Comparison of the EQ-HWB and EQ-HWB-S With Other Preference-Based Measures Among United States Informal Caregivers.

OBJECTIVES: Several measures have been used or developed to capture the health and well-being of caregivers, including the EQ Health and Well-being (EQ-HWB) and its short form, EQ-HWB-S. This study aimed to evaluate the psychometric properties and construct validity of the EQ-HWB/EQ-HWB-S in a US caregiver population. METHODS: A cross-sectional survey was conducted involving 504 caregivers. Eligible participants were 18+ years old, provided unpaid care to a relative/friend aged 18+ in the past 6 months, and spent on average of at least 1 hour per week caregiving. Survey included the following measures: EQ-HWB, Adult Social Care Outcomes Toolkit for Carers-Carer, CarerQol, and EQ-5D-5L. Psychometric properties were assessed using response distributions, floor/ceiling effects, Spearman's correlation for convergent validity, and effect sizes (ES) for known-group validity based on caregiving situations and intensity. RESULTS: The average age of caregivers was 49.2 (SD = 15.4), with 57.5% being female. More than half (54.4%) reported high caregiving intensity, and 68.3% lived with the care recipient. The EQ-HWB-S index showed a strong positive correlation with the EQ-5D-5L (rs = 0.72), Adult Social Care Outcomes Toolkit for Carers (rs = 0.54), and CarerQol (rs = 0.54) indices. Notably, the EQ-HWB-S index showed the largest ES among measures in differentiating caregiving scenarios with a large ES for caregiver's general health (d = 1.00) and small ES for caregiving intensity (d = 0.39). CONCLUSIONS: Results support construct validity of EQ-HWB and EQ-HWB-S as measures for assessing health and well-being of adult informal caregivers in comparison with other validated instruments. Differing levels of known-group validity across anchors emphasize the importance of selecting appropriate measures for caregivers, depending on research question and/or intervention aims.

Putting a Dollar Value on Informal Care Time Provided to People Living With Dementia: A Discrete Choice Experiment.

OBJECTIVES: Informal care represents a significant cost driver in dementia but monetizing informal care hours to inform cost-of-illness or economic evaluation studies remains a challenge. This study aimed to use a discrete choice experiment to estimate the value of informal care time provided to people with dementia in Australia accounting for positive and negative impacts of caregiving. METHODS: Attributes and levels were derived from a literature review, interviews with carers, and advice received from an advisory group. Attributes included 4 positive and negative caregiving experiences, in addition to "hours of care provided" and the "monetary compensation from the government." A D-efficient design was constructed with 2 generic alternatives that represented hypothetical informal caregiving situations. The discrete choice experiment survey was administered online to a representative sample of the Australian general population and a group of informal carers of people with dementia. The willingness to accept estimates were calculated for the 2 samples separately using the mixed logit model in the willingness to pay space. RESULTS: Based on 700 respondents included in the analysis (n = 488 general public, n = 212 informal carers), the mean willingness to accept for an additional hour of informal care, corrected for the positive and negative impacts of informal care, was $21 (95% CI 18-23) for the general public and $20 (95% CI 16-25) for the informal carers sample. CONCLUSION: The estimates generated in this study can be used to inform future cost-of-illness studies and economic evaluations, ensuring that informal care time is considered in future policy and funding decisions.

How Well Does the EQ-5D-Y-5L Describe Children With Intellectual Disability?: "There's a Lot More to My Child Than That She Can't Wash or Dress Herself."

OBJECTIVES: The EQ-5D-5L is a generic health utility instrument for measuring health-related quality of life (HRQoL), with self-report and proxy report versions for children (EQ-5D-Y-5L). Children with intellectual disability (ID) are a heterogeneous population whose impairments and comorbidities place them at risk of poor HRQoL. This study aimed to describe the content validity and suitability for children with ID of a proxy report version of the EQ-5D-Y-5L as seen by their caregivers. METHODS: A proxy report EQ-5D-Y-5L was administered to caregivers of children with ID. Using cognitive think-aloud interviewing, participants were encouraged to provide the reasoning for their choices, assess the questions' relevance, comprehensibility, and comprehensiveness, and comment on the tool's strengths and weaknesses. Qualitative content analysis used both directed (deductive) and conventional (inductive) methods. RESULTS: There were 28 interviews with 30 caregivers of children with ID (aged 8-22 years, 17 boys, with autism spectrum disorder, cerebral palsy, Down syndrome, and rare genetic disorders). The EQ-5D-Y-5L was considered clear, concise, and largely relevant, but insufficiently comprehensive for this population. Interviewees sought clarification of the definition of HRQoL, whether it included unchanging impairments (vs fluctuating health states), and what basis of comparison to use (child or peer). Many interviewees suggested inclusion of questions for other domains, including communication and social engagement, equipment and human supports required, and a wider range of mental health questions. CONCLUSIONS: The study suggests that further work is required to ensure accurate responses to the EQ-5D-Y-5L from caregivers of children with ID and to describe these children adequately.

The Ceiling Effects of EQ-5D-3L and 5L in General Population Health Surveys: A Systematic Review and Meta-Analysis.

OBJECTIVES: This review aims to examine the ceiling effects of EQ-5D-3L (3L) and EQ-5D-5L (5L) in general adult populations and identify the factors influencing these effects. METHODS: We searched 8 databases for observational studies published in English from inception to 24 July 2023. Ceiling effects were calculated by dividing the number of participants reporting full health at dimension or profile level by the total sample size. Subgroup analysis and meta-regression using the metafor package in R software were performed. RESULTS: We identified 94 studies from 70 articles, including 4 543 647 adults across 37 countries. The global pooled proportion of individuals reporting full health ("11111") was 56% (95% CI 51%-62%) for 3L and 49% (95% CI 44%-54%) for 5L. The self-care dimension showed the highest ceiling effects (3L: 97%; 5L: 94%), whereas pain/discomfort had the lowest (3L: 69%; 5L: 60%). The ceiling effects in East/South-East Asia were higher than in Europe by 25% (95% CI 18%-32%) in 3L and 9% (95% CI -2%-20%) in 5L. Adjusting for mean age and proportion of males, significant regional differences persisted in the overall profile level of 3L, in all 3L dimensions (except for self-care), and 5L dimensions (except for pain/discomfort and anxiety/depression). CONCLUSIONS: This review highlights significant ceiling effects in the EQ-5D, especially in Asian populations. The 5L version exhibited fewer ceiling effects than the 3L, indicating its superiority for general population surveys. Further research is crucial to understand the disparities in self-reported health outcomes between Asians and other populations.

Psychometric Properties of the EQ-5D-Y-5L for Children With Intellectual Disability.

OBJECTIVES: The EQ-5D-Y-5L is a generic preference-based measure of health-related quality of life for children. This study aimed to describe the distributional properties, test-retest reliability, and convergent validity of the EQ-5D-Y-5L in children with intellectual disability (ID). METHODS: Caregivers of children with ID (aged 4 to 18 years) completed an online survey, including a proxy-report EQ-5D-Y-5L, the Quality-of-life Inventory-Disability, and disability-appropriate measures corresponding to the EQ-5D dimensions: mobility, self-care (SC), usual activities (UA), pain/discomfort (PD), and worry/sadness/unhappiness. Twenty-one participants repeated the EQ-5D-Y-5L a few weeks later. Test-retest reliability was computed using weighted kappa and intraclass correlation coefficients, and convergent validity using Spearman's and Pearson's correlation coefficients. RESULTS: Caregivers of 234 children completed the survey, with <1% missing values. Only 1.7% reported "no problems" on all dimensions (11111). The dimensions with the lowest percentage of "no problems" were SC and UA (both 8%). Test-retest reliability coefficients were fair to substantial for 4 dimensions (weighted kappa .30 to .79) but low for PD and overall health, as measured by the visual analog scale (EQ-VAS). Convergent validity was strong (Spearman's correlation .65 to .87) for mobility, SC, and PD; moderate to strong for worry/sadness/unhappiness (.47 to .60) and the EQ-VAS (Pearson's correlation .49); and weak to moderate for UA (.21 to .52). CONCLUSIONS: Convergent validity was generally good; test-retest reliability varied. Children with ID had lower scores on SC and UA than other populations, and their EQ-VAS could fluctuate greatly, indicating poorer and less stable health-related quality of life.

Understanding caregiver burden from multiple perspectives: dyadic agreement between caregiver and care recipient.

PURPOSE: Caregiver burden (CB) is typically self-assessed by caregivers. However, an emerging concept is assessment of CB by the recipients of care, i.e., the patient. The specific objectives are (1) to assess the level of agreement between care recipients' and caregivers' view on CB, across financial, physical, emotional, and social domains; (2) to explore two care recipient perspectives: their self-perceived burden (CR-SPB), and their interpretation of the caregiver's view (Proxy-CB). METHODS: Data were collected from 504 caregiver-care recipient dyads in the U.S. using an online Qualtrics panel. The survey assessed caregiver burden using CarerQol and newly developed items. The level of agreement between responses was quantified using weighted kappa (κ) coefficients for individual items and intraclass correlation coefficients (ICC) for index/summary scores. RESULTS: The average age of caregivers was 49.2 years, and 62.7 years for care recipients. Dyads most commonly consisted of spouses/partners (34.5%); 68.3% lived together. Proxy-CB aligned more closely with caregiver's view, with moderate to substantial agreement across CB domains (from κ = 0.48 for emotional to κ = 0.66 for financial). In the same perspective, the CarerQol-7D Index showed moderate agreement (ICC = 0.58) and the summary score of CB items substantial agreement (ICC = 0.76). Care recipients generally overestimated  CB in the Proxy-CB perspective, while they underestimated it in the CR-SPB perspective. CONCLUSION: Results demonstrate there is a difference between perspectives. Strong agreement in Proxy-CB perspective suggests that care recipients can potentially substitute for caregivers depending on the domain. CR-SPB agrees less with caregivers and may provide complementary information.

Scoring the EQ-HWB-S: can we do it without value sets? A non-parametric item response theory analysis.

BACKGROUND: Only one pilot value set (UK) is currently available for the EQ Health and Wellbeing Instrument short version (EQ-HWB-S). As an alternative to preference-weighted scoring, we examined whether a level summary score (LSS) is appropriate for the EQ-HWB-S using Mokken scaling analyses. METHODS: Data from patients, carers and the general population collected during the developmental phase of the EQ-HWB-S in Australia, US and UK were used, noting 3 of 9 items have since undergone revision. EQ-HWB-S data fit was examined using R package Mokken scaling's monotone homogeneity model, utilizing the automated item selection procedure (AISP) as well as Loevinger's scaling coefficients for items and the scale (HS). Manifest monotonicity was assessed by examining whether the cumulative probability for responses at or above each response level did not decrease across the summary score. RESULTS: EQ-HWB-S data were available for 3340 respondents: US = 903, Australia = 514 and UK = 1923. Mean age was 50 ± 18 and 1841 (55%) were female. AISP placed all 9 items of the EQ-HWB-S on a single scale when the lower bound was set to < 0.448. Strong scalability (HS = 0.561) was found for the EQ-HWB-S as a single scale. Stronger scales were formed by separating the psychosocial items (n = 6, HS = 0.683) and physical sensation items (n = 3, HS = 0.713). No violations of monotonicity were found except for the items mobility and daily activities for the subgroups with long-term conditions and UK subjects, respectively. DISCUSSION: As EQ-HWB-S items formed a strong scale and subscales based on Mokken analysis, LSS is a promising weighting-free approach to scoring.

Eliciting a value set for the Swedish Capability-Adjusted Life Years instrument (CALY-SWE).

PURPOSE: Our aim was to elicit a value set for Capability-Adjusted Life Years Sweden (CALY-SWE); a capability-grounded quality of life instrument intended for use in economic evaluations of social interventions with broad consequences beyond health. METHODS: Building on methods commonly used in the quality-adjusted life years EQ-5D context, we collected time-trade off (TTO) and discrete choice experiment (DCE) data through an online survey from a general population sample of 1697 Swedish participants. We assessed data quality using a score based on the severity of inconsistencies. For generating the value set, we compared different model features, including hybrid modeling of DCE and TTO versus TTO data only, censoring of TTO answers, varying intercept, and accommodating for heteroskedasticity. We also assessed the models' DCE logit fidelity to measure agreement with potentially less-biased DCE data. To anchor the best capability state to 1 on the 0 to 1 scale, we included a multiplicative scaling factor. RESULTS: We excluded 20% of the TTO answers of participants with the largest inconsistencies to improve data quality. A hybrid model with an anchor scale and censoring was chosen to generate the value set; models with heteroskedasticity considerations or individually varying intercepts did not offer substantial improvement. The lowest capability weight was 0.114. Health, social relations, and finance and housing attributes contributed the largest capability gains, followed by occupation, security, and political and civil rights. CONCLUSION: We elicited a value set for CALY-SWE for use in economic evaluations of interventions with broad social consequences.

Development of algorithms for estimating the Child Health Utility 9D from Caregiver Priorities and Child Health Index of Life with Disability.

PURPOSE: The primary aim was to determine Child Health Utility 9D (CHU9D) utilities from the Caregiver Priorities and Child Health Index of Life with Disabilities (CPCHILD) for non-ambulatory children with cerebral palsy (CP). METHODS: One hundred and eight surveys completed by Australian parents/caregivers of children with CP were analysed. Spearman's coefficients were used to investigate the correlations between the two instruments. Ordinary least square, robust MM-estimator, and generalised linear models (GLM) with four combinations of families and links were developed to estimate CHU9D utilities from either the CPCHILD total score or CPCHILD domains scores. Internal validation was performed using 5-fold cross-validation and random sampling validation. The best performing algorithms were identified based on mean absolute error (MAE), concordance correlation coefficient (CCC), and the difference between predicted and observed means of CHU9D. RESULTS: Moderate correlations (ρ 0.4-0.6) were observed between domains of the CHU9D and CPCHILD instruments. The best performing algorithm when considering the CPCHILD total score was a generalised linear regression (GLM) Gamma family and logit link (MAE = 0.156, CCC = 0.508). Additionally, the GLM Gamma family logit link using CPCHILD comfort and emotion, quality of life, and health domain scores also performed well (MAE = 0.152, CCC = 0.552). CONCLUSION: This study established algorithms for estimating CHU9D utilities from CPCHILD scores for non-ambulatory children with CP. The determined algorithms can be valuable for estimating quality-adjusted life years for cost-utility analysis when only the CPCHILD instrument is available. However, further studies with larger sample sizes and external validation are recommended to validate these findings.

How do common conditions impact health-related quality of life for children? Providing guidance for validating pediatric preference-based measures.

BACKGROUND: There is increasing interest in the validation of pediatric preference-based health-related quality of life measurement instruments. It is critical that children with various degrees of health-related quality of life (HRQoL) impact are included in validation studies. To inform patient sample selection for validation studies from a pragmatic perspective, this study explored HRQoL impairments between known-groups and HRQoL changes over time across 27 common chronic child health conditions and identified conditions with the largest impact on HRQoL. METHODS: The health dimensions of two common preference-based HRQoL measures, the EQ-5D-Y and CHU9D, were constructed using Pediatric Quality of Life Inventory items that overlap conceptually. Data was from the Longitudinal Study of Australian Children, a nationally representative sample with over 10,000 children at baseline. Seven waves of data were included for the analysis, with child age ranging from 2 to18 years. Impacts to specific health dimensions and overall HRQoL between those having a specific condition versus not were compared using linear mixed effects models. HRQoL changes over time were obtained by calculating the HRQoL differences between two consecutive time points, grouped by "Improved" and "Worsened" health status. Comparison among various health conditions and different age groups (2-4 years, 5-12 years and 13-18 years) were made. RESULTS: Conditions with the largest statistically significant total HRQoL impairments of having a specific condition compared with not having the condition were recurrent chest pain, autism, epilepsy, anxiety/depression, irritable bowel, recurrent back pain, recurrent abdominal pain, and attention deficit hyperactivity disorder (ADHD) for the total sample (2-18 years). Conditions with largest HRQoL improvement over time were anxiety/depression, ADHD, autism, bone/joint/muscle problem, recurrent abdominal pain, recurrent pain in other part, frequent headache, diarrhea and day-wetting. The dimensions included in EQ-5D-Y and CHU9D can generally reflect HRQoL differences and changes. The HRQoL impacts to specific health dimensions differed by condition in the expected direction. The conditions with largest HRQoL impacts differed by age group. CONCLUSIONS: The conditions with largest HRQoL impact were identified. This information is likely to be valuable for recruiting patient samples when validating pediatric preference-based HRQoL instruments pragmatically.

Examining the psychometric properties of a split version of the EQ-5D-5L anxiety/depression dimension in patients with anxiety and/or depression.

PURPOSE: This study explored differences in self-reported responses and the psychometric performance of the composite EQ-5D-5L anxiety/depression (A/D) dimension compared with a split version of the dimension where 'anxiety' and 'depression' are measured separately. METHODS: People with anxiety and/or depression who visited the Amanuel Mental Specialized Hospital in Ethiopia completed the standard EQ-5D-5L with the added subdimensions. Correlation analysis was used to examine convergent validity with validated measures of depression (PHQ-9) and anxiety (GAD-7), while ANOVA was used to assess known-groups' validity. Agreement between ratings for composite and split dimensions was compared using percent agreement and Cohen's Kappa, while the proportion of 'no problems' reports was compared using the chi-square test. Discriminatory power analysis was undertaken using the Shannon index (H') and Shannon Evenness index (J'). Open-ended questions explored participants' preferences. RESULTS: Of the 462 respondents, 30.5% reported no problems with the composite A/D, while 13.2% reported no problems on both subdimensions. Agreement between ratings for composite and split dimensions was highest for respondents with comorbid anxiety and depression. The depression subdimension had higher correlation with PHQ-9 (r = 0.53) and GAD-7 (r = 0.33) than the composite A/D dimension (r = 0.36 and r = 0.28, respectively). The split subdimensions and composite A/D could adequately differentiate respondents based on their severity of anxiety or depression. Slightly better informativity was observed in EQ-4D-5L + anxiety (H' = 5.4; J' = 0.47) and EQ-4D-5L + depression (H' = 5.31; J' = 0.46) than EQ-5D-5L (H' = 5.19; J' = 0.45). CONCLUSIONS: Adopting two subdimensions within the EQ-5D-5L tool appears to perform slightly better than the standard EQ-5D-5L.

What is measured by the composite, single-item pain/discomfort dimension of the EQ-5D-5L? An exploratory analysis.

PURPOSE: This study examines the EQ-5D-5L pain/discomfort dimension by drawing comparisons with five other pain and discomfort items (pain severity, discomfort severity, pain frequency, discomfort frequency and pain interference) collected in the Australian psychometric study for the EQ Health and Wellbeing instrument. METHODS: Participants, recruited via a market research company, completed an online survey. Methods of analyses included the assessment of descriptive statistics, variation in reporting patterns using chi-square tests and cross-tabulations, correlation analyses, ordered univariate logistic regression, and discriminatory power analyses (Shannon index (H') and Shannon Evenness index (J')). RESULTS: Survey data from 514 participants were used. Compared with EQ-5D-5L pain/discomfort, there was a higher proportion of respondents reporting some level of impairment on at least one of the pain severity and discomfort severity items (74% versus 81%). Correlation with EQ-5D-5L pain/discomfort was strongest for pain severity (r = 0.83) and weakest for discomfort frequency (r = 0.41); the same inferences were drawn for predictive ability. Adding any additional pain or discomfort items to the EQ-5D-5L increased the absolute informativity (H') but not the relative informativity (J'). When replacing EQ-5D-5L pain/discomfort with separate pain and/or discomfort items - i.e., adding items to a modified 'EQ-4D-5L'-absolute informativity increased, while relative informativity increased only when pain interference and frequency-related items (independently or in combination) were added. CONCLUSION: The EQ-5D-5L pain/discomfort dimension captures aspects of pain more than aspects of discomfort. Potential reasons include the absence of descriptors or because pain is mentioned first in the composite item.

Feasibility of self-reported health related quality of life assessment with older people in residential care: insights from the application of eye tracking technology.

PURPOSE: Increasingly there are calls to routinely assess the health-related quality of life (HRQoL) of older people receiving aged care services, however the high prevalence of dementia and cognitive impairment remains a challenge to implementation. Eye-tracking technology facilitates detailed assessment of engagement and comprehension of visual stimuli, and may be useful in flagging individuals and populations who cannot reliably self-complete HRQoL instruments. The aim of this study was to apply eye-tracking technology to provide insights into self-reporting of HRQoL among older people in residential care with and without cognitive impairment. METHODS: Residents (n = 41), recruited based on one of three cognition subgroups (no, mild, or moderate cognitive impairment), completed the EQ-5D-5L on a computer with eye tracking technology embedded. Number and length of fixations (i.e., eye gaze in seconds) for key components of the EQ-5D-5L descriptive system were calculated. RESULTS: For all dimensions, participants with no cognitive impairment fixated for longer on the Area of Interest (AOI) for the response option they finally chose, relative to those with mild or moderate cognitive impairment. Participants with cognitive impairment followed similar fixation patterns to those without. There was some evidence that participants with cognitive impairment took longer to complete and spent relatively less time attending to the relevant AOIs, but these differences did not reach statistical significance generally. CONCLUSIONS: This exploratory study applying eye tracking technology provides novel insights and evidence of the feasibility of self-reported HRQoL assessments in older people in aged care settings where cognitive impairment and dementia are highly prevalent.

Understanding the measurement relationship between EQ-5D-5L, PROMIS-29 and PROPr.

PURPOSE: Many generic patient-reported instruments are available for the measurement of health outcomes, including EQ-5D-5L, and the Patient-Reported Outcome Measurement Information System (PROMIS). Assessing their measurement characteristics informs users about the consistency between, and limits of, evidence produced. The aim was to assess the measurement relationship between the EQ-5D-5L descriptive system and value sets, the PROMIS-29 and PROPr (PROMIS value set). METHODS: Data were extracted from a cross-sectional survey administering measures of quality of life online in Australia. Descriptive analysis, agreement and construct validity assessment methods were used to compare instruments at the item, domain and value set level. RESULTS: In total, 794 Australians completed the survey. Convergent validity analysis found that similar dimensions across instruments were highly correlated (> 0.50), but the PROMIS-29 assesses additional health concepts not explicitly covered by EQ-5D (sleep and fatigue). Known-group assessment found that EQ-5D-5L and PROPr were able to detect those with and without a condition (ES range 0.78-0.83) but PROPr could more precisely detect differing levels of self-reported health. Both instruments were sensitive to differences in levels of pain. DISCUSSION: There is some consistency in what the EQ-5D-5L, PROMIS-29 and PROPr measure. Differences between value set characteristics can be linked to differences what is measured and the valuation approaches used. This has implications for the use of each in assessing health outcomes, and the results can inform decisions about which instrument should be used in which context.

Developing an Australian utility value set for MacNew-7D health states.

BACKGROUND: A new preference-based measure (MacNew-7D) has recently been developed to allow condition-specific data to be used to capture the quality of life in health economic evaluations in cardiology; however, a general population value set has not yet been developed. This study developed a population utility value set for the MacNew-7D heart disease-specific instrument. METHODS: The discrete choice experiments (DCE) technique was chosen as the preference-elicitation method. The DCE asked respondents to compare two options and to state their preferences. The survey was conducted using an online panel of respondents, with quota sampling using age groups, sex and jurisdictions to achieve representativeness of the Australian population. The total design consisted of 200 choice sets, of which each respondent answered eight. Additionally, each respondent answered two quality control choice sets. The best-fitting models were selected on the basis of consistency, parsimony, and goodness of fit. RESULTS: In total, 1903 respondents were included in the analyses. The MacNew-7D utility value set ranged from -0.4456 to 1.000 for health states defined by the classification system. The best-fitting model retained all levels for five dimensions and collapsed one adjacent level for the other two dimensions. Findings were robust to sensitivity analyses related to the inclusion or exclusion of dominancy and repeat tasks. CONCLUSION: Findings indicated that the MacNew-7D utility value set is likely suitable for estimating quality-adjusted life years derived from the MacNew heart disease health-related quality-of-life questionnaire. This value set was derived from an Australian population-based sample and may not be generalisable to dissimilar populations.

Deciding Between SF-6Dv2 Health States: A Think-Aloud Study of Decision-Making Strategies Used in Discrete Choice Experiments.

OBJECTIVE: This study aimed to gain insight into decision-making strategies individuals used when evaluating pairs of SF-6Dv2 health states in discrete choice experiments (DCEs). METHODS: This qualitative, cross-sectional, noninterventional study asked participants to use a think-aloud approach to compare SF-6Dv2 health states in DCEs. Thematic analysis focused on comprehension and cognitive strategies used to compare health states and make decisions. RESULTS: Participants (N = 40) used 3 main strategies when completing DCEs: (1) trading, (2) reinterpretation, and (3) relying on previous experience. Trading was the most common strategy, used by everyone at least once, and involved prioritizing key attributes, such as preferring a health state with significant depression but no bodily pain. Reinterpretation was used by 17 participants and involved reconstructing health states by changing underlying assumptions (eg, rationalizing selecting a health state with significant pain because they could take pain medications). Finally, some (n = 13) relied on previous experience when making decisions on some choice tasks. Participants with experience dealing with pain, for instance, prioritized health states with the least impact in this dimension. CONCLUSIONS: Qualitatively evaluating the decision-making strategies used in DCEs allows researchers to evaluate whether the tasks and attributes are interpreted accurately. The findings from this study add to the understanding of the generation of SF-6Dv2 health utility weights and the validity of these weights (e.g., reinterpreting health states could undermine the validity of DCEs and utility weights), and the overall usefulness of the SF-6Dv2. The methodology described in this study can and should be carried forth in valuing other health utility measures, not just the SF-6Dv2.

Generation, Selection, and Face Validation of Items for a New Generic Measure of Quality of Life: The EQ-HWB.

OBJECTIVES: This article aims to describe the generation and selection of items (stage 2) and face validation (stage 3) of a large international (multilingual) project to develop a new generic measure, the EQ-HWB (EQ Health and Wellbeing), for use in economic evaluation across health, social care, and public health to estimate quality-adjusted life-years. METHODS: Items from commonly used generic, carer, social care, and mental health quality of life measures were mapped onto domains or subdomains identified from a literature review. Potential terms and items were reviewed and refined to ensure coverage of the construct of the domains/subdomain (stage 2). Input on the potential item pool, response options, and recall period was sought from 3 key stakeholder groups. The pool of candidate items was tested in qualitative interviews with potential future users in an international face validation study (stage 3). RESULTS: Stage 2 resulted in the generation of 687 items. Predetermined selection criteria were applied by the research team resulting in 598 items being dropped, leaving 89 items that were reviewed by key stakeholder groups. Face validation (stage 3) tested 97 draft items and 4 response scales. A total of 47 items were retained and 14 were modified, whereas 3 were added to the candidate pool of items. This resulted in a 64-item set. CONCLUSIONS: This international multiculture, multilingual study with a common methodology identified many items that performed well across all countries. These were taken to the psychometric testing along with modified and new items for the EQ-HWB.

Developing a New Generic Health and Wellbeing Measure: Psychometric Survey Results for the EQ-HWB.

OBJECTIVES: The development of measures such as the EQ-HWB (EQ Health and Wellbeing) requires selection of items. This study explored the psychometric performance of candidate items, testing their validity in patients, social carer users, and carers. METHODS: Article and online surveys that included candidate items (N = 64) were conducted in Argentina, Australia, China, Germany, United Kingdom, and the United States. Psychometric assessment on missing data, response distributions, and known group differences was undertaken. Dimensionality was explored using exploratory and confirmatory factor analysis. Poorly fitting items were identified using information functions, and the function of each response category was assessed using category characteristic curves from item response theory (IRT) models. Differential item functioning was tested across key subgroups. RESULTS: There were 4879 respondents (Argentina = 508, Australia = 514, China = 497, Germany = 502, United Kingdom = 1955, United States = 903). Where missing data were allowed, it was low (UK article survey 2.3%; US survey 0.6%). Most items had responses distributed across all levels. Most items could discriminate between groups with known health conditions with moderate to large effect sizes. Items were less able to discriminate across carers. Factor analysis found positive and negative measurement factors alongside the constructs of interest. For most of the countries apart from China, the confirmatory factor analysis model had good fit with some minor modifications. IRT indicated that most items had well-functioning response categories but there was some evidence of differential item functioning in many items. CONCLUSIONS: Items performed well in classical psychometric testing and IRT. This large 6-country collaboration provided evidence to inform item selection for the EQ-HWB measure.

The EQ-HWB: Overview of the Development of a Measure of Health and Wellbeing and Key Results.

OBJECTIVES: Existing measures for estimating quality-adjusted life-years are mostly limited to health-related quality of life. This article presents an overview of the development the EQ-HWB (EQ Health and Wellbeing), which is a measure that encompasses health and wellbeing. METHODS: Stages: (1) Establishing domains through reviews of the qualitative literature informed by a conceptual framework. (2) Generation and selection of items to cover the domains. (3) Face validation of these items through qualitative interviews with 168 patients, social care users, general population, and carers across 6 countries (Argentina, Australia, China, Germany, United Kingdom, United States). (4) Extensive psychometric testing of candidate items (using classical, factor analysis, and item response theory methods) on > 4000 respondents in the 6 countries. Stakeholders were consulted throughout. RESULTS: A total of 32 subdomains grouped into 7 high-level domains were identified from the qualitative literature and 97 items generated to cover them. Face validation eliminated 36 items, modified 14, and added 3. Psychometric testing of 64 items found little difference in missing data or problems with response distribution, the conceptual model was confirmed except in China, and most items performed well in the item response theory in all countries. Evidence was presented to stakeholders in 2 rounds of consultation to inform the final selection of items for the EQ-HWB (25-item) and the short version of EQ-HWB (9-items). CONCLUSIONS: EQ-HWB measures have been developed internationally for evaluating interventions in health, public health, and social care including the impact on patients, social care users, and carers.

Evidence on the relationship between PROMIS-29 and EQ-5D: a literature review.

PURPOSE: EQ-5D and PROMIS-29 are both concise, generic measures of patient-reported outcomes accompanied by preference weights that allow the estimation of quality-adjusted life years (QALYs). Both instruments are candidates for use in economic evaluation. However, they have different features in terms of the domains selected to measure respondents' self-perceived health and the characteristics of (and methods used to obtain) the preference weights. It is important to understand the relationship between the instruments and the implications of choosing either for the evidence used in decision-making. This literature review aimed to synthesise existing evidence on the relationship between PROMIS-29 (and measures based on it, such as PROMIS-29+2) and EQ-5D (both EQ-5D-3L and EQ-5D-5L). METHODS: A literature review was conducted in PubMed and Web of Science to identify studies investigating the relationship between PROMIS-29 and EQ-5D-based instruments. RESULTS: The literature search identified 95 unique studies, of which nine studies met the inclusion criteria, i.e. compared both instruments. Six studies examined the relationship between PROMIS-29 and EQ-5D-5L. Three main types of relationship have been examined in the nine studies: (a) comparing PROMIS-29 and EQ-5D as descriptive systems; (b) mapping PROMIS-29 domains to EQ-5D utilities; and (c) comparing and transforming PROMIS-29 utilities to EQ-5D utilities. CONCLUSION: This review has highlighted the lack of evidence regarding the relationship between PROMIS-29 and EQ-5D. The impact of choosing either instrument on the evidence used in cost-effectiveness analysis is currently unclear. Further research is needed to understand the relationship between the two instruments.