et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ [Internet]. 2008;336:924–6. Available from: https://www.bmj.com/content/336/7650/924. [DOI] [PMC free article] [PubMed]
- 57.Mokkink LB, Prinsen CAC, Patrick DL, Alonso J, Bouter LM, De Vet HCW, et al. COSMIN methodology for systematic reviews of Patient-Reported Outcome Measures (PROMs)—user manual [Internet]. Amsterdam; 2018. Available from: https://www.cosmin.nl/.
- 58.Åström M, Åström M, Åström M, Krig S, Ryding S, Cleland N, et al. EQ-5D-Y-5L as a patient-reported outcome measure in psychiatric inpatient care for children and adolescents—a cross-sectional study. Health Qual Life Outcomes. 2020;18:1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 59.Bergfors S, Åström M, Burström K, Egmar AC. Measuring health-related quality of life with the EQ-5D-Y instrument in children and adolescents with asthma. Acta Paediatrica Int J Paediatr. 2015;104:167–73. [DOI] [PubMed] [Google Scholar]
- 60.Burström K, Bartonek A, Broström EW, Sun S, Egmar AC. EQ-5D-Y as a health-related quality of life measure in children and adolescents with functional disability in Sweden: testing feasibility and validity. Acta Paediatrica Int J Paediatr. 2014;103:426–35. [DOI] [PubMed] [Google Scholar]
- 61.Burström K, Svartengren M, Egmar AC. Testing a Swedish child-friendly pilot version of the EQ-5D instrument—initial results. Eur J Public Health. 2011;21:178–83. [DOI] [PubMed] [Google Scholar]
- 62.Canaway AG, Frew EJ. Measuring preference-based quality of life in children aged 6–7 years: a comparison of the performance of the CHU-9D and EQ-5D-Y—the WAVES pilot study. Qual Life Res. 2013;22:173–83. [DOI] [PubMed] [Google Scholar]
- 63.Chen G, Flynn T, Stevens K, Brazier J, Huynh E, Sawyer M, et al. Assessing the health-related quality of life of Australian adolescents: an empirical comparison of the child health utility 9D and EQ-5D-Y instruments. Value Health. 2015;18:432–8. 10.1016/j.jval.2015.02.014. [DOI] [PubMed] [Google Scholar]
- 64.Eidt-Koch D, Mittendorf T, Greiner W. Cross-sectional validity of the EQ-5D-Y as a generic health outcome instrument in children and adolescents with cystic fibrosis in Germany. BMC Pediatr. 2009;9:55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 65.Gusi N, Perez-Sousa MA, Gozalo-Delgado M, Olivares PR. Validez y fiabilidad de la versión proxy del EQ-5D-Y en español. An Pediatr (Engl Ed). 2014;81:212–9. [DOI] [PubMed] [Google Scholar]
- 66.Hsu CN, Lin H-W, Pickard AS, Tain Y-L. EQ-5D-Y for the assessment of health-related quality of life among Taiwanese youth with mild-to-moderate chronic kidney disease. Int J Qual Health Care. 2018;30:298–305. [DOI] [PubMed] [Google Scholar]
- 67.Kim SKSK, Jo MW, Kim SH. A cross sectional survey on health-related quality of life of elementary school students using the Korean version of the EQ-5D-Y. PeerJ. 2017;2017:1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 68.Kim SK, Jo MW, Kim SH. Health-related quality of life by allergy symptoms in elementary school students. Health Qual Life Outcomes. 2018;16:1–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 69.Klaassen RJ, Krahn M, Gaboury I, Hughes J, Anderson R, Grundy P, et al. Evaluating the ability to detect change of health-related quality of life in children with Hodgkin disease. Cancer. 2010;116:1608–14. [DOI] [PubMed] [Google Scholar]
- 70.Matza LS, Secnik K, Mannix S, Sallee FR. Parent-proxy EQ-5D ratings of children with attention-deficit hyperactivity disorder in the US and the UK. Pharmacoeconomics. 2005;23:777–90. [DOI] [PubMed] [Google Scholar]
- 71.Mayoral K, Rajmil L, Murillo M, Garin O, Pont A, Alonso J, et al. Measurement properties of the online EuroQol-5D-youth instrument in children and adolescents with type 1 diabetes mellitus: Questionnaire study. J Med Internet Res. 2019;21(11):e14947. [DOI] [PMC free article] [PubMed]
- 72.Pan CW, Zhong H, Li J, Suo C, Wang P. Measuring health-related quality of life in elementary and secondary school students using the Chinese version of the EQ-5D-Y in rural China. BMC Public Health. 2020;20:1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 73.Robles N, Rajmil L, Rodriguez-Arjona D, Azuara M, Codina F, Raat H, et al. Development of the web-based Spanish and Catalan versions of the Euroqol 5D-Y (EQ-5D-Y) and comparison of results with the paper version. Health Qual Life Outcomes. 2015;13:1–9. Available from: ???. [DOI] [PMC free article] [PubMed]
- 74.Ryan JM, McKay E, Anokye N, Noorkoiv M, Theis N, Lavelle G. Comparison of the CHU-9D and the EQ-5D-Y instruments in children and young people with cerebral palsy: a cross-sectional study. BMJ Open. 2020;10(9):e037089. [DOI] [PMC free article] [PubMed]
- 75.Scalone L, Tomasetto C, Matteucci MC, Selleri P, Broccoli S, Pacelli B, et al. Assessing quality of life in children and adolescents: development and validation of the Italian version of the EQ-5D-Y. Ital J Public Health. 2011;8:331–41. [Google Scholar]
- 76.Verstraete J, Lloyd A, Scott D, Jelsma J. How does the EQ-5D-Y proxy version 1 perform in 3, 4 and 5-year-old children? Health Qual Life Outcomes. 2020;18:1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 77.Vitale MG, Levy DE, Johnson MG, Gelijns AC, Moskowitz AJ, Roye BP, et al. Assessment of quality of life in adolescent patients with orthopaedic problems: are adult measures appropriate? J Pediatr Orthop. 2001;21:622–8. [PubMed] [Google Scholar]
- 78.Willems DCM, Joore MA, Nieman FHM, Severens JL, Wouters EFM, Hendriks JJE. Using EQ-5D in children with asthma, rheumatic disorders, diabetes, and speech/language and/or hearing disorders. Int J Technol Assess Health Care. 2009;25:391–9. [DOI] [PubMed] [Google Scholar]
- 79.Shiroiwa T, Fukuda T, Shimozuma K. Psychometric properties of the Japanese version of the EQ-5D-Y by self-report and proxy-report: reliability and construct validity. Qual Life Res. 2019;28:3093–105. 10.1007/s11136-019-02238-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 80.Scott D, Scott C, Jelsma J, Abraham D, Verstraete J. Validity and feasibility of the self-report EQ-5D-Y as a generic health-related quality of life outcome measure in children and adolescents with juvenile idiopathic arthritis in Western Cape, South Africa. South Afr J Physiotherapy. 2019;75:1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 81.Scott D, Ferguson GD, Jelsma J. The use of the EQ-5D-Y health related quality of life outcome measure in children in the Western Cape, South Africa: psychometric properties, feasibility and usefulness—a longitudinal, analytical study. Health Qual Life Outcomes. 2017;15:1–14. 10.1186/s12955-017-0590-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 82.Wong CKH, Cheung PWH, Luo N, Cheung JPY. A head-to-head comparison of five-level (EQ-5D-5L-Y) and three-level EQ-5D-Y questionnaires in paediatric patients. Eur J Health Econ. 2019;20:647–56. 10.1007/s10198-018-1026-7. [DOI] [PubMed] [Google Scholar]
- 83.Amien R, Scott D, Verstraete J. Performance of the EQ-5D-Y interviewer administered version in young children. Children. 2022;9:1–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 84.Fitriana TS, Purba FD, Rahmatika R, Muhaimin R, Sari NM, Bonsel G, et al. Comparing measurement properties of EQ-5D-Y-3L and EQ-5D-Y-5L in paediatric patients. Health Qual Life Outcomes. 2021;19:1–12. 10.1186/s12955-021-01889-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 85.Fitriana TS, Purba FD, Stolk E, Busschbach JJV. EQ-5D-Y-3L and EQ-5D-Y-5L proxy report: psychometric performance and agreement with self-report. Health Qual Life Outcomes. 2022;20:1–13. 10.1186/s12955-022-01996-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 86.Lin J, Wong CKH, Cheung JPY, Cheung PWH, Luo N. Psychometric performance of proxy-reported EQ-5D youth version 5-level (EQ-5D-Y-5L) in comparison with three-level (EQ-5D-Y-3L) in children and adolescents with scoliosis. Eur J Health Econ. 2022;23:1383–95. 10.1007/s10198-022-01435-z. [DOI] [PubMed] [Google Scholar]
- 87.Mayoral K, Garin O, Lizano-Barrantes C, Pont A, Caballero-Rabasco AM, Praena-Crespo M, et al. Measurement properties of the EQ-5D-Y administered through a smartphone app in children with asthma: a longitudinal questionnaire study. Health Qual Life Outcomes. 2022;20:1–16. 10.1186/s12955-022-01955-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 88.Pei W, Yue S, Zhi-Hao Y, Ruo-Yu Z, Bin W, Nan L. Testing measurement properties of two EQ-5D youth versions and KIDSCREEN-10 in China. Eur J Health Econ. 2021;22:1083–93. 10.1007/s10198-021-01307-y. [DOI] [PubMed] [Google Scholar]
- 89.Pérez-Sousa MÁ, Olivares PR, Ramírez-Vélez R, Gusi N. Comparison of the psychometric properties of the EQ-5D-3L-Y and EQ-5D-5L-Y instruments in Spanish children and adolescents. Value Health. 2021;24:1799–806. 10.1016/j.jval.2021.07.006. [DOI] [PubMed] [Google Scholar]
- 90.Zhou W, Shen A, Yang Z, Wang P, Wu B, Herdman M, et al. Patient-caregiver agreement and test–retest reliability of the EQ-5D-Y-3L and EQ-5D-Y-5L in paediatric patients with haematological malignancies. Eur J Health Econ. 2021;22:1103–13. 10.1007/s10198-021-01309-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 91.Verstraete J, Marthinus Z, Dix-Peek S, Scott D. Measurement properties and responsiveness of the EQ-5D-Y-5L compared to the EQ-5D-Y-3L in children and adolescents receiving acute orthopaedic care. Health Qual Life Outcomes. 2022;20:1–14. 10.1186/s12955-022-01938-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 92.Verstraete J, Scott D. Comparison of the EQ-5D-Y-5L, EQ-5D-Y-3L and PedsQL in children and adolescents. J Patient Rep Outcomes. 2022. 10.1186/s41687-022-00480-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 93.Verstraete J, Amien R, Scott D. Comparing measurement properties of the English EQ-5D-Y 3-level version with the 5-level version in South Africa. Value Health Reg Issues. 2022;30:140–7. 10.1016/j.vhri.2021.12.004. [DOI] [PubMed] [Google Scholar]
- 94.Verstraete J, Scott D. The performance of the EQ-5D-Y-5L compared to the EQ-5D-Y-3L in children and adolescents with cerebral palsy (CP). Dialog Health. 2022;1:100032. 10.1016/j.dialog.2022.100032. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 95.Amien R, Scott D, Verstraete J. The validity and reliability of the interviewer-administered EQ-5D-Y-3L version in young children. Health Qual Life Outcomes. 2023;21(19):1–13. [DOI] [PMC free article] [PubMed]
- 96.Downs J, Norman R, Mulhern B, Jacoby P, Reddihough D, Choong CS, et al. Psychometric properties of the EQ-5D-Y-5L for children with intellectual disability. Value Health. 2024;27:776–83. [DOI] [PubMed] [Google Scholar]
- 97.He S, Chen L, Wang P, Xi X. Psychometric comparison of EQ-5D-Y, CHU-9D, and PedsQL 4.0 in Chinese children and adolescents with functional dyspepsia: a multicenter study. Value Health. 2023;26:1754–62. [DOI] [PubMed] [Google Scholar]
- 98.Hsu CN, Tain YL, Lu PC, Lin HW. Comparisons of EQ-5D-Y and PedsQL in pediatric patients with mild-to-moderate chronic kidney disease in longitudinal analyses. Health Qual Life Outcomes. 2023;21:1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 99.Jones R, O’Loughlin R, Xiong X, Bahrampour M, Devlin N, Hiscock H, et al. Comparative psychometric performance of common generic paediatric health-related quality of life instrument descriptive systems: results from the australian paediatric multi-instrument comparison study. Pharmacoeconomics. 2024;42:39–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 100.Mihalopoulos C, Chen G, Scott JG, Bucholc J, Allen C, Coghill D, et al. Assessing outcomes for cost-utility analysis in children and adolescents with mental health problems: are multiattribute utility instruments fit for purpose? Value Health. 2023;26:733–41. [DOI] [PubMed] [Google Scholar]
- 101.Ngwira LG, Maheswaran H, Verstraete J, Petrou S, Niessen L, Smith SC. Psychometric performance of the Chichewa versions of the EQ-5D-Y-3L and EQ-5D-Y-5L among healthy and sick children and adolescents in Malawi. J Patient Rep Outcomes. 2023;7:1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 102.O’Loughlin R, Jones R, Chen G, Mulhern B, Hiscock H, Devlin N, et al. Comparing the psychometric performance of generic paediatric health-related quality of life instruments in children and adolescents with ADHD, anxiety and/or depression. Pharmacoeconomics. 2024;42:57–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 103.Perry N, Boulton KA, Hodge A, Ong N, Phillips N, Howard K, et al. A psychometric investigation of health-related quality of life measures for paediatric neurodevelopment assessment: reliability and concurrent validity of the PEDS-QL, CHU-9D, and the EQ-5D-Y. Autism Res. 2024;17:972–88. [DOI] [PubMed] [Google Scholar]
- 104.van Heusden A, Rivero-Arias O, Herdman M, Hiscock H, Devlin N, Dalziel K, et al. Psychometric performance comparison of the adapted versus original versions of the EQ-5D-Y-3L and -Y-5L in proxy respondents for 2- to 4-year-olds. Pharmacoeconomics. 2024;42:129–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 105.Verstraete J, Booth C, Booth J, Peer S, McGuire J, Kritzinger F, et al. Fighting to breathe and fighting for health-related quality of life: measuring the impact of being dependent on technology for breathing on the child and their caregiver. Patient. 2024;17:65–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 106.Xu RH, Dai Y, Ng SSM, Tsang HWH, Zhang S, Dong D. Assessing validity of the EQ-5D-5L proxy in children and adolescents with Duchenne muscular dystrophy or spinal muscular atrophy. Eur J Health Econ. 2024;25:103–15. [DOI] [PubMed] [Google Scholar]
- 107.Xu RH, Zhu L, Sun R, Tan RLY, Luo N, Zou S, et al. Investigating the psychometric properties of the EQ-5D-Y-3L, EQ-5D-Y-5L, CHU-9D, and PedsQL in children and adolescents with osteogenesis imperfecta. Eur J Pediatr. 2022;181:4049–58. [DOI] [PubMed] [Google Scholar]
- 108.Zhou W, Shen A, Yang Z, Wang P, Wu B, Herdman M, et al. Validity and responsiveness of EQ-5D-Y in children with haematological malignancies and their caregivers. Eur J Health Econ. 2024:25(8):1361–70. [DOI] [PubMed]
- 109.Chiarotto A, Terwee CB, Kamper SJ, Boers M, Ostelo RW. Evidence on the measurement properties of health-related quality of life instruments is largely missing in patients with low back pain: a systematic review. J Clin Epidemiol. 2018;102:23–37. 10.1016/j.jclinepi.2018.05.006. [DOI] [PubMed] [Google Scholar]
- 110.Kreimeier S, Greiner W. EQ-5D-Y as a health-related quality of life instrument for children and adolescents: the instrument’s characteristics, development, current use, and challenges of developing its value set. Value Health. 2019;22:31–7. [DOI] [PubMed] [Google Scholar]
- 111.Reeve BB, Wyrwich KW, Wu AW, Velikova G, Terwee CB, Snyder CF, et al. ISOQOL recommends minimum standards for patient-reported outcome measures used in patient-centered outcomes and comparative effectiveness research. Qual Life Res. 2013;22:1889–905. [DOI] [PubMed] [Google Scholar]
- 112.Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977;33:159–74. [PubMed] [Google Scholar]
- 113.Konerding U. What does Cronbach’s alpha tell us about the EQ-5D? A methodological commentary to “psychometric properties of the EuroQol Five-Dimensional Questionnaire (EQ-5D-3L) in caregivers of autistic children.” Qual Life Res. 2013;22:2939–40. [DOI] [PubMed] [Google Scholar]
- 114.Buchholz I, Janssen MF, Kohlmann T, Feng Y-S. A systematic review of studies comparing the measurement properties of the three-level and five-level versions of the EQ-5D. Pharmacoeconomics. 2018;36:645–61. 10.1007/s40273-018-0642-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 115.Kreimeier S, Oppe M, Ramos-Goñi JM, Cole A, Devlin N, Herdman M, et al. Valuation of EuroQol Five-Dimensional Questionnaire, youth version (EQ-5D-Y) and EuroQol five-dimensional questionnaire, three-level version (EQ-5D-3L) health states: the impact of wording and perspective. Value Health. 2018;21:1291–8. 10.1016/j.jval.2018.05.002. [DOI] [PubMed] [Google Scholar]
- 116.Ramos-Goñi JM, Oppe M, Stolk E, Shah K, Kreimeier S, Rivero-Arias O, et al. International valuation protocol for the EQ-5D-Y-3L. Pharmacoeconomics. 2020;38:653–63. 10.1007/s40273-020-00909-3. [DOI] [PubMed] [Google Scholar]
- 117.Gagnier JJ, Lai J, Mokkink LB, Terwee CB. COSMIN reporting guideline for studies on measurement properties of patient-reported outcome measures. Qual Life Res. 2021;30:2197–218. 10.1007/s11136-021-02822-4. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
All data included in this study are freely available upon request to the study authors.