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

There is a need for tools that can provide a brief assessment of functioning for children with neurodevelopmental conditions, including health‐related quality of life (HR‐QoL). This study evaluated the psychometric properties of three commonly used and well known HR‐QoL measures in a cohort of children presenting to clinical developmental assessment services. The most common diagnoses received in these assessment services were autism spectrum disorders. Findings showed good internal consistency for the PedsQL and the CHU‐9D, but not the EQ‐5D‐Y. This research also found that the CHU‐9D, EQ‐5D‐Y, and PedsQL correlated with relevant functioning domains assessed by the VABS‐III. Overall, the measures showed that children with neurodevelopmental conditions experienced poor HR‐QoL. The majority of children (>86%) met cut‐off criteria for significant health concerns on the PedsQL. On the EQ‐5D‐Y and CHU‐9D, they showed reduced HR‐QoL particularly on domains relating to school and homework, being able to join in activities, looking after self, and doing usual activities. This study supports the use of the CHU‐9D and PedsQL in this population to assess and potentially track HR‐QoL in a broad neurodevelopment paediatric population.


Lay Summary
There is a need for brief assessments of health-related quality of life (HR-QoL) for children with neurodevelopmental conditions.This study evaluated the psychometric properties of three commonly used and well-known HR-QoL measures in a cohort of children presenting for clinical developmental assessment.The most common diagnoses received in these assessment services were autism spectrum disorders.Findings supported the reliability of the PedsQL and the CHU-9D, but not the EQ-5D-Y.This research found that the CHU-9D, EQ-5D-Y, and PedsQL correlated with relevant functioning domains on the VABS-III, another commonly used functioning measure.Measures showed that children with neurodevelopmental

INTRODUCTION
The International Classification of Functioning, Disability and Health, known more commonly as ICF, emphasizes frameworks for understanding how an individual's health condition impacts their interactions with their environment (WHO, 2001).The ICF framework has had significant influence on the conceptualisation and quantification of the impact of disability, shifting focus from the assessment of symptoms to impact across domains of functioning (Sharma, 2016).Such frameworks have become important for tracking outcomes for the delivery of services aimed to improve functional independence.For example, the National Disability Insurance Scheme [NDIS] (2020) in Australia uses the framework to track the effectiveness of assessments and supports for people with disabilities across domains of communication, social interaction, learning, mobility, self-care and selfmanagement (NDIA, 2020).
International awareness is growing to better recognize and support functioning in people with neurodevelopmental conditions (NDCs).In conjunction with the NDIS in Australia, several countries including England, Denmark, France, and Hungary have adopted specific legal frameworks to support daily functioning in people with disabilities (Della Fina & Cera, 2015).As a result of this increased focus on functional outcomes, there is growing interest in the different measures of, and conceptualizations for, functioning (Bölte, de Schipper, Holtmann, et al., 2014;Bölte, de Schipper, Robison, et al., 2014;Purcell et al., 2015;Sareen et al., 2005;Saulnier & Klaiman, 2022).One of the most widely used measures of adaptive functioning is the Vineland Adaptive Behavior Scales, third edition (VABS-III).The VABS-III is considered a gold-standard tool in both clinical and research settings (Sparrow et al., 2016) and assesses four domains of functioning: Communication, Daily Living Skills, Socialization, and Motor Skills (Sparrow et al., 2016).Although the VABS-III is well established as a measure of adaptive functioning, limitations have been noted, such as its time to completion (NDIA, 2021).
There is a need for tools that can provide a brief assessment of different aspects of adaptive functioning.A related but distinct concept to adaptive functioning is health related quality of life (HR-QoL).HR-QoL is a multi-dimensional concept that measures how a person's physical, psychological and social wellbeing is impacted by their health condition (Hand, 2016).HR-QoL is a dynamic concept, defined as the assessment of an individual's perceived wellbeing over time and is not dependent on physical health alone, but rather the interaction of many relationships in an individuals life (Lin et al., 2013).Factors that influence adaptive functioning also impact HR-QoL in many disabilities.For example, HR-QoL is greatly reduced in children who have communication difficulties (Nicola & Watter, 2015), poor mobility (Bray et al., 2017), and limited socialization (Katsiana et al., 2020).
Previous research indicates that children with NDCs experience considerable challenges in their HR-QoL compared to their neurotypical peers, with higher rates of depression (Gurney et al., 2006), chronic conditions (Kuhlthau et al., 2010) and poorer physical health (more sleep problems and allergies) (Katsiana et al., 2020).Furthermore, it has been established that impaired adaptive functioning in early childhood is associated with poorer HR-QoL outcomes over time (Forde et al., 2022).
Discrepancies between proxy and child self-report have been found in previous studies (Khanna et al., 2022;Petrou, 2003;Sheldrick et al., 2012).Children with NDCs can experience challenges when reporting on their own quality of life (i.e., communication difficulties, cognitive impairment, young age) (Petrou, 2003;Tavernor et al., 2013).Therefore, proxy reports have utility in these populations and are commonly used in assessment services.The EuroQol-5D (Youth; EQ-5D-Y) (Ravens-Sieberer et al., 2010), the Child Health Utility-9D instrument (CHU-9D) (Stevens, 2012), and the Pediatric Quality of Life Inventory™ 4.0 (PedsQL) (Varni et al., 2003), are widely used assessments designed to measure various aspects of HR-QoL such as daily activities, mobility and socialization.The EQ-5D-Y, CHU-9D and PedsQL are generic HR-QoL questionnaires that can involve a caregiver providing a proxy report on their child's experiences in their daily life, including their routine and activities (i.e., doing schoolwork, playing), their feelings (i.e., feeling worried, sad, or pain) and mobility (i.e., walking or running), amongst others.Domains within HR-QoL measures are also captured in the six functional activity areas specified in the NDIS Act (see Figure 1), suggesting that measures assessing HR-QoL could support assessment of adaptive functioning as defined by leading health organizations.For this reason, HR-QoL assessments warrant investigation as a complementary means of assessing adaptive functioning and health status in children with NDCs.There has, however, been limited research into the psychometric properties of the CHU-9D, EQ-5D-Y, and PedsQL in children who present with a diverse range of NDCs to assessment clinics (Raghunandan et al., 2023;Stokes et al., 2017;Xiong et al., 2023).
The aim of this study was to evaluate the psychometric properties of established HR-QoL instruments in a cohort of children attending a diagnostic assessment service for neurodevelopmental concerns.Based on their use in the paediatric literature generally (Lindvall et al., 2021;Ravens-Sieberer et al., 2010;Varni et al., 2003), the CHU-9D, EQ-5D-Y and the PedsQL were selected as measures to explore in a NDC cohort.The psychometric properties of these instruments, specifically, internal consistency, item-total correlations, concurrent, discriminant validity and floor and ceiling effects were investigated.The relationship between these HR-QoL measures with a goldstandard assessment of adaptive functioning, the VABS-III, was examined to ascertain the concurrent validity of the HR-QoL measures.We also assessed the discriminant validity (whether two scales are empirically distinct) between these HR-QoL measures and the strengths and difficulties questionnaire (SDQ) (Goodman, 2001;Rönkkö & Cho, 2022) which captures emotional and behavioral difficulties and has been widely used in NDC populations (Bryant et al., 2020).

METHOD Participants and setting
Participants in this study were 204 children aged between 1.63 and 16.55 years (M = 6.95,SD = 3.76; Male: 148, Female: 56), recruited from The Child Development Unit (CDU) using the Sydney Child Neurodevelopment Registry (Boulton et al., 2023), in partnership with the Clinic for Autism and Neurodevelopmental Research.Informed consent was obtained from parents or caregivers.The CDU is a tertiary public health unit that provides developmental diagnostic assessments for children and adolescents who present with complex developmental concerns.The CDU assesses children from a wide range of regional and metropolitan areas across New South Wales regardless of family income, resulting in a vast cultural, demographic and linguistically diverse clinical population.Children with developmental concerns are referred to the CDU by their paediatrician for developmental assessment.Multidisciplinary assessments are carried out by diagnostic specialists, including paediatricians, neuropsychologists, social workers, speech pathologists, nurses, and occupational therapists.On average, about 75% of children receive an autism related diagnosis and over 50% receive more than one diagnosis.

Measures
Health-related quality of life (HR-QoL) The EuroQol-5D -Youth (EQ-5D-Y) The EQ-5D-Y (version 2.0) is a self-or proxy-completed HR-QoL assessment designed for children and adolescents aged from 4 to 15 years (Ravens-Sieberer et al., 2010).The proxy report version was completed by caregivers for this study.The EQ-5D-Y takes 2-5 minutes to complete and consists of 5 items, rated using a 3-level Likert scale (1 = no problems; 3 = a lot of problems).Higher scores indicate greater problems in perceived HR-QoL.Domains assess mobility, looking after oneself, doing usual activities, having pain or discomfort and feeling worried, sad, or unhappy.The EQ-5D-Y also includes a Visual Analogue Scale (VAS), where participants can visually rate their level of health from 0 to 100, 100 being the best possible health status.After completion, a value set is calculated which is coded as a 5-digit number describing a patient's health status, which is then converted to a single score on a 0 (dead) to 1 (full health) utility scale.Both item scores and VAS scores of the EQ-5D-Y were evaluated, and, due to the current lack of Australian utility index scores, we calculated utility scores using recently developed German value sets, as recommended by Kreimeier et al. (2022).
The EQ-5D-Y has shown adequate reliability in the 5 domain-5 level version (EQ-5D-5L) in patients with chronic health conditions (a = 0.85-0.79)(Seng et al., 2020;Tran et al., 2012).It has also shown excellent convergent validity with other HR-QoL tools including the PedsQL (Bergfors et al., 2015).However, prior studies have reported high ceiling effects in the EQ-5D-Y in large studies of typically developing children and those with asthma (Mayoral et al., 2022;Ravens-Sieberer et al., 2010).

The Child Health Utility-9D (CHU-9D)
The CHU-9D is a proxy or child self-report questionnaire developed for children and adolescents aged 7-17 years, and has been validated for use in child and adolescent populations in Australia and the United Kingdom (Chen et al., 2015;Stevens & Ratcliffe, 2012).The proxyreport was completed by caregivers for this study.The CHU-9D takes approximately 2-7 minutes to complete and contains 9 items, (feeling worried, sad, pain, tired, annoyed, schoolwork/homework, sleep and daily routine), with responses rated using a 5-point Likert scale indicating levels of increasing severity, (e.g., 1 = do not feel worried; 5 = feel very worried).Higher scores on the CHU-9D correspond to a poorer perceived HR-QoL.The instrument is further scored using algorithms that are preference-based and produce utility weights (Furber & Segal, 2015), on a 0 (dead) to 1 (full health) scalewhere higher scores represented better utility based QOL.The current study used CHU-9D item scores, total scores and Australian utility index scores.With respect to existing psychometric properties for the CHU-9D, measures of internal consistency have been adequate (a = >0.7) in a cohort of neurotypical schoolaged children (Lindvall et al., 2021), and the CHU-9D has been reported to be able to discriminate between different paediatric groups in terms of mental health severity (Wolf et al., 2021).
The Pediatric Quality of Life Inventory (PedsQL) The PedsQL is designed for children and adolescents aged from 2 to 18 years and takes 5 minutes to complete (Varni et al., 2003).It is conducted as a self-report tool or through proxy-report.The proxy-report version used corresponded to the age of the child (i.e., 2-4 years, 5-7, 8-12, and 13-18 years), and was completed by caregivers for this study.The PedsQL consists of 23 items assessing 4 broad scales including: physical functioning, emotional functioning, social functioning and school functioning.Questions are rated using a 5-point Likert scale, (0 = never; 4 = almost always), with the total score computed as the sum of all answered items.Item scores on the PedsQL are reversed and linearly transformed to a 0-100 scale, with higher scores on the PedsQL indicating better perceived HR-QoL.The current study used total scores and scale scores of the PedsQL.Clinical cut-off scores have been proposed for the PedsQL to provide an indication of the presence of chronic health conditions.Recommended cut-off scores are 83 and below for those aged under 8 years, and 78 and below for those aged over 8 years on the total functioning domain (Huang et al., 2009).
The PedsQL proxy report form has demonstrated excellent reliability in a paediatric cohort of cancer patients (a = 0.93) (Varni et al., 2002).In addition, moderate correlations have been observed between the PedsQL and other HR-QoL questionnaires (r = 0.4-0.6) in children attending clinical hospital settings (Amedro et al., 2021).The PedsQL has also been found to be effective at discriminating between healthy and ill children (Viecili & Weiss, 2015).Of note, the PedsQL has previously been used in NDC cohorts and has been validated in autistic children, including with parents of autistic children (Katsiana et al., 2020;Lee et al., 2016;Stokes et al., 2017).

Adaptive functioning
The Vineland Adaptive Behavior Scales, third edition (VABS-III) The VABS-III is designed for use from 0 to 90 years of age, and is administered as a semi-structured interview by a trained assessor, or by proxy-report questionnaire (teacher or parent/caregiver reported).The VABS-III uses a 3-point Likert scale (0 = never; 2 = usually).The comprehensive version of the VABS-III comprises 502 items, and takes approximately 40-60 minutes to complete (Cordeiro et al., 2020).There is good concordance between the Interview and Parent/Caregiver versions (Sparrow et al., 2016).Standard scores are generated from responses in four domains: Communication, Daily Living Skills, Socialization, and Motor Skills.A completed VABS-III reveals an overall Adaptive Behavior Composite (ABC) score with a normative mean of 100 and a standard deviation of 15.A lower score indicates reduced adaptive functioning capacity.The ABC serves as a measure of an individual's adaptive functioning, as well as an extension to evaluate against The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) criteria for NDCs.The VABS-III has demonstrated consistent psychometric properties of excellent internal consistency (from a = 0.94-0.99)and moderate to strong concurrent validity (r = 0.32-0.83)on the comprehensive forms in a paediatric cohort (Sparrow et al., 2016).The VABS-III has been found to be useful in assessment of adaptive functioning in NDCs (Faja et al., 2023).

The Strengths and Difficulties Questionnaire (SDQ)
The Strengths and Difficulties Questionnaire (SDQ) is a behavior and mental health questionnaire that takes 3-5 minutes to complete and is used in children and adolescents aged 2-17 years (Goodman, 2001).The SDQ was conducted in this study by proxy report.The SDQ consists of 25 questions assessing areas of emotional symptoms, conduct problems, hyperactivity/inattention, peer relationships, and prosocial behavior which are scored on a 3-point Likert scale and summed to produce domain level and total scores.Higher scores indicate more problem behaviors.A UK study has established satisfactory Abbreviations: CHU-9D, the child health utility instrument-9D; EQ-5D-Y, the EuroQol-5D; PedsQL, the Pediatric Quality of Life inventory 4.0.
internal consistency of the SDQ in a sample of children from the general population (a = 0.73) (Goodman, 2001).

Procedure
The VABS-III assessments were performed in The Child Development Unit (CDU) at The Children's Hospital at Westmead, by clinical staff trained to administer and score each measure.The HR-QoL measures were emailed to caregivers approximately 1 month prior to their child's appointment and completed via the Research Enterprise Data Capture (REDCap) platform, an electronic data collection system endorsed by the University of Sydney (Harris et al., 2009;Harris et al., 2019) as part of the Sydney Child Neurodevelopment Research Registry.Families received an email reminder to complete questionnaires 1 week prior to their appointment, and those who did not do so before their appointment were required to complete it on the day of assessment as per the process described in Patel et al. (2021).The HR-QoL measures were sent out as part of a nationally recommended transdiagnostic protocol of questionnaire-based assessments, to collect information on the broad needs of children and families attending the CDU (Boulton et al., 2021).In relation to community involvement, we consulted with the community about the relevancy and selection of measures, but did not consult with autistic children, families or the broader community about the specific psychometric evaluation used in this research (Boulton et al., 2021).It is the goal of future research to involve children and families in the design, implementation, and evaluation of future studies.

Statistical analysis
All statistical analyses were performed on SPSS ® Statistics, v27.0 ©IBM.Only children who had data on one or more of the HR-QoL measures (i.e., CHU-9D, EQ-5D-Y, PedsQL) were included in this study.The minimum sample size required for reliability analysis was drawn from Bujang et al. (2018) and was based on a predicted minimum value score of 0.70 and a null hypothesis set at 0.50, which is N = 87 at 90% power and N = 66 at 80% power for a 25-item questionnaire (Bujang et al., 2018).Internal consistency, item-total correlations, concurrent validity, discriminant validity and floor and ceiling effects were evaluated for each HR-QoL instrument in the study population.Given the CHU-9D and EQ-5D-Y have been validated to date in children aged 7-17 and 4-15 respectively, analyses were initially conducted only on children in those age ranges.Exploratory analyses were then conducted using data from all respondents.
Internal consistency of the CHU-9D, EQ-5D-Y, and the PedsQL was evaluated using Cronbach's coefficient alpha, to determine the degree at which all the item scores within a measure correlate with one another.A Cronbach's alpha value of 0.7 was considered acceptable, between 0.7-0.9 was considered adequate and greater than >0.9 was considered excellent (Nunnally, 1967).Pearson's correlation coefficients (r) were used to calculate item-total correlations of HR-QoL measures.Items with correlations below <0.3 were considered poor, correlations between 0.3-0.5 were considered moderate and correlations above >0.5 were considered strong (Cohen, 1988).
Concurrent validity (with VABS-III) and discriminant validity (with the SDQ and age) were evaluated using Pearson's correlation coefficients (r) and evaluating the strength of the correlations between different measures (Eid et al., 2015;Lenhard & Lenhard, 2014;Steiger, 1980).For the CHU9D and EQ-5D-Y, utility scores were used to assess concurrent validity and for the Peds-QL, domain scores were correlated with corresponding VABS-III domains.In addition, we looked at concurrent validity between VABS-III and conceptually related items/domains on the HR-QoL measures (see Figure 1 for a description of conceptually overlapping domains of assessments and the NDIS domains.Coefficients of <0.20 were defined as very weak, 0.20-0.39 as weak, 0.40-0.59as moderate, 0.6-0.79 as strong, and >0.8 as very strong (Swinscow, 1996).Floor and ceiling effects on the CHU-9D, EQ-5D-Y and the PedsQL were considered present if ≥15% of scores were at floor or ceiling (Terwee et al., 2007).Finally, as cut-off scores have been proposed for the total functioning domain of the PedsQL, we also looked at how many children in this sample scored above proposed cut-offs which is indicative of potential health care needs (Huang et al., 2009).

Demographics
Participants were 204 children who received a developmental assessment at the CDU (see Table 1).The most common diagnoses reported in the sample were Autism Spectrum Disorder (ASD) and Attention Deficit/ Hyperactivity Disorder (ADHD), see Supplementary Figure S1.As part of routine clinical care, caregivers completed a range of standardized assessments and questionnaires, including one or more of the primary outcome measures on HR-QoL, including the CHU-9D (N = 200), EQ-5D-Y (N = 203) and PedsQL (N = 193), and an assessment of the child's adaptive functioning, the VABS-III (N = 105).With respect to the VABS-III, 46 participants completed the comprehensive interview form, and 59 completed the parent/caregiver survey.There were no statistically significant differences between T A B L E 2 Descriptive statistics of participant responses for the HR-QoL measures.Abbreviations: CHU-9D, the child health utility instrument-9D; EQ-5D-Y, the EuroQol-5D; PedsQL, the Pediatric Quality of Life inventory 4.0.
these versions on any domain ( p = >0.05),so VABS-III scores were combined for further analyses.

Descriptive statistics -HR-QoL measures
Table 2 displays descriptive statistics for each measure in specified age groups.In this cohort, 87.5% (n = 112) of the sample under 8 and 87.7% (n = 57) over 8 years of age, met cut-off criteria for possible health concerns and greater health care needs on the PedsQL.Poorer HR-QoL was observed in CHU-9D domains of "school work/homework" and "able to join in activities" and in EQ-5D-Y domains assessing "looking after myself" and "doing usual activities."

Floor and ceiling effects
Tables S1 displays floor and ceiling effects for the HR-QoL instruments.Ceiling effects were detected for the EQ VAS and the utility scores in the 4-15 year age group, as well as the School Functioning domain of the PedsQL 2-4 years.No other floor or ceiling effects were observed across measures.

Internal consistency and item-total correlations
Internal consistency and item-total correlations for each HR-QoL measure are displayed in Table 3.The CHU-9D showed adequate internal consistency, both in the validated age range (7-17 years), and in the larger sample of children aged 2-17 years.Item-total correlations were varied, with weak correlations for items assessing feeling 'pain' and strong correlations in items assessing "daily routine."The EQ-5D-Y revealed questionable internal consistency in both age groups.Item-total correlations showed poor correlations for items assessing feeling "worried, sad, or unhappy" and strong correlations in both age groups for items assessing "doing usual activities." For the PedsQL, there was adequate to excellent levels of internal consistency for all age groups.Strong item-total correlations were found within 13-18 year olds for items assessing "feeling sad" and low correlations were seen within the 2-4 year age group for items assessing "doing activities with other children".

Concurrent validity
Correlations between the HR-QoL measures and the VABS-III are displayed in Tables 4-8.Weak to strong correlations were found between these measures.There were moderate correlations between the VABS-III Communication domain with the PedsQL Social Functioning Scales in the 5-7 age group (see Table 4), and between Daily Living Skills domain on the VABS-III and Physical Functioning on the PedsQL in the 8-12 years and 13-18 age group (see Table 5).The CHU-9D and EQ-5D-Y revealed negative correlations with all the VABS-III domains, this inverse relationship is a result of the opposite direction in which these measures are scored.

Discriminant validity
The SDQ domain of "conduct problems," "emotional problems" and total scores were found to have moderate and strong correlations with the CHU-9D total score (Table 9).No strong correlations were found between the SDQ domain and with the EQ-5D-Y VAS or total scores (Table 10).Weak correlation between the SDQ "prosocial behavior" domain with the PedsQL total score was detected, but otherwise had no significant correlations (Table 11).Further testing of the statistical significance using Fisher's r-to-z transformation revealed negative z-scores indicating that the correlation coefficient between the CHU-9D and EQ-5D-Y total scores and SDQ total scores are weaker than their correlations with the VABS-3.This suggests that the CHU-9D and EQ-5D-Y share more similarity with the VABS-3 constructs and show discriminant validity the SDQ.The z-score for the PedsQL indicates that there is a stronger relationship with the correlations on the SDQ total score, suggesting that these measures have overlapping constructs (see Supplementary Table S2).
T A B L E 3 Internal consistency and item-total correlations of HR-QoL measures.

DISCUSSION
This study evaluated the psychometric properties of three commonly used and well-known HR-QoL measures in a cohort of children presenting to clinical developmental assessment services.The most common diagnoses received in these assessment services are autism spectrum disorders (Boulton et al., 2023).Overall, findings showed that the PedsQL and the CHU-9D showed good internal consistency, but the EQ-5D-Y showed lower levels of internal consistency.In regard to quality of life, the majority of children (>86%) met cut-off criteria for significant health concerns on the PedsQL.On the EQ-5D-Y and CHU-9D, children showed reduced HR-QoL, particularly on domains relating to school and homework, being able to join in activities, looking after self, and doing usual activities.Scores on the CHU-9D and EQ-5D-Y also showed reduced HR-QoL that was similar or worse to previous studies (Furber & Segal, 2015;Ten Hoopen et al., 2020) and in children with life disabilities (Khanna et al., 2013;Stevens & Ratcliffe, 2012).This research did, however, find that the CHU-9D and PedsQL correlated with different functioning domains assessed by the VABS-III.This may suggest that these measures assess different aspects of wellbeing.Our study shows that these measures may have potential to monitor HR-QoL in a broad neurodevelopment paediatric population and to identify areas of HR-QoL where children require support.
The CHU-9D is a promising HR-QoL tool to use in neurodiverse populations.The CHU-9D revealed adequate internal consistency and no significant floor or ceiling effects.A review by Rowen et al. (2021) found that psychometric evidence of the CHU-9D is based on a limited amount of research (N = 12), with no past study investigating its use in neurodevelopment populations.Overall, the average CHU-9D proxy-reported scores indicated poorer HR-QoL in "school work/homework," as well as "ability to join in activities."Therefore, these factors can result in a decreased HR-QoL (Webster et al., 2018).Children in this study displayed lower HR-QoL scores when compared to previous studies using the CHU-9D with cohorts who had long-standing disabilities (Furber & Segal, 2015;Stevens & Ratcliffe, 2012).The utility scores from this population do show more similarity to scores taken from mental health cohorts (Wolf et al., 2021).The current study is the first to apply the CHU-9D in a diverse neurodevelopmental population within clinical practice, a group whose functioning and HR-QoL needs are extensive but have remained largely unexplored with the CHU-9D up until now.
In contrast to the results of the CHU-9D, there was less psychometric support for use of the EQ-5D-Y.The measure showed poor internal consistency and some evidence of ceiling effects.In regard to its scores, the EQ-5D-Y proxy-reported scores indicated children experienced prominent difficulties in "Looking after myself" and "Doing usual activities."This is consistent with previous studies (Khanna et al., 2013) and reinforces that children with NDCs experience considerable difficulties doing daily tasks independently.Although there is a scarcity of research using the EQ-5D-Y in paediatric NDC populations, especially in autism, the EQ VAS scores are similar to the few that currently exist (Khanna et al., 2013;Ten Hoopen et al., 2020), and difficulties in these specific domains have been reported previously (Domellof et al., 2014).The difficulties reflected in the scores of the EQ-5D-Y suggests that the EQ-5D-Y can provide information on certain aspects of QoL, however, its poor internal consistency should be taken into consideration.
The PedsQL provided excellent internal consistency results but some evidence of ceiling effect in the 2-4 year age group on the School Functioning domain.Ceiling effects in the PedsQL have been reported previously (DeCarlo et al., 2020); (Varni et al., 2007) and suggests caregivers had difficulty providing comment on the School Functioning domain, likely due to children in this group not attending school yet.Furthermore, on the PedsQL children were found to show poorer HR-QoL in areas of psychosocial functioning, social functioning, and in older age groups, emotional functioning, indicating that increased understanding of one's functional limitations can impact HR-QoL.The caregiver reported scores were lower than those found in a previous study with a much smaller autistic cohort (N = 61) (Katsiana et al., 2020) and in a study of children with severe language impairment (Nicola & Watter, 2015).In addition, we found that the PedsQL showed stronger correlations to a measure of mental health than adaptive functioning.This study suggests that HR-QoL impacts may be much greater than what has previously been reported.Previous studies recruited cohorts specifically for a research study, however, this research cohort we report on here is integrated in clinical practices in public health settings.It may show more impacts given the cohort has been waiting for a diagnostic assessment for many years and has reported many disadvantages associated with the social determinants of health (Boulton et al., 2023).
In terms of both clinical and broader implications, the CHU-9D, EQ-5D-Y, and PedsQL have practical advantages and fulfill the easy to use, standardized, generic, quick and capacity-focused requirements outlined in the NDIS framework (see Figure 1).By successfully assessing areas of difficulty in QoL, clinical teams can implement tailored assessments to suit individual needs and improve long-term health outcomes.

Relationship with the VABS-III
The second part of this study was to explore how HR-QoL measures related to a well-established measure of functioning, the VABS-III.Overall, the CHU-9D exhibited moderate correlations with the VABS-III on domains assessing Daily Living Skills and Socialization.The CHU-9D domains (i.e., daily routine, able to join in activities), thematically correspond to the VABS-III Daily Living Skills and Socialization domains much better than the Communication and Motor domains.Similarly, the EQ-5D-Y had moderate correlations to the VABS-III Daily Living Skills subscale.The EQ-5D-Y There is an absence of previous literature reporting on the comparison of these HR-QoL measures to the VABS-III.The current study is the first to correlate these measures in a large neurodiverse population in a clinical environment.Unfortunately, we found no single measure of HR-QoL to correlate with every VABS-III domain.However, collectively, these HR-QoL measures address each domain on the VABS-III and have valuable properties to complement the VABS-III in assessment services.The CHU-9D, EQ-5D-Y, and PedsQL fulfill the easy to use, generic and quick requirements and importantly have a particular focus on wellbeing rather than diagnostic specificity.

Strength and limitations
These HR-QoL measures have practical advantages, being relatively quick to administer with a low response burden on participants.The current study has investigated the psychometric properties of the CHU-9D, EQ-5D-Y, and PedsQL in a relatively large cohort of children with diverse presentations of neurodevelopmental conditions that has not been previously reported on.These measures have been found to have adequate to excellent psychometric properties in a paediatric NDC cohort.Previous research in this area has tended to focus on particular neurodevelopmental conditions, such as intellectual disabilities (Viecili & Weiss, 2015), ASD (Viecili & Weiss, 2015), and in child mental health cohorts (Furber & Segal, 2015).The present research utilizes data to provide insight into how these measures perform in children with diverse presentations of NDCs, who will likely be seeking NDIS funding following assessment.
The results of this study can be generalized to other assessment settings in countries outside of Australia.
The current study is not without its limitations.First, there is debate surrounding the challenges when assessing HR-QoL in children.This includes research indicating that proxy reports and child self-reports are inconsistent, with a meta-analysis finding correlations between proxy and selfreports to be 0.22 (Achenbach et al., 1987).Previous studies have identified this within subjective domains such as emotional functioning, compared to observable domains such as physical activity (White-Koning et al., 2005).This disparity may be explained by biases or lack of knowledge or understanding of the child's condition (DeCarlo et al., 2020).In addition, recall bias is a potential issue in relation to parents or caregivers recalling past events inaccurately, potentially hindering the reliability of the responses gathered in the measures.Additionally the interplay of factors such as parental mental health are worthy of consideration (Chuang et al., 2014).Future studies may wish to compare the results we present here with those of neurotypical children to identify different response patterns to these scales as well as address any differences in proxy reporting.In terms of the study methodology, we acknowledge that the CHU-9D and EQ-5D-Y have not been validated across all of the age ranges in this study.However, comparable results were observed when we looked separately at the subset of participants within the validated age ranges, suggesting that these measures may have applicability for a wider age group.We acknowledge that other psychometric properties, such as structural validity, test-retest reliability and the overall factor structure of these measures were not assessed.Future research may wish to use item response analysis to conduct a more detailed analysis to investigate further the internal consistency and discriminant validity results provided in this study.

CONCLUSIONS
These brief, easy to access HR-QoL measures can provide a broader overview of wellbeing to inform assessment.This study has selected some of the most common HR-QoL instruments and tested their utility to determine how they are associated with the current gold-standard measure of adaptive functioning, the VABS-III.Overall, it has shown that the PedsQL and CHU-9D were reliable measures and each HR-QoL measure related to different subscales of the VABS-III.Thus, these measures can be included in a clinician's toolkit as an additional source of information regarding children with neurodevelopmental conditions as well as their potential use as treatment outcomes.
T A B L E 9 *p < 0.05; **p < 0.001.T A B L E 1 0 Authors N.P., K.B., N.S., A.J.G designed the research study, authors, N.P., A.H., N.O., and N.S. supported data collection, authors K.B and A.G. conducted analysis and drafted the first manuscript, Authors RR and KH contributed to analysis and interpretation.All authors reviewed drafts and contributed to the final version of the submitted manuscript.Discriminant validity of the PedsQL with the SDQ.
T A B L E 1 1