Advertisement
Value in Health
ISPOR
Advanced searchSave search
Patient-Reported Outcomes| Volume 24, ISSUE 9, P1308-1318, September 2021

Download started.

Ok

New Zealand Population Norms for the EQ-5D-5L Constructed From the Personal Value Sets of Participants in a National Survey

Open AccessPublished:June 26, 2021DOI:https://doi.org/10.1016/j.jval.2021.04.1280
New Zealand Population Norms for the EQ-5D-5L Constructed From the Personal Value Sets of Participants in a National Survey
Previous ArticlePatient-Reported Outcomes After Treatment Discontinuation: Commercial Clinical Trial Data From Four Cancer Types
Next ArticleReference Values and Psychometric Properties of the Quality of Life After Traumatic Brain Injury-Overall Scale in Italy, The Netherlands, and the United Kingdom

      Abstract

      Objectives

      To derive New Zealand (NZ) population norms for the EQ-5D-5L and to examine the association between participants’ sociodemographic characteristics and their health-related quality of life.

      Methods

      Data from the 2018 NZ EQ-5D-5L valuation study (n = 2468) were used. Each participant’s 5-digit profile was converted to a single utility value using their personal value set. The profiles, mean utility values, and mean EuroQol visual analog scale (EQ-VAS) scores were summarized by dimension and disaggregated by age group and gender. Multivariable logistic and Tobit regressions were used to investigate the association between participants’ sociodemographic characteristics and the EQ-5D-5L dimensions, utility values, and EQ-VAS scores.

      Results

      The mean utility value was 0.847 and the mean EQ-VAS score was 74.8. Of the 3125 possible EQ-5D-5L profiles, 25 profiles represented the current health status of the majority of participants (78%). The odds of having problems with anxiety or depression was greatest for people aged 18 to 24 years and decreased with age. People with a long-term disability or chronic illness had greater odds of problems on all dimensions and lower (poorer) utility values and EQ-VAS scores. Age, ethnicity, employment status, long-term disability, and chronic illness were associated with utility.

      Conclusion

      EQ-5D-5L population norms were derived for the NZ population using the personal value sets of 2468 participants. Consistent with other countries’ population norms, EQ-5D-5L utility values and EQ-VAS scores were associated with age, employment status, long-term disability, and chronic illness. These norms will support resource allocation decision making and help in understanding the health-related quality of life of the NZ population.

      Keywords

      Introduction

      The EQ-5D, which represents health-related quality of life (HRQoL) on 5 generic dimensions, is the world’s most widely used health descriptive system.
      • Brazier J.
      • Ratcliffe J.
      • Saloman J.
      • Tsuchiya A.
      Measuring and Valuing Health Benefits for Economic Evaluation.
      Oxford University Press, Oxford, United Kingdom2017
      It is used for clinical and population health studies and economic evaluations internationally,
      • Brazier J.
      • Ratcliffe J.
      • Saloman J.
      • Tsuchiya A.
      Measuring and Valuing Health Benefits for Economic Evaluation.
      Oxford University Press, Oxford, United Kingdom2017
      ,
      • Devlin N.
      • Parkin D.
      • Janssen B.
      Methods for Analysing and Reporting EQ-5D Data.
      Springer Nature, Cham, Switzerland2020
      including by researchers and policy makers in New Zealand (NZ). Since 1999, NZ’s Pharmaceutical Management Agency has used the original 3-level version of the system, the EQ-5D-3L, for cost-utility analysis to decide which medicines and devices to buy for the country.
      Pharmaceutical Management Agency (PHARMAC)
      Guidelines for Funding Applications to PHARMAC.
      PHARMAC, 2019
      https://pharmac.govt.nz/medicine-funding-and-supply/make-an-application/pharmconnect-make-a-medicine-funding-application/pharmconnect-supporting-information/#guidelines
      NZ’s Ministry of Health uses the EQ-5D-3L for creating patient-reported health outcome measures.
      • Shuker C.
      • Bohm G.
      • Hamblin R.
      • et al.
      Progress in public reporting in New Zealand since the Ombudsman’s ruling, and an invitation.
      N Z Med J. 2017; 130: 11-22
      Fundamental to these important applications in NZ are the EQ-5D-3L social value set and population norms that were created in 1999 from a survey of the general population.
      • Devlin N.J.
      • Hansen P.
      • Kind P.
      • Williams A.
      Logical inconsistencies in survey respondents’ health state valuations-a methodological challenge for estimating social tariffs.
      Health Econ. 2003; 12: 529-544
      A social value set consists of values for all possible EQ-5D health states (or profiles) and is intended to represent the “average” preferences of the general population. Population norms (ie, reference data or normative data) describe the HRQoL of the general population and comprise EQ-5D profiles (participants’ current health status, as described on each of the 5 dimensions), EQ-5D utility values (the values associated with these profiles, from the social value set), and EuroQol visual analog scale (EQ-VAS) scores, presented by age and gender. These norms are used to compare the HRQoL of patients with particular conditions vis-à-vis the HRQoL of average members of the relevant population subgroup (eg, specific age group) without the condition.
      • Szende A.
      • Janssen B.
      • Cabases J.
      Self-Reported Population Health: An International Perspective Based on EQ-5D.
      Springer Nature, Dordrecht, the Netherlands2014
      ,
      • Van den Berg B.
      SF−6D population norms.
      Health Econ. 2012; 21: 1508-1512
      In 2009, a revised version of the EQ-5D-3L, the EQ-5D-5L, was introduced internationally that increased the number of levels on each of the 5 dimensions from 3 to 5, thereby increasing the number of representable health states from 243 (35) to 3125 (55).
      • Herdman M.
      • Gudex C.
      • Lloyd A.
      • et al.
      Development and preliminary testing of the new five-level version of EQ-5D (EQ-5D-5L).
      Qual Life Res. 2011; 20: 1727-1736
      In 2018, an EQ-5D-5L social value set for NZ was created that involved representative participants from the NZ general population expressing their HRQoL preferences using an online valuation tool. The tool comprised an adaptive discrete choice experiment and a binary search algorithm to identify any health states worse than dead.
      • Sullivan T.
      • Hansen P.
      • Ombler F.
      • Derrett S.
      • Devlin N.
      A new tool for creating personal and social EQ-5D-5L value sets, including valuing ‘dead’.
      Soc Sci Med. 2020; 246112707
      Central to this methodology was the creation of a personal value set for each of the 2468 participants (ie, 3125 health state values for each person), from which the social value set was derived—in effect, by averaging the 2468 personal value sets.
      These 2468 personal value sets captured the heterogeneity of New Zealanders’ health state preferences and their HRQoL, from which population norms can be constructed. In addition to being the first NZ EQ-5D-5L population norms to be derived, this is the first time population norms have been constructed from personal value sets. In contrast, utility values (ie, the values associated with each participant’s EQ-5D profile) are usually obtained from the social value set instead of from each individual’s personal value set (based on their personal weights).
      • Devlin N.
      • Parkin D.
      • Janssen B.
      Methods for Analysing and Reporting EQ-5D Data.
      Springer Nature, Cham, Switzerland2020
      The construction, and hence general availability, of NZ population norms will serve 3 important inter-related purposes. NZ population norms will complement existing NZ HRQoL health data, strengthen the usefulness of the NZ EQ-5D-5L social value set (constructed from the same data set) for cost-utility analysis and patient-reported health outcome measures in NZ, and help researchers and policy makers to better understand the HRQoL of the NZ population. Given the extensive amount of international EQ-5D research over the past 3 decades, there is a lot of interest in the population health characteristics of different populations.
      • Szende A.
      • Janssen B.
      • Cabases J.
      Self-Reported Population Health: An International Perspective Based on EQ-5D.
      Springer Nature, Dordrecht, the Netherlands2014
      Currently, EQ-5D-5L population norms exist for China, Germany, Hong Kong, Italy, Japan, Poland, Quebec, Singapore, South Australia, Spain, and Vietnam.
      • Yang Z.
      • Busschbach J.
      • Liu G.
      • Luo N.
      EQ-5D-5L norms for the urban Chinese population in China.
      Health Qual Life Outcomes. 2018; 16: 210
      , 
      • Grochtdreis T.
      • Dams J.
      • Konig H.H.
      • Konnopka A.
      Health-related quality of life measured with the EQ-5D-5L: estimation of normative index values based on a representative German population sample and value set.
      Eur J Health Econ. 2019; 20: 933-944
      , 
      • Wong E.L.Y.
      • Cheung A.W.L.
      • Wong A.Y.K.
      • Xu R.H.
      • Ramos-Goni J.M.
      • Rivero-Arias O.
      Normative profile of health-related quality of life for Hong Kong general population using preference-based instrument EQ-5D-5L.
      Value Health. 2019; 22: 916-924
      , 
      • Scalone L.
      • Cortesi P.A.
      • Ciampichini R.
      • Cesana G.
      • Mantovani L.G.
      Health related quality of life norm data of the Italian general population: results using the EQ-5D-3L and EQ-5D-5L instruments.
      Epidemiol Biostat Public Health. 2015; 12
      , 
      • Shiroiwa T.
      • Fukuda T.
      • Ikeda S.
      • et al.
      Japanese population norms for preference-based measures: EQ-5D-3L, EQ-5D-5L, and SF-6D.
      Qual Life Res. 2016; 25: 707-719
      , 
      • Golicki D.
      • Niewada M.
      EQ-5D-5L Polish population norms.
      Arch Med Sci. 2017; 13: 191-200
      , 
      • Poder T.G.
      • Carrier N.
      • Kouakou C.R.C.
      Quebec health-related quality-of-life population norms using the EQ-5D-5L: decomposition by sociodemographic data and health problems.
      Value Health. 2020; 23: 251-259
      , 
      • Abdin E.
      • Subramaniam M.
      • Vaingankar J.A.
      • Luo N.
      • Chong S.A.
      Measuring health-related quality of life among adults in Singapore: population norms for the EQ-5D.
      Qual Life Res. 2013; 22: 2983-2991
      , 
      • McCaffrey N.
      • Kaambwa B.
      • Currow D.C.
      • Ratcliffe J.
      Health-related quality of life measured using the EQ-5D-5L: South Australian population norms.
      Health Qual Life Outcomes. 2016; 14: 133
      , 
      • Garcia-Gordillo M.A.
      • Adsuar J.C.
      • Olivares P.R.
      Normative values of EQ-5D-5L: in a Spanish representative population sample from Spanish health survey, 2011.
      Qual Life Res. 2016; 25: 1313-1321
      , 
      • Nguyen L.H.
      • Tran B.X.
      • Le Q.N.H.
      • Tran T.T.
      • Latkin C.A.
      Quality of life profile of general Vietnamese population using EQ-5D-5L.
      Health Qual Life Outcomes. 2017; 15: 199
      NZ EQ-5D-5L population norms will join this collection.
      The objective of this article was to report on the construction of NZ population norms for the EQ-5D-5L, including the results of investigating any associations between participants’ sociodemographic characteristics and HRQoL.

      Methods

      Data

      The data were generated in an earlier study that involved an online valuation tool for creating personal EQ-5D-5L value sets distributed to a representative sample of NZ adults in February and March 2018.
      • Sullivan T.
      • Hansen P.
      • Ombler F.
      • Derrett S.
      • Devlin N.
      A new tool for creating personal and social EQ-5D-5L value sets, including valuing ‘dead’.
      Soc Sci Med. 2020; 246112707
      Participants were asked to complete 3 tasks: (1) report their own health on the EQ-5D-5L questionnaire and EQ-VAS (explained below); (2) complete a discrete choice experiment (DCE) and identify any health states worse than dead, from which their weights for the levels on each dimension of the EQ-5D-5L were determined; and (3) provide their sociodemographic information. Ethical approval was obtained from the University of Otago Human Ethics Committee (D17/297). The tool’s extensive data quality checks resulted in a “high-quality” subsample of 2468 participants whose personal value sets were, in effect, averaged to create a social value set for NZ. For details about this earlier study, including the methods for creating the personal value sets and the data quality checks, see Sullivan et al.
      • Sullivan T.
      • Hansen P.
      • Ombler F.
      • Derrett S.
      • Devlin N.
      A new tool for creating personal and social EQ-5D-5L value sets, including valuing ‘dead’.
      Soc Sci Med. 2020; 246112707

      Variables Used to Derive the Population Norms

      EQ-5D-5L dimensions and profiles

      Each participant’s HRQoL in terms of their current health status was measured using the EQ-5D-5L questionnaire, resulting in a 5-digit profile for each participant. The EQ-5D-5L has 5 dimensions—mobility, self-care, usual activities, pain/discomfort and anxiety/depression—and each dimension has 5 levels of severity, ranging from “no problems” to “unable to do/extreme problems.” Thus, each EQ-5D-5L health state, of which there are 3125 (55) theoretically possible, can be summarized by a 5-digit profile corresponding to the 5 dimensions and 5 levels—ranging from 11111 (full health, where a participant reports no problems on all 5 dimensions) to 55555 (worst health, ie, maximum problems on all 5 dimensions).
      EuroQol Research Foundation
      EQ-5D-5L User Guide.
      2019
      https://euroqol.org/publications/user-guides

      Utility Values

      For each participant, their personal EQ-5D-5L value set was created by applying their preference weights from their DCE to each of the 3125 health states—with values anchored at unity for full health and zero for dead and negative values for any states worse than dead.
      • Sullivan T.
      • Hansen P.
      • Ombler F.
      • Derrett S.
      • Devlin N.
      A new tool for creating personal and social EQ-5D-5L value sets, including valuing ‘dead’.
      Soc Sci Med. 2020; 246112707
      Thus, each participant’s 5-digit profile, representing their HRQoL in terms of their current health, can be easily valued by consulting their personal value set (of 3125 values). In this article, the value of a participant’s 5-digit profile is referred to as their “utility value” (to differentiate it from their EQ-5D-5L value set of 3125 values).

      EQ-VAS Scores

      In addition to reporting their HRQoL on the EQ-5D-5L (resulting in their 5-digit profile), participants also marked their current HRQoL on the EQ-VAS—a visual analog scale (VAS) with scores ranging from 0 (the “worst imaginable health state”) to 100 (the “best imaginable health state”).

      Sociodemographic Variables

      These (self-reported) sociodemographic variables were included: gender (male, female, gender diverse), age (measured as a categorical variable in 10-year age groups except for 18-24 and ≥65), ethnicity, education level, employment status, individual income, living arrangements (live alone or with others), region, long-term disability (ie, a condition that prevents a person from doing everyday things that other people can do, lasting ≥6 months), and chronic disease (ie, a physical or mental illness diagnosed by a doctor, expected to last for ≥6 months; symptoms may come and go or be present all the time). These variables are commonly used in population norm studies,
      • Szende A.
      • Janssen B.
      • Cabases J.
      Self-Reported Population Health: An International Perspective Based on EQ-5D.
      Springer Nature, Dordrecht, the Netherlands2014
      with long-term disability and chronic disease included owing to their high prevalence in NZ and known association with HRQoL.
      Statistics New Zealand
      Disability.
      https://www.stats.govt.nz/information-releases/disability-survey-2013
      Date: 2013
      Date accessed: September 16, 2020
      , 
      Ministry of Health
      Annual data Explorer 2018/19: New Zealand health survey [data file].
      http://minhealthnz.shinyapps.io/nz-health-survey-2018-19-annual-data-explorer/
      Date: 2019
      Date accessed: September 16, 2020
      , 
      • Stanley J.
      • Semper K.
      • Millar E.
      • Sarfati D.
      Epidemiology of multimorbidity in New Zealand: a cross-sectional study using national-level hospital and pharmaceutical data.
      BMJ Open. 2018; 8
      To simplify the interpretation of results, the response categories for several of the sociodemographic variables were combined. The 6 most common ethnicity groups recognized by Statistics NZ were used to categorize ethnicity (people could identify with multiple ethnicities); “region” was aggregated into 5 main regions including “outside NZ” for NZ residents residing overseas; to align with the way similar countries (eg, Australia)
      • McCaffrey N.
      • Kaambwa B.
      • Currow D.C.
      • Ratcliffe J.
      Health-related quality of life measured using the EQ-5D-5L: South Australian population norms.
      Health Qual Life Outcomes. 2016; 14: 133
      report educational attainment (ie, up to secondary school, trade/certificate/diploma, degree or higher), “no qualifications” and “secondary school” were combined to form the “low” education category; “other postsecondary school qualification” was categorized as “medium” education and “university degree or equivalent” was categorized as “high” education; and with respect to employment status, the “student” and “homemaker” categories were combined, “self-employed” was included in “other” (as the number of hours of paid work was unknown), and “not in paid work” included people who were receiving a benefit, people who were unemployed but not on a benefit, and people who were not working.

      Statistical Analysis

      HRQoL, as measured by the EQ-5D-5L, utility values, and EQ-VAS scores, was summarized for all participants together and disaggregated by age groups and gender. Utility values and EQ-VAS scores were specified in terms of their means, standard deviations (SDs), medians (50th percentile). and interquartile ranges (25th and 75th percentiles), with their distributions represented via histograms. The number, percentages, and cumulative percentages were calculated for the 25 most commonly reported EQ-5D-5L profiles. The mean utility values and EQ-VAS scores for each of these profiles were calculated by summing the participants’ utility values and EQ-VAS scores associated with these profiles and taking their respective means.
      Responses on the EQ-5D-5L dimensions were recoded into 0 = “no problems” and 1 = “slight, moderate, severe or extreme problems”. Multivariable logistic regression models were used to examine the association between gender, age group, ethnicity, education, region, living arrangement, employment, income, long-term disability, and chronic disease and each of the EQ-5D-5L dimensions (“slight, moderate, severe or extreme problems” vs “no problems”) separately. Owing to the ceiling effects in the utility and EQ-VAS scores (ie, scores >1 for utility and >100 for EQ-VAS are impossible), multivariable Tobit regression models
      • Austin P.C.
      • Escobar M.
      • Kopec J.A.
      The use of the Tobit model for analyzing measures of health status.
      Qual Life Res. 2000; 9: 901-910
      were used to examine the association between these 2 scores (separately) and the sociodemographic variables. Statistical analyses were performed using STATA/SE 16.1 (StataCorp LLC, College Station, TX).

      Results

      Participants

      The 2468 participants’ sociodemographic characteristics are reported in Table 1, along with the characteristics of the NZ population (where data are available) for comparison purposes. Of these 2468 people, 52.9% were female, 22.5% aged ≥65 years, most identified as NZ European (63.3%), 33.8% reported a low level of education (ie, no qualifications or secondary school only), and 39.1% were in full-time paid work. More than half of the participants (54.7%) reported having at least 1 chronic illness or disease lasting ≥6 months and 25.7% had a long-term disability. The sample is representative of the NZ general population, and this conclusion is supported by the comparisons shown in Table 1 in terms of age, gender, ethnicity, and health status (as defined by long-term disability and chronic illness or disease).
      Statistics New Zealand
      Disability.
      https://www.stats.govt.nz/information-releases/disability-survey-2013
      Date: 2013
      Date accessed: September 16, 2020
      , 
      Ministry of Health
      Annual data Explorer 2018/19: New Zealand health survey [data file].
      http://minhealthnz.shinyapps.io/nz-health-survey-2018-19-annual-data-explorer/
      Date: 2019
      Date accessed: September 16, 2020
      , 
      • Stanley J.
      • Semper K.
      • Millar E.
      • Sarfati D.
      Epidemiology of multimorbidity in New Zealand: a cross-sectional study using national-level hospital and pharmaceutical data.
      BMJ Open. 2018; 8
      A direct comparison of participants’ education and income data with census data is difficult because the census data include ≥15 years, whereas participants in this survey were >18 years old. Because most young people in NZ attend school until they are 17 years old
      Ministry of Education
      Education counts.
      https://www.educationcounts.govt.nz/indicators/main/student-engagement-participation/retention_of_students_in_senior_secondary_schools
      Date: 2020
      Date accessed: September 16, 2020
      (school in NZ is compulsory from 6 to 16 years and available for free), often followed by postsecondary school study (and hence will be on low or nil income), this may account for the lower number of people in the “no qualifications/secondary school” and “≤$20 000 income” groups in our study than the census data. Nonetheless, it is likely that participants with low education are underrepresented and participants with an income of >$70 000 are overrepresented.
      Table 1Sociodemographic characteristics of the participants (n = 2468) and New Zealand census data.
      CharacteristicParticipantsNew Zealand census data (2013)
      n%%
      Gender
       Male115746.947.9
       Female130652.952.1
       Gender diverse50.2Unrecorded
      Age (y)
       18-2425210.212.8
       25-3443917.816.1
       35-4444017.817.9
       45-5438115.418.8
       55-6440216.315.4
       ≥6555422.519.0
      Ethnicity
      Sums to >100% as some people identify with multiple ethnicities.
       New Zealand European156363.364.3
       Māori39015.814.1
       Pacific1084.46.9
       Asian34013.811.1
       MELAA502.01.2
       Others27911.313.6
      Education level
      New Zealand population statistics include people ≥ 15 years.
       No qualifications/secondary school (low)83433.854.2
       Other postsecondary school qualifications (medium)55522.525.7
       University degree or equivalent (high)107943.720.1
      Employment status
       Full-time work (≥30 h/wk)96539.1
      2017 September quarter employment and unemployment rates were 67.8% and 4.5%.
       Part-time work (<30 h/wk)38015.4
       Not in paid work (including people on a benefit)2379.6
       Student/Homemaker33013.4
       Retired49119.9
       Others (including self-employed)652.6
      Individual income
       ≤ $20 00055222.438.2
       $20 001-$30 00042517.213.7
       $30 001-$50 00053721.821.4
       $50 001-$70 00041416.812.9
       $70 001-$100 00033613.67.8
       ≥ $100 0012048.36.0
      Living arrangement
       Living alone32813.3
       Living with others214086.7
      Region
       Northern89836.4
       Midland45218.3
       Central56723.0
       Southern54722.1
       Outside New Zealand40.2
      Long-term disability (lasting ≥ 6 mo)
       No183474.376.0
       Yes63425.724.0
      Chronic disease
       No111945.3
       Yes134954.7
      Note: Adapted from “A new tool for creating personal and social EQ-5D-5L value sets, including valuing ‘dead’
      • Brazier J.
      • Ratcliffe J.
      • Saloman J.
      • Tsuchiya A.
      Measuring and Valuing Health Benefits for Economic Evaluation.
      Oxford University Press, Oxford, United Kingdom2017
      , Social Science and Medicine, 246:112707, p. 5. Copyright 2019 by Elsevier B.V.
      MELAA indicates Middle Eastern/Latin American/African.
      ∗ Sums to >100% as some people identify with multiple ethnicities.
      † New Zealand population statistics include people ≥ 15 years.
      ‡ 2017 September quarter employment and unemployment rates were 67.8% and 4.5%.

      EQ-5D-5L Profiles, Utility Values, and EQ-VAS Scores for the NZ Population

      The mean utility values and EQ-VAS scores are reported in Table 2 together with their SDs, medians and interquartile ranges. Appendix Figure 1 in Supplemental Materials found at https://doi.org/10.1016/j.jval.2021.04.1280 illustrates the distribution of the utility values and EQ-VAS scores. Both the utility values and EQ-VAS scores are skewed to the left with utility values ranging from −1.924 to 1.000 and EQ-VAS scores from 1 to 100. A total of 37 participants (1.5%) had a negative utility value and 543 (22%) had a full-health utility value of 1 (ie, profile = 11111). Nevertheless, only 60 participants (2.4%) reported a score of 100 on the EQ-VAS, with most scores clustered at 80 (16%).
      Table 2HRQoL by EQ-5D-5L dimension, utility value and EQ-VAS score, disaggregated by age group with column percentages shown.
      DimensionAge group (years)
      18-2425-3435-4445-5455-64≥ 65Total
      n (%)n (%)n (%)n (%)n (%)n (%)n (%)
      Mobility
       No problems219 (86.9)386 (87.9)360 (81.8)274 (71.9)241 (60.0)300 (54.2)1780 (72.1)
       Slight problems27 (10.7)37 (8.4)60 (13.6)65 (17.1)98 (24.4)151 (27.3)438 (17.7)
       Moderate problems6 (2.4)13 (3.0)16 (3.6)30 (7.9)48 (11.9)69 (12.5)182 (7.4)
       Severe problems0 (0.0)3 (0.7)1 (0.2)9 (2.4)12 (3.0)31 (5.6)56 (2.3)
       Unable to0 (0.0)0 (0.0)3 (0.7)3 (0.8)3 (0.7)3 (0.5)12 (0.5)
      Self-care
       No problems245 (97.2)414 (94.3)417 (94.8)339 (89.0)356 (88.6)484 (87.4)2255 (91.4)
       Slight problems6 (2.4)17 (3.9)19 (4.3)35 (9.2)35 (8.7)49 (8.8)161 (6.5)
       Moderate problems1 (0.4)4 (0.9)2 (0.5)7 (1.8)8 (2.0)16 (2.9)38 (1.5)
       Severe problems0 (0.0)4 (0.9)1 (0.2)0 (0.0)1 (0.2)4 (0.7)10 (0.4)
       Unable to0 (0.0)0 (0.0)1 (0.2)0 (0.0)2 (0.5)1 (0.2)4 (0.2)
      Usual activities
       No problems189 (75.0)369 (84.1)348 (79.1)271 (71.1)245 (60.9)310 (56.0)1732 (70.2)
       Slight problems43 (17.1)45 (10.3)68 (15.5)66 (17.3)103 (25.6)166 (30.0)491 (19.9)
       Moderate problems15 (6.0)18 (4.1)12 (2.7)34 (8.9)36 (9.0)54 (9.7)169 (6.8)
       Severe problems4 (1.6)7 (1.6)10 (2.3)10 (2.6)13 (3.2)22 (4.0)66 (2.7)
       Unable to1 (0.4)0 (0.0)2 (0.5)0 (0.0)5 (1.2)2 (0.4)10 (0.4)
      Pain/discomfort
       No problems134 (53.2)236 (53.8)191 (43.4)128 (33.6)106 (26.4)151 (27.3)946 (38.3)
       Slight problems91 (36.1)153 (34.9)187 (42.5)167 (43.8)180 (44.8)278 (50.2)1056 (42.8)
       Moderate problems21 (8.3)43 (9.8)50 (11.4)66 (17.3)82 (20.4)103 (18.6)365 (14.8)
       Severe problems5 (2.0)6 (1.4)11 (2.5)15 (3.9)27 (6.7)22 (4.0)86 (3.5)
       Extreme problems1 (0.4)1 (0.2)1 (0.2)5 (1.3)7 (1.7)0 (0.0)15 (0.6)
      Anxiety/depression
       No problems77 (30.6)191 (43.5)224 (50.9)193 (50.7)231 (57.5)408 (73.6)1324 (53.6)
       Slight problems79 (31.3)160 (36.4)141 (32.0)114 (29.9)106 (26.4)113 (20.4)713 (28.9)
       Moderate problems55 (21.8)54 (12.3)56 (12.7)53 (13.9)48 (11.9)32 (5.8)298 (12.1)
       Severe problems34 (13.5)23 (5.2)15 (3.4)16 (4.2)12 (3.0)0 (0.0)100 (4.1)
       Extreme problems7 (2.8)11 (2.5)4 (0.9)5 (1.3)5 (1.2)1 (0.2)33 (1.3)
      Utility value
       Mean0.8030.8700.8750.8230.8150.8650.847
       Standard deviation0.3380.2020.1910.2830.2690.1830.240
       50th percentile (median)0.9110.9400.9440.9110.9060.9250.926
       25th percentile0.7510.8390.8460.7780.7690.8280.811
       75th percentile0.9841.0001.0000.9820.9790.9840.988
      EQ-VAS score
       Mean72.174.473.873.374.578.474.8
       Standard deviation19.017.617.518.618.716.618.0
       50th percentile (median)78.080.077.080.080.080.080.0
       25th percentile65.070.065.065.065.070.067.5
       75th percentile85.088.085.085.090.090.090.0
      EQ-VAS indicates EuroQol visual analog scale; HRQoL, health-related quality of life.
      In Table 2, HRQoL, as measured by the number and percentage for the EQ-5D-5L dimensions and the central tendency and spread for the utility values and EQ-VAS scores, is presented by age group. The EQ-5D-5L dimension for which the highest proportion of participants reported problems was pain or discomfort (61.7%), followed by anxiety or depression (46.4%), usual activities (29.8%), mobility (27.9%), and, with the lowest proportion, self-care (8.6%). The mean utility value was 0.847 (SD 0.240) and the mean EQ-VAS score was 74.8 (SD 18.0). Negative utility values ranged from −1.924 to −0.005. There was some variability in the EQ-5D-5L dimensions, utility values, and EQ-VAS scores between all age groups. Similar results were obtained when age groups were disaggregated by gender (see Appendix Tables 1 and 2 in Supplemental Materials found at https://doi.org/10.1016/j.jval.2021.04.1280).
      The 25 most commonly reported EQ-5D-5L profiles, with their respective mean utility values and EQ-VAS scores, are presented in Table 3. Of the 3125 possible profiles, 289 (9%) were reported by participants, with the majority (78%) represented by 25 profiles. The 4 most common profiles were as follows (in descending order): 11111 (n = 543, 22.0%), 11121 (n = 337, 13.7%), 11112 (n = 219, 8.9%), and 11122 (n = 209, 8.5%). These profiles had mean utility values of 1.000, 0.952, 0.942, and 0.871 and mean EQ-VAS scores of 84.7, 81.2, 78.4, and 77.5, respectively.
      Table 3Most commonly reported EQ-5D-5L profiles, with associated mean utility values and EQ-VAS scores.
      EQ-5D-5L profile
      Digits represent the response levels (1-5) for each of the 5 EQ-5D-5L dimensions (mobility, self-care, usual activities, pain/discomfort and anxiety/depression).
      		NumberPercentageCumulative percentageMean utility valueMean EQ-VAS score
      1111154322.022.01.00084.7
      1112133713.735.70.95281.2
      111122198.944.50.94278.4
      111222098.553.00.87177.5
      21221702.855.80.87977.5
      11113662.758.50.87471.1
      21121662.761.20.91083.5
      11123652.663.80.82570.2
      11221391.665.40.91278.3
      11222331.366.70.85373.5
      11131311.368.00.84877.9
      21222271.169.10.81275.0
      11132261.170.10.82568.5
      21122261.171.20.82474.7
      21231230.972.10.83969.3
      11114190.872.90.72765.7
      11223180.773.60.67267.0
      11232160.674.30.78363.8
      21232160.674.90.74374.1
      11133140.675.50.77365.7
      11211140.676.10.94587.4
      21131140.676.60.87370.9
      31231140.677.20.76065.4
      11231130.577.70.88570.4
      21123130.578.20.76072.2
      EQ-VAS indicates EuroQol visual analog scale.
      ∗ Digits represent the response levels (1-5) for each of the 5 EQ-5D-5L dimensions (mobility, self-care, usual activities, pain/discomfort and anxiety/depression).

      Association Between Sociodemographic Characteristics and EQ-5D-5L Dimensions, Utility Values and EQ-VAS Scores

      The results of the logistic regression models showing the association between reported problems in the EQ-5D-5L dimensions and selected sociodemographic characteristics are presented in Table 4. Compared with people aged 18 to 24 years, people aged ≥45 years had higher odds of mobility problems, with the highest odds in the ≥65 years age group (odds ratio [OR] 2.85; 95% confidence interval [CI] 1.60-5.06). Similar results were found with the pain/discomfort dimension. Compared with the youngest age group (18-24 years), the odds of problems with pain/discomfort increased for people aged ≥35 years, with OR of 2.49 (95% CI 1.54–4.03) higher for people aged ≥65 years. In contrast, compared with 18- to 24-year-olds, people in all other age groups had lower odds of problems with anxiety/depression, and the odds decreased with age, with the oldest age group having the lowest odds (OR 0.15; 95% CI 0.09–0.24).
      Table 4Multivariable logistic regression models of reported problems in the EQ-5D-5L dimensions and sociodemographic characteristics.
      VariableMobilitySelf-careUsual activitiesPain/discomfortAnxiety/depression
      OR (95% CI)Global P-value
      The global P-value is shown; evidence of differences can be determined by the confidence intervals.
      		OR (95% CI)Global P-valueOR (95% CI)Global P-valueOR (95% CI)Global P-valueOR (95% CI)Global P-value
      Gender (Ref. Female and Gender diverse)
      Owing to small numbers, the gender diverse category (n = 5) was combined with the female category.
       Male0.96 (0.77, 1.21)0.7531.19 (0.85, 1.65)0.3070.92 (0.74, 1.14)0.4290.94 (0.78, 1.14)0.5200.84 (0.69, 1.01)0.064
      Age group (Ref. 18-24)
       25-340.94 (0.56, 1.58)<0.0012.41 (0.98, 5.92)0.1520.63 (0.41, 0.98)0.1131.13 (0.80, 1.59)<0.0010.69 (0.48, 0.99)<0.001
       35-441.38 (0.83, 2.27)1.82 (0.73, 4.59)0.81 (0.52, 1.25)1.70 (1.19, 2.42)0.47 (0.33, 0.69)
       45-541.86 (1.13, 3.06)2.90 (1.20, 7.02)0.91 (0.58, 1.41)2.24 (1.55, 3.25)0.38 (0.26, 0.56)
       55-642.52 (1.54, 4.13)2.09 (0.86, 5.07)1.06 (0.68, 1.64)2.68 (1.82, 3.94)0.24 (0.16, 0.35)
       ≥652.85 (1.60, 5.06)2.63 (0.98, 7.05)1.11 (0.66, 1.89)2.49 (1.54, 4.03)0.15 (0.09, 0.24)
      Ethnicity
      A reference group is used for each category as some people identified with multiple ethnicities.
       New Zealand European (Ref non-New Zealand European)0.90 (0.58, 1.38)0.3441.36 (0.75, 2.48)0.8331.32 (0.87, 1.98)0.2481.23 (0.87, 1.74)0.2572.12 (1.49, 3.02)<0.001
       Māori (Ref non-Māori)0.91 (0.62, 1.34)1.09 (0.64, 1.84)0.86 (0.60, 1.24)1.00 (0.73, 1.37)1.00 (0.73, 1.38)
       Pacific (Ref non-Pacific)1.30 (0.72, 2.32)1.58 (0.69, 3.59)1.25 (0.71, 2.18)1.02 (0.64, 1.60)0.84 (0.53, 1.33)
       Asian (Ref non-Asian)0.57 (0.32, 1.01)1.04 (0.42, 2.59)0.90 (0.53, 1.52)1.07 (0.71, 1.61)1.68 (1.10, 2.56)
       MELAA (Ref non-MELAA)0.78 (0.31, 1.95)0.77 (0.16, 3.75)0.90 (0.37, 2.18)1.44 (0.73, 2.87)1.34 (0.67, 2.66)
       Others (Ref not Others)0.93 (0.56, 1.54)1.42 (0.69, 2.91)1.12 (0.69, 1.81)0.86 (0.57, 1.29)1.78 (1.17, 2.69)
      Education level (Ref. Low)
       Medium0.83 (0.63, 1.09)0.2241.07 (0.72, 1.58)0.8800.92 (0.70, 1.20)0.7700.94 (0.73, 1.21)0.3811.06 (0.83, 1.35)0.629
       High1.04 (0.81, 1.35)1.10 (0.75, 1.62)1.00 (0.78, 1.29)0.86 (0.69, 1.07)0.94 (0.75, 1.17)
      Region (Ref. Northern and outside New Zealand)
      Owing to small numbers, the outside NZ category (n = 4) was combined with the Northern category.
       Midland1.17 (0.86, 1.60)0.3631.37 (0.86, 2.17)0.1320.89 (0.66, 1.21)0.3021.04 (0.80, 1.35)0.9870.91 (0.70, 1.17)0.707
       Central1.26 (0.95, 1.67)1.60 (1.05, 2.43)1.16 (0.89, 1.53)1.00 (0.78, 1.27)0.99 (0.78, 1.26)
       Southern1.03 (0.77, 1.38)1.12 (0.71, 1.77)0.92 (0.69, 1.22)1.03 (0.80, 1.32)0.88 (0.69, 1.13)
      Living arrangement (Ref. Living alone)
       Living with others1.08 (0.80, 1.47)0.6020.68 (0.47, 1.00)0.0560.92 (0.69, 1.24)0.6021.15 (0.86, 1.53)0.3410.87 (0.66, 1.15)0.324
      Employment status (Ref. Full-time work)
       Part-time work1.06 (0.73, 1.52)0.0501.19 (0.65, 2.18)0.2481.01 (0.71, 1.44)0.0040.76 (0.57, 1.01)0.2900.86 (0.65, 1.15)<0.001
       Not in paid work1.22 (0.79, 1.88)2.05 (1.11, 3.76)1.69 (1.11, 2.58)1.09 (0.72, 1.65)1.87 (1.26, 2.79)
       Student/homemaker0.74 (0.48, 1.14)1.20 (0.61, 2.37)1.00 (0.68, 1.49)0.85 (0.62, 1.18)1.29 (0.93, 1.80)
       Retired1.67 (1.07, 2.61)1.36 (0.68, 2.71)1.87 (1.20, 2.92)0.95 (0.62, 1.46)0.67 (0.44, 1.01)
       Other (including self-employed)1.20 (0.63, 2.28)1.82 (0.77, 4.29)2.05 (1.10, 3.84)1.20 (0.65, 2.21)1.09 (0.61, 1.93)
      Individual income (Ref. ≤ $20 000)
       $20 000-$30 0000.89 (0.64, 1.24)0.2571.08 (0.71, 1.66)0.4900.77 (0.56, 1.06)0.0100.98 (0.71, 1.33)0.0901.02 (0.76, 1.38)0.101
       $30 001-$50 0000.70 (0.50, 1.00)0.66 (0.40, 1.10)0.53 (0.38, 0.74)0.78 (0.58, 1.06)0.80 (0.60, 1.08)
       $50 001-$70 0000.87 (0.58, 1.29)0.77 (0.41, 1.43)0.63 (0.43, 0.92)0.74 (0.53, 1.03)0.84 (0.61, 1.17)
       $70 001-$100 0000.98 (0.63, 1.51)0.88 (0.44, 1.78)0.68 (0.45, 1.04)0.82 (0.57, 1.18)0.62 (0.43, 0.90)
       ≥ $100 0010.68 (0.41, 1.13)0.80 (0.36, 1.75)0.56 (0.34, 0.91)0.57 (0.38, 0.85)0.89 (0.59, 1.33)
      Long-term disability (Ref. No)
       Yes5.25 (4.14, 6.66)<0.0015.04 (3.50, 7.25)<0.0013.99 (3.17, 5.02)<0.0012.12 (1.65, 2.73)<0.0011.61 (1.28, 2.03)<0.001
      Chronic disease (Ref. No)
       Yes2.42 (1.90, 3.08)<0.0012.18 (1.39, 3.44)<0.0013.03 (2.39, 3.83)<0.0012.24 (1.85, 2.70)<0.0012.52 (2.06, 3.07)<0.001
      CI indicates confidence interval; MELAA, Middle Eastern/Latin American/African; OR, odds ratio; Ref., reference.
      ∗ The global P-value is shown; evidence of differences can be determined by the confidence intervals.
      † Owing to small numbers, the gender diverse category (n = 5) was combined with the female category.
      ‡ A reference group is used for each category as some people identified with multiple ethnicities.
      § Owing to small numbers, the outside NZ category (n = 4) was combined with the Northern category.
      With respect to ethnicity, NZ Europeans compared with non-NZ Europeans, and Asians compared with non-Asians had higher odds of problems with anxiety/depression (OR 2.12; 95% CI 1.49–3.02; OR 1.68; 95% CI 1.10–2.56, respectively). Compared with people in full-time paid work, the odds of problems with usual activities for people not in paid work or who were retired were 1.69 (95% CI 1.11–2.58) and 1.87 (95% CI 1.20–2.92) times higher, with people not in paid work also having higher odds of problems with anxiety/depression (OR 1.87; 95% CI 1.26–2.79). People with an income of $30 001 to $70 000 or >$100 000 (compared with people earning ≤$20 000) had lower odds of problems with usual activities. Finally, the odds of problems on all 5 dimensions were higher for people with a long-term disability or chronic disease, with the odds as high as 5.25 (95% CI 4.14–6.66) for mobility, for people with a disability, and 3.03 (95% CI 2.39–3.83) for usual activities, for people with a chronic disease.
      The results of the multivariable Tobit models for utility values, EQ-VAS scores, and selected sociodemographic characteristics are presented in Table 5. In line with the reported problems as discussed earlier, the results indicate that age, ethnicity, employment status, long-term disability, and chronic illness affect a person’s utility. Compared with people aged 18 to 24, older age groups (except for 45- to 54-year-olds) had, on average, higher utility values. The greatest difference was in the oldest age group, where the mean utility value was 0.104 higher for people aged ≥65 years than for 18- to 24-year-olds. People who identified as NZ European or Asian had lower mean utility (−0.089 [95% CI −0.130 to −0.048] and −0.058 [95% CI −0.108, to −0.007]) than non-NZ Europeans and non-Asians respectively, whereas people not in paid work had lower utility than people in full-time paid employment (−0.102 [95% CI −0.147 to −0.056]). The largest difference in utility was for people with a long-term disability or a chronic illness. They had −0.157 (95% CI −0.184 to −0.130) and −0.127 (95% CI −0.151 to −0.104) lower mean utility than people without a disability or chronic illness.
      Table 5Multivariable Tobit models of utility values and EQ-VAS scores and sociodemographic characteristics.
      VariableUtility valuesEQ-VAS scores
      Coefficient (95% CI)P-value
      The global P-value is shown; evidence of differences can be determined by the confidence intervals.
      		Coefficient (95% CI)P-value
      Gender (Ref. Female)0.0470.898
       Male0.020 (−0.002, 0.042)−0.3 (−1.7, 1.1)
       Gender diverse−0.194 (−0.417, 0.029)0.9 (−13.5, 15.4)
      Age group (Ref. 18–24 years)0.006<0.001
       25-340.049 (0.006, 0.092)1.0 (−1.6, 3.7)
       35-440.064 (0.021, 0.108)1.3 (−1.4, 4.1)
       45-540.037 (−0.008, 0.082)3.1 (0.3, 5.9)
       55-640.067 (0.021, 0.113)6.7 (3.8, 9.6)
       ≥650.104 (0.048, 0.160)10.1 (6.6, 13.7)
      Ethnicity
      A reference group is used for each category as some people identified with multiple ethnicities.
      		<0.0010.213
       New Zealand European (Ref. non-New Zealand European)−0.089 (−0.130, −0.048)0.4 (−2.2, 2.9)
       Māori (Ref. non-Māori)−0.015 (−0.053, 0.022)0.5 (−1.9, 2.8)
       Pacific (Ref. non-Pacific)−0.002 (−0.058, 0.054)−2.9 (−6.3, 0.5)
       Asian (Ref. non-Asian)−0.058 (−0.108, −0.007)0.5 (−2.6, 3.7)
       MELAA (Ref. non-MELAA)−0.038 (−0.120, 0.045)4.0 (−1.1, 9.2)
       Others (Ref. not others)−0.051 (−0.100, −0.003)1.4 (−1.6, 4.4)
      Education level (Ref. Low)0.2320.746
       Medium0.012 (−0.016, 0.041)0.7 (−1.1, 2.5)
       High0.023 (−0.003, 0.049)0.2 (−1.4, 1.9)
      Region (Ref. Northern)0.3860.370
       Midland0.000 (−0.030, 0.031)0.2 (−1.7, 2.1)
       Central−0.011 (−0.040, 0.017)−0.6 (−2.4, 1.1)
       Southern0.002 (−0.027, 0.031)−0.1 (−1.9, 1.7)
       Outside New Zealand0.305 (−0.030, 0.640)15.2 (−0.9, 31.3)
      Living arrangement (Ref. Living alone)0.7960.728
       Living with others0.004 (−0.028, 0.036)−0.4 (−2.4, 1.7)
      Employment status (Ref. Full-time work)<0.0010.006
       Part-time work−0.001 (−0.036, 0.034)0.3 (−1.9, 2.4)
       Not in paid work−0.102 (−0.147, −0.056)−4.4 (−7.3, −1.5)
       Student/homemaker−0.006 (−0.046, 0.033)1.4 (−1.1, 3.9)
       Retired−0.008 (−0.056, 0.041)−0.6 (−3.6, 2.4)
       Other (including self-employed)−0.036 (−0.105, 0.033)−2.1 (−6.4, 2.2)
      Individual income (Ref. ≤ $20 000)0.0560.043
       $20 000-$30 0000.027 (−0.008, 0.062)1.4 (−0.8, 3.6)
       $30 001-$50 0000.054 (0.019, 0.089)3.1 (0.9, 5.3)
       $50 001-$70 0000.048 (0.008, 0.087)1.8 (−0.6, 4.3)
       $70 001-$100 0000.057 (0.014, 0.100)3.5 (0.8, 6.2)
       ≥ $100 0010.049 (0.000, 0.098)4.1 (1.0, 7.1)
      Long-term disability (Ref. No)<0.001<0.001
       Yes−0.157 (−0.184, −0.130)−9.4 (−11.1, −7.7)
      Chronic disease (Ref. No)<0.001<0.001
       Yes−0.127 (−0.151, −0.104)−9.1 (−10.6, −7.7)
      CI indicates confidence interval; EQ-VAS, EuroQol visual analog scale; MELAA, Middle Eastern/Latin American/African; OR, odds ratio; Ref., reference.
      ∗ The global P-value is shown; evidence of differences can be determined by the confidence intervals.
      † A reference group is used for each category as some people identified with multiple ethnicities.
      In terms of the EQ-VAS, age, employment status, individual income, long-term disability, and chronic illness affected the scores. Similar to the utility results, people aged ≥45 years had higher EQ-VAS scores than 18- to 24-year-olds, and people with an individual income of $30 001 to $50 000 or >$70 000 (compared with people earning ≤$20 000) also had higher EQ-VAS scores. In contrast, people not in paid work reported lower EQ-VAS scores (−4.4 [95% CI −7.3 to −1.5]) than people in full-time paid employment, and people with a long-term disability or a chronic illness had much lower EQ-VAS scores—−9.4 (95% CI −11.1 to −7.7) and −9.1 (95% CI −10.6 to −7.7), respectively—than those without a disability or chronic illness.

      Discussion

      This study is the first to construct NZ EQ-5D-5L population norms. It is also the first time that population norms have been constructed from personal value sets—based on 2468 personal EQ-5D-5L value sets obtained from a representative sample of the NZ population.
      • Sullivan T.
      • Hansen P.
      • Ombler F.
      • Derrett S.
      • Devlin N.
      A new tool for creating personal and social EQ-5D-5L value sets, including valuing ‘dead’.
      Soc Sci Med. 2020; 246112707
      In addition, associations between participants’ sociodemographic characteristics and HRQoL were examined. The results of this study can be used to supplement existing data (eg, morbidity), compare specific population groups (eg, patients) with the general population, and inform healthcare research and policy making.
      The mean EQ-5D-5L utility value for the NZ population is 0.847. This value is close to the mean EQ-5D-5L utility value for Quebec (0.824),
      • Poder T.G.
      • Carrier N.
      • Kouakou C.R.C.
      Quebec health-related quality-of-life population norms using the EQ-5D-5L: decomposition by sociodemographic data and health problems.
      Value Health. 2020; 23: 251-259
      but it is lower than the values for China (0.957),
      • Yang Z.
      • Busschbach J.
      • Liu G.
      • Luo N.
      EQ-5D-5L norms for the urban Chinese population in China.
      Health Qual Life Outcomes. 2018; 16: 210
      Japan (0.9551),
      • Shiroiwa T.
      • Fukuda T.
      • Ikeda S.
      • et al.
      Japanese population norms for preference-based measures: EQ-5D-3L, EQ-5D-5L, and SF-6D.
      Qual Life Res. 2016; 25: 707-719
      Singapore (0.95),
      • Abdin E.
      • Subramaniam M.
      • Vaingankar J.A.
      • Luo N.
      • Chong S.A.
      Measuring health-related quality of life among adults in Singapore: population norms for the EQ-5D.
      Qual Life Res. 2013; 22: 2983-2991
      Hong Kong (0.919),
      • Wong E.L.Y.
      • Cheung A.W.L.
      • Wong A.Y.K.
      • Xu R.H.
      • Ramos-Goni J.M.
      • Rivero-Arias O.
      Normative profile of health-related quality of life for Hong Kong general population using preference-based instrument EQ-5D-5L.
      Value Health. 2019; 22: 916-924
      Italy (0.915),
      • Scalone L.
      • Cortesi P.A.
      • Ciampichini R.
      • Cesana G.
      • Mantovani L.G.
      Health related quality of life norm data of the Italian general population: results using the EQ-5D-3L and EQ-5D-5L instruments.
      Epidemiol Biostat Public Health. 2015; 12
      South Australia (0.91),
      • McCaffrey N.
      • Kaambwa B.
      • Currow D.C.
      • Ratcliffe J.
      Health-related quality of life measured using the EQ-5D-5L: South Australian population norms.
      Health Qual Life Outcomes. 2016; 14: 133
      Vietnam (0.91),
      • Nguyen L.H.
      • Tran B.X.
      • Le Q.N.H.
      • Tran T.T.
      • Latkin C.A.
      Quality of life profile of general Vietnamese population using EQ-5D-5L.
      Health Qual Life Outcomes. 2017; 15: 199
      Spain (0.897),
      • Garcia-Gordillo M.A.
      • Adsuar J.C.
      • Olivares P.R.
      Normative values of EQ-5D-5L: in a Spanish representative population sample from Spanish health survey, 2011.
      Qual Life Res. 2016; 25: 1313-1321
      Poland (0.888),
      • Golicki D.
      • Niewada M.
      EQ-5D-5L Polish population norms.
      Arch Med Sci. 2017; 13: 191-200
      and Germany (0.88).
      • Grochtdreis T.
      • Dams J.
      • Konig H.H.
      • Konnopka A.
      Health-related quality of life measured with the EQ-5D-5L: estimation of normative index values based on a representative German population sample and value set.
      Eur J Health Econ. 2019; 20: 933-944
      Notably, 22% (n = 543) of the participants reported no problems on all 5 dimensions (ie, 11111). This is a much lower proportion of people reporting perfect health than other countries, such as Vietnam (67.4%),
      • Nguyen L.H.
      • Tran B.X.
      • Le Q.N.H.
      • Tran T.T.
      • Latkin C.A.
      Quality of life profile of general Vietnamese population using EQ-5D-5L.
      Health Qual Life Outcomes. 2017; 15: 199
      Spain (62.4%),
      • Garcia-Gordillo M.A.
      • Adsuar J.C.
      • Olivares P.R.
      Normative values of EQ-5D-5L: in a Spanish representative population sample from Spanish health survey, 2011.
      Qual Life Res. 2016; 25: 1313-1321
      Japan (55%),
      • Shiroiwa T.
      • Fukuda T.
      • Ikeda S.
      • et al.
      Japanese population norms for preference-based measures: EQ-5D-3L, EQ-5D-5L, and SF-6D.
      Qual Life Res. 2016; 25: 707-719
      China (54%),
      • Yang Z.
      • Busschbach J.
      • Liu G.
      • Luo N.
      EQ-5D-5L norms for the urban Chinese population in China.
      Health Qual Life Outcomes. 2018; 16: 210
      Hong Kong (46%),
      • Wong E.L.Y.
      • Cheung A.W.L.
      • Wong A.Y.K.
      • Xu R.H.
      • Ramos-Goni J.M.
      • Rivero-Arias O.
      Normative profile of health-related quality of life for Hong Kong general population using preference-based instrument EQ-5D-5L.
      Value Health. 2019; 22: 916-924
      South Australia (42.8%),
      • McCaffrey N.
      • Kaambwa B.
      • Currow D.C.
      • Ratcliffe J.
      Health-related quality of life measured using the EQ-5D-5L: South Australian population norms.
      Health Qual Life Outcomes. 2016; 14: 133
      Poland (38.5%),
      • Golicki D.
      • Niewada M.
      EQ-5D-5L Polish population norms.
      Arch Med Sci. 2017; 13: 191-200
      Italy (38%),
      • Scalone L.
      • Cortesi P.A.
      • Ciampichini R.
      • Cesana G.
      • Mantovani L.G.
      Health related quality of life norm data of the Italian general population: results using the EQ-5D-3L and EQ-5D-5L instruments.
      Epidemiol Biostat Public Health. 2015; 12
      and Germany (30.6%),
      • Grochtdreis T.
      • Dams J.
      • Konig H.H.
      • Konnopka A.
      Health-related quality of life measured with the EQ-5D-5L: estimation of normative index values based on a representative German population sample and value set.
      Eur J Health Econ. 2019; 20: 933-944
      although it is similar to Quebec (20.8%).
      • Poder T.G.
      • Carrier N.
      • Kouakou C.R.C.
      Quebec health-related quality-of-life population norms using the EQ-5D-5L: decomposition by sociodemographic data and health problems.
      Value Health. 2020; 23: 251-259
      If there is a ceiling effect—where more of the population is at the ceiling—then the measure is less able to represent the health states of the population. The greater variation in profiles (ie, fewer people reporting 11111) could suggest the NZ sample is more representative in terms of describing the population’s HRQoL, especially if countries are considered to be similar in terms of their overall health status (eg, Germany and NZ).
      OECD. 2019
      Health at a Glance 2019: OECD Indicators.
      OECD Publishing, Paris, France2019
      Nevertheless, as discussed below, the mean EQ-VAS scores do not necessarily equate to mean utility values. For example, in the South Australian study,
      • McCaffrey N.
      • Kaambwa B.
      • Currow D.C.
      • Ratcliffe J.
      Health-related quality of life measured using the EQ-5D-5L: South Australian population norms.
      Health Qual Life Outcomes. 2016; 14: 133
      42.8% of participants reported “no problems” on the EQ-5D descriptive system but only 7.2% of participants reported >90 on the EQ-VAS. Further work is needed to determine how much variance is caused by ceiling effects, how much is caused by selection bias of healthier people into the samples, and how much could be a real effect.
      When broken down by dimension, pain/discomfort (61.7%) and anxiety/depression (46.4%) had the highest proportions of participants reporting problems, followed by mobility (27.9%), usual activities (29.8%), and self-care (8.6%). Similar patterns were found in China,
      • Yang Z.
      • Busschbach J.
      • Liu G.
      • Luo N.
      EQ-5D-5L norms for the urban Chinese population in China.
      Health Qual Life Outcomes. 2018; 16: 210
      Hong Kong,
      • Wong E.L.Y.
      • Cheung A.W.L.
      • Wong A.Y.K.
      • Xu R.H.
      • Ramos-Goni J.M.
      • Rivero-Arias O.
      Normative profile of health-related quality of life for Hong Kong general population using preference-based instrument EQ-5D-5L.
      Value Health. 2019; 22: 916-924
      and Quebec.
      • Poder T.G.
      • Carrier N.
      • Kouakou C.R.C.
      Quebec health-related quality-of-life population norms using the EQ-5D-5L: decomposition by sociodemographic data and health problems.
      Value Health. 2020; 23: 251-259
      Interestingly, although the NZ, Quebecer,
      • Poder T.G.
      • Carrier N.
      • Kouakou C.R.C.
      Quebec health-related quality-of-life population norms using the EQ-5D-5L: decomposition by sociodemographic data and health problems.
      Value Health. 2020; 23: 251-259
      and Spanish
      • Garcia-Gordillo M.A.
      • Adsuar J.C.
      • Olivares P.R.
      Normative values of EQ-5D-5L: in a Spanish representative population sample from Spanish health survey, 2011.
      Qual Life Res. 2016; 25: 1313-1321
      populations recorded a high proportion (61.7%, 67.9%, and 71.7% respectively) of people with pain/discomfort, other countries reported low proportions of people with pain/discomfort, ranging from 10.0% to 44.4%.
      • Yang Z.
      • Busschbach J.
      • Liu G.
      • Luo N.
      EQ-5D-5L norms for the urban Chinese population in China.
      Health Qual Life Outcomes. 2018; 16: 210
      , 
      • Grochtdreis T.
      • Dams J.
      • Konig H.H.
      • Konnopka A.
      Health-related quality of life measured with the EQ-5D-5L: estimation of normative index values based on a representative German population sample and value set.
      Eur J Health Econ. 2019; 20: 933-944
      , 
      • Wong E.L.Y.
      • Cheung A.W.L.
      • Wong A.Y.K.
      • Xu R.H.
      • Ramos-Goni J.M.
      • Rivero-Arias O.
      Normative profile of health-related quality of life for Hong Kong general population using preference-based instrument EQ-5D-5L.
      Value Health. 2019; 22: 916-924
      , 
      • Scalone L.
      • Cortesi P.A.
      • Ciampichini R.
      • Cesana G.
      • Mantovani L.G.
      Health related quality of life norm data of the Italian general population: results using the EQ-5D-3L and EQ-5D-5L instruments.
      Epidemiol Biostat Public Health. 2015; 12
      ,
      • Golicki D.
      • Niewada M.
      EQ-5D-5L Polish population norms.
      Arch Med Sci. 2017; 13: 191-200
      ,
      • McCaffrey N.
      • Kaambwa B.
      • Currow D.C.
      • Ratcliffe J.
      Health-related quality of life measured using the EQ-5D-5L: South Australian population norms.
      Health Qual Life Outcomes. 2016; 14: 133
      ,
      • Nguyen L.H.
      • Tran B.X.
      • Le Q.N.H.
      • Tran T.T.
      • Latkin C.A.
      Quality of life profile of general Vietnamese population using EQ-5D-5L.
      Health Qual Life Outcomes. 2017; 15: 199
      Consistent with the population norms for China,
      • Yang Z.
      • Busschbach J.
      • Liu G.
      • Luo N.
      EQ-5D-5L norms for the urban Chinese population in China.
      Health Qual Life Outcomes. 2018; 16: 210
      Japan,
      • Shiroiwa T.
      • Fukuda T.
      • Ikeda S.
      • et al.
      Japanese population norms for preference-based measures: EQ-5D-3L, EQ-5D-5L, and SF-6D.
      Qual Life Res. 2016; 25: 707-719
      and Hong Kong,
      • Wong E.L.Y.
      • Cheung A.W.L.
      • Wong A.Y.K.
      • Xu R.H.
      • Ramos-Goni J.M.
      • Rivero-Arias O.
      Normative profile of health-related quality of life for Hong Kong general population using preference-based instrument EQ-5D-5L.
      Value Health. 2019; 22: 916-924
      the younger age groups were more likely to report problems with anxiety/depression. Nevertheless, this pattern was not observed in some regions, such as Quebec, Singapore, South Australia, Spain, and Vietnam.
      • Poder T.G.
      • Carrier N.
      • Kouakou C.R.C.
      Quebec health-related quality-of-life population norms using the EQ-5D-5L: decomposition by sociodemographic data and health problems.
      Value Health. 2020; 23: 251-259
      , 
      • Abdin E.
      • Subramaniam M.
      • Vaingankar J.A.
      • Luo N.
      • Chong S.A.
      Measuring health-related quality of life among adults in Singapore: population norms for the EQ-5D.
      Qual Life Res. 2013; 22: 2983-2991
      , 
      • McCaffrey N.
      • Kaambwa B.
      • Currow D.C.
      • Ratcliffe J.
      Health-related quality of life measured using the EQ-5D-5L: South Australian population norms.
      Health Qual Life Outcomes. 2016; 14: 133
      , 
      • Garcia-Gordillo M.A.
      • Adsuar J.C.
      • Olivares P.R.
      Normative values of EQ-5D-5L: in a Spanish representative population sample from Spanish health survey, 2011.
      Qual Life Res. 2016; 25: 1313-1321
      , 
      • Nguyen L.H.
      • Tran B.X.
      • Le Q.N.H.
      • Tran T.T.
      • Latkin C.A.
      Quality of life profile of general Vietnamese population using EQ-5D-5L.
      Health Qual Life Outcomes. 2017; 15: 199
      It is difficult to compare EQ-VAS scores with utility values. For example, although 543 participants (22%) reported no problems on all dimensions (ie, 11111), only 60 (2.4%) reported a “perfect” EQ-VAS score (100). This finding may, at least in part, be underpinned by there being “other” aspects of health beyond the 5 EQ-5D dimensions that matter to people and reflected in the EQ-VAS. Of interest (concern) are the 37 participants (1.5%) with a negative utility value—meaning that, on the day they completed the survey, their health state could be categorized as worse than dead.
      Nevertheless, consistent with findings from other countries, the NZ utility values and EQ-VAS scores were associated with sociodemographic characteristics, including age, employment status, long-term disability, and chronic illness.
      • Yang Z.
      • Busschbach J.
      • Liu G.
      • Luo N.
      EQ-5D-5L norms for the urban Chinese population in China.
      Health Qual Life Outcomes. 2018; 16: 210
      ,
      • Grochtdreis T.
      • Dams J.
      • Konig H.H.
      • Konnopka A.
      Health-related quality of life measured with the EQ-5D-5L: estimation of normative index values based on a representative German population sample and value set.
      Eur J Health Econ. 2019; 20: 933-944
      ,
      • Poder T.G.
      • Carrier N.
      • Kouakou C.R.C.
      Quebec health-related quality-of-life population norms using the EQ-5D-5L: decomposition by sociodemographic data and health problems.
      Value Health. 2020; 23: 251-259
      ,
      • McCaffrey N.
      • Kaambwa B.
      • Currow D.C.
      • Ratcliffe J.
      Health-related quality of life measured using the EQ-5D-5L: South Australian population norms.
      Health Qual Life Outcomes. 2016; 14: 133
      ,
      • Garcia-Gordillo M.A.
      • Adsuar J.C.
      • Olivares P.R.
      Normative values of EQ-5D-5L: in a Spanish representative population sample from Spanish health survey, 2011.
      Qual Life Res. 2016; 25: 1313-1321
      In particular, and unsurprisingly, a long-term disability or chronic illness was found to detrimentally impact a person’s HRQoL. Nevertheless, contrary to findings from countries in the Asian region, including Hong Kong,
      • Wong E.L.Y.
      • Cheung A.W.L.
      • Wong A.Y.K.
      • Xu R.H.
      • Ramos-Goni J.M.
      • Rivero-Arias O.
      Normative profile of health-related quality of life for Hong Kong general population using preference-based instrument EQ-5D-5L.
      Value Health. 2019; 22: 916-924
      Japan,
      • Shiroiwa T.
      • Fukuda T.
      • Ikeda S.
      • et al.
      Japanese population norms for preference-based measures: EQ-5D-3L, EQ-5D-5L, and SF-6D.
      Qual Life Res. 2016; 25: 707-719
      and Singapore,
      • Abdin E.
      • Subramaniam M.
      • Vaingankar J.A.
      • Luo N.
      • Chong S.A.
      Measuring health-related quality of life among adults in Singapore: population norms for the EQ-5D.
      Qual Life Res. 2013; 22: 2983-2991
      compared with people aged 18 to 24 years, older age groups had higher utility and EQ-VAS scores (except for the 45-54 years age group with respect to utility and the 25-44 years age group with respect to EQ-VAS), with the greatest difference in people aged ≥65 years. A possible reason is the proportionately higher number of young people reporting problems on the anxiety/depression dimension (eg, 69.4% for 18- to 24-year-olds compared with 26.3% of the ≥65-year-olds), which is reflective of NZ population statistics.
      Ministry of Health
      Annual data Explorer 2018/19: New Zealand health survey [data file].
      http://minhealthnz.shinyapps.io/nz-health-survey-2018-19-annual-data-explorer/
      Date: 2019
      Date accessed: September 16, 2020
      Our findings also revealed that people who were not in paid work had lower HRQoL than the NZ general population, which is consistent with previous studies.
      • Wong E.L.Y.
      • Cheung A.W.L.
      • Wong A.Y.K.
      • Xu R.H.
      • Ramos-Goni J.M.
      • Rivero-Arias O.
      Normative profile of health-related quality of life for Hong Kong general population using preference-based instrument EQ-5D-5L.
      Value Health. 2019; 22: 916-924
      ,
      • McCaffrey N.
      • Kaambwa B.
      • Currow D.C.
      • Ratcliffe J.
      Health-related quality of life measured using the EQ-5D-5L: South Australian population norms.
      Health Qual Life Outcomes. 2016; 14: 133
      ,
      • Nguyen L.H.
      • Tran B.X.
      • Le Q.N.H.
      • Tran T.T.
      • Latkin C.A.
      Quality of life profile of general Vietnamese population using EQ-5D-5L.
      Health Qual Life Outcomes. 2017; 15: 199
      Marital status has been shown to affect HRQoL
      • Wong E.L.Y.
      • Cheung A.W.L.
      • Wong A.Y.K.
      • Xu R.H.
      • Ramos-Goni J.M.
      • Rivero-Arias O.
      Normative profile of health-related quality of life for Hong Kong general population using preference-based instrument EQ-5D-5L.
      Value Health. 2019; 22: 916-924
      ,
      • Scalone L.
      • Cortesi P.A.
      • Ciampichini R.
      • Cesana G.
      • Mantovani L.G.
      Health related quality of life norm data of the Italian general population: results using the EQ-5D-3L and EQ-5D-5L instruments.
      Epidemiol Biostat Public Health. 2015; 12
      ,
      • McCaffrey N.
      • Kaambwa B.
      • Currow D.C.
      • Ratcliffe J.
      Health-related quality of life measured using the EQ-5D-5L: South Australian population norms.
      Health Qual Life Outcomes. 2016; 14: 133
      but data on this characteristic were not collected in this study. Nevertheless, we collected information about whether people lived alone or with others and found a tendency toward reduced problems with self-care for people living with others compared with those living alone.
      Overall, the difference in profiles, utility values, and EQ-VAS scores across countries is unsurprising. Countries and cultures regard and report health differently, highlighting the importance of each country having their own value set.
      • Szende A.
      • Janssen B.
      • Cabases J.
      Self-Reported Population Health: An International Perspective Based on EQ-5D.
      Springer Nature, Dordrecht, the Netherlands2014
      Relative to population data for other countries, including China,
      • Yang Z.
      • Busschbach J.
      • Liu G.
      • Luo N.
      EQ-5D-5L norms for the urban Chinese population in China.
      Health Qual Life Outcomes. 2018; 16: 210
      Hong Kong,
      • Wong E.L.Y.
      • Cheung A.W.L.
      • Wong A.Y.K.
      • Xu R.H.
      • Ramos-Goni J.M.
      • Rivero-Arias O.
      Normative profile of health-related quality of life for Hong Kong general population using preference-based instrument EQ-5D-5L.
      Value Health. 2019; 22: 916-924
      Poland,
      • Golicki D.
      • Niewada M.
      EQ-5D-5L Polish population norms.
      Arch Med Sci. 2017; 13: 191-200
      South Australia,
      • McCaffrey N.
      • Kaambwa B.
      • Currow D.C.
      • Ratcliffe J.
      Health-related quality of life measured using the EQ-5D-5L: South Australian population norms.
      Health Qual Life Outcomes. 2016; 14: 133
      and Vietnam,
      • Nguyen L.H.
      • Tran B.X.
      • Le Q.N.H.
      • Tran T.T.
      • Latkin C.A.
      Quality of life profile of general Vietnamese population using EQ-5D-5L.
      Health Qual Life Outcomes. 2017; 15: 199
      the NZ sample has more participants with “low education” and “high education.” Nevertheless, it is difficult to directly compare data as each country categorizes education differently. For example, China
      • Yang Z.
      • Busschbach J.
      • Liu G.
      • Luo N.
      EQ-5D-5L norms for the urban Chinese population in China.
      Health Qual Life Outcomes. 2018; 16: 210
      categorizes educational attainment as “primary or lower,” “junior and senior high school,” and “college or higher”—similar to Hong Kong
      • Wong E.L.Y.
      • Cheung A.W.L.
      • Wong A.Y.K.
      • Xu R.H.
      • Ramos-Goni J.M.
      • Rivero-Arias O.
      Normative profile of health-related quality of life for Hong Kong general population using preference-based instrument EQ-5D-5L.
      Value Health. 2019; 22: 916-924
      —whereas South Australia’s
      • McCaffrey N.
      • Kaambwa B.
      • Currow D.C.
      • Ratcliffe J.
      Health-related quality of life measured using the EQ-5D-5L: South Australian population norms.
      Health Qual Life Outcomes. 2016; 14: 133
      categories are “up to secondary school,” “trade/certificate/diploma,” and “degree or higher,” which more closely aligns with the categories used in our study. The NZ population sample includes more unemployed people than Italy,
      • Scalone L.
      • Cortesi P.A.
      • Ciampichini R.
      • Cesana G.
      • Mantovani L.G.
      Health related quality of life norm data of the Italian general population: results using the EQ-5D-3L and EQ-5D-5L instruments.
      Epidemiol Biostat Public Health. 2015; 12
      South Australia,
      • McCaffrey N.
      • Kaambwa B.
      • Currow D.C.
      • Ratcliffe J.
      Health-related quality of life measured using the EQ-5D-5L: South Australian population norms.
      Health Qual Life Outcomes. 2016; 14: 133
      and Vietnam.
      • Nguyen L.H.
      • Tran B.X.
      • Le Q.N.H.
      • Tran T.T.
      • Latkin C.A.
      Quality of life profile of general Vietnamese population using EQ-5D-5L.
      Health Qual Life Outcomes. 2017; 15: 199
      Moreover, more than 50% of the NZ sample have a chronic health condition, in contrast to 6.3% and 30.3% of the Vietnamese
      • Nguyen L.H.
      • Tran B.X.
      • Le Q.N.H.
      • Tran T.T.
      • Latkin C.A.
      Quality of life profile of general Vietnamese population using EQ-5D-5L.
      Health Qual Life Outcomes. 2017; 15: 199
      and Hong Kong populations,
      • Wong E.L.Y.
      • Cheung A.W.L.
      • Wong A.Y.K.
      • Xu R.H.
      • Ramos-Goni J.M.
      • Rivero-Arias O.
      Normative profile of health-related quality of life for Hong Kong general population using preference-based instrument EQ-5D-5L.
      Value Health. 2019; 22: 916-924
      respectively—bearing in mind that unemployment and chronic health conditions may be defined differently in countries with different social services and health systems and therefore may not be strictly comparable.
      The different methods used to create value sets and/or the populations surveyed could also explain the variation in mean utility values between countries. The NZ EQ-5D-5L value set was created using the potentially all pairwise rankings of all possible alternatives method, which is a type of adaptive DCE (see Hansen and Ombler
      • Hansen P.
      • Ombler F.
      A new method for scoring additive multi-attribute value models using pairwise rankings of alternatives.
      J Multi Criteria Dec Anal. 2008; 15: 87-107
      2008 for details), and a binary search algorithm to identify states worse than dead, implemented by 1000minds software (www.1000minds.com). Other countries used the “EuroQol Valuation Technology,”
      • Oppe M.
      • Devlin N.J.
      • van Hout B.
      • Krabbe P.F.
      • de Charro F.
      A program of methodological research to arrive at the new international EQ-5D-5L valuation protocol.
      Value Health. 2014; 17: 445-453
      with some countries using hybrid models incorporating both time trade-off and DCE data from the EuroQol Valuation Technology to construct their value sets (eg, Germany, Hong Kong, Poland, and Spain
      • Ludwig K.
      • von der Schulenburg J.M.
      • Greiner W.
      German value set for the EQ-5D-5L.
      Pharmacoecon. 2018; 36: 663-674
      , 
      • Wong E.L.
      • Ramos-Goni J.M.
      • Cheung A.W.
      • Wong A.Y.
      • Rivero-Arias O.
      Assessing the use of a feedback module to model EQ-5D-5L health states values in Hong Kong.
      Patient Patient Centered Outcomes Res. 2018; 11: 235-247
      , 
      • Golicki D.
      • Jakubczyk M.
      • Graczyk K.
      • Niewada M.
      Valuation of EQ-5D-5L health states in Poland: the first EQ-VT-based study in Central and Eastern Europe.
      Pharmacoecon. 2019; 37: 1165-1176
      , 
      • Ramos-Goñi J.M.
      • Craig B.M.
      • Oppe M.
      • et al.
      Handling data quality issues to estimate the Spanish EQ-5D-5L value set using a hybrid interval regression approach.
      Value Health. 2018; 21: 596-604.31
      ).
      Similarly, comparing NZ’s EQ-5D-3L population norms with the newly derived EQ-5D-5L population norms is not practicable given the fundamental differences between the value sets (ie, 3 vs 5 levels), the different methods used to create the 2 value sets (VAS vs DCE), the number of participants (396 vs 2468), and the fact that the EQ-5D-3L value set was created more than 20 years ago. Nevertheless, an important strength of the EQ-5D-5L study relative to the earlier EQ-5D-3L study is the higher proportion of Māori (the indigenous population of NZ) who took part in the survey: 15.8% versus 7.6%.
      The strength of this study lies in its unique study design, whereby a large sample (n = 2468) of personal EQ-5D-5L value sets representative of the NZ general population was used to construct population norms and individual and mean utility values and EQ-VAS scores. Given it has been more than 20 years since a NZ EQ-5D value set (ie, the EQ-5D-3L) and associated population norms was created,
      • Devlin N.J.
      • Hansen P.
      • Kind P.
      • Williams A.
      Logical inconsistencies in survey respondents’ health state valuations-a methodological challenge for estimating social tariffs.
      Health Econ. 2003; 12: 529-544
      the newly derived NZ EQ-5D-5L population norms will be useful for healthcare funders, providers, researchers, and policy makers. Population norms can inform policy making and assist decision making, including valuing HRQoL in economic evaluations, comparing population groups, and identifying disease burden both within NZ and internationally.
      • Devlin N.J.
      • Brooks R.
      EQ-5D and the EuroQol Group: past, present and future.
      Appl Health Econ Health Pol. 2017; 15: 127-137
      This study has identified particular sociodemographic factors that influence the HRQoL of the NZ population. Further research is currently underway to investigate the relationship more fully between individual utility and sociodemographic characteristics.
      Similar to other studies of this type, a potential limitation is the possibility of selection bias when conducting online surveys. Although the sample is generally representative of the NZ population in terms of the main sociodemographic characteristics, there may have been selection bias in recruiting the sample—for example, people without access or the confidence to complete an online survey or people who were unwell at the time. Potential bias can also arise when variables, such as income and education, are based on self-report. In this study, participants’ identities were kept anonymous with the intention of mitigating such bias. Although difficult to compare directly with national statistics (given the disparity in age groups), there were proportionately more people with higher education (and higher income) in this study. Because utility values and EQ-VAS scores are known to increase with education level,
      • Grochtdreis T.
      • Dams J.
      • Konig H.H.
      • Konnopka A.
      Health-related quality of life measured with the EQ-5D-5L: estimation of normative index values based on a representative German population sample and value set.
      Eur J Health Econ. 2019; 20: 933-944
      ,
      • Wong E.L.Y.
      • Cheung A.W.L.
      • Wong A.Y.K.
      • Xu R.H.
      • Ramos-Goni J.M.
      • Rivero-Arias O.
      Normative profile of health-related quality of life for Hong Kong general population using preference-based instrument EQ-5D-5L.
      Value Health. 2019; 22: 916-924
      ,
      • Poder T.G.
      • Carrier N.
      • Kouakou C.R.C.
      Quebec health-related quality-of-life population norms using the EQ-5D-5L: decomposition by sociodemographic data and health problems.
      Value Health. 2020; 23: 251-259
      the overrepresentation of people with high education could influence the results. Finally, this study used cross-sectional data which provides a snapshot of the NZ general population at a certain point in time—thus, restricting the investigation of casual relationships between HRQoL and other sociodemographic characteristics over time.

      Conclusions

      This study reports the first NZ population norms for the EQ-5D-5L, which, for the first time ever, were constructed from the personal EQ-5D-5L value sets of a large and representative sample. Consistent with other countries’ population norms, the NZ population’s EQ-5D-5L values and EQ-VAS scores are associated with age, employment status, long-term disability, and chronic illness. The results from this study will support health funders and policy makers in their resource allocation decision making and help researchers and policy makers to better understand the HRQoL of the NZ population.

      Article and Author Information

      Author Contributions: Concept and design: Sullivan, Turner, Derrett, Hansen
      Acquisition of data: Sullivan, Derrett, Hansen
      Analysis and interpretation of data: Sullivan, Turner, Derrett
      Drafting of the manuscript: Sullivan, Turner, Hansen
      Critical revision of the paper for important intellectual content: Sullivan, Turner, Derrett, Hansen
      Statistical analysis: Turner
      Obtaining funding: Sullivan
      Administrative, technical, or logistical support: Hansen
      Conflict of Interest Disclosures: Drs Sullivan and Derrett are members of the EuroQol Group. Dr Hansen is a coinventor and co-owner of the 1000minds software mentioned in the article and used for the discrete choice experiment survey that the data in the article are derived from. Dr Hansen also reported multiple patents issued (#7552104, 200423, 526447, 607753, and 527785) in the United States, Australia, and New Zealand for the PAPRIKA method. No other disclosures were reported.
      Funding/Support: The Department of Preventive and Social Medicine, University of Otago provided funding for data collection and software administration; 1000minds Ltd supplied the software; the Health Research Council of New Zealand provided funding for the analysis (HRC19/640).
      Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

      Acknowledgment

      Thank you to Hui Yee Yeo and Georgia McCarty for their comments on the paper.

      Supplemental Material

      References

        • Brazier J.
        • Ratcliffe J.
        • Saloman J.
        • Tsuchiya A.
        Measuring and Valuing Health Benefits for Economic Evaluation.
        Oxford University Press, Oxford, United Kingdom2017
        View in Article
        • Devlin N.
        • Parkin D.
        • Janssen B.
        Methods for Analysing and Reporting EQ-5D Data.
        Springer Nature, Cham, Switzerland2020
        View in Article
        • Pharmaceutical Management Agency (PHARMAC)
        Guidelines for Funding Applications to PHARMAC.
        PHARMAC, 2019
        https://pharmac.govt.nz/medicine-funding-and-supply/make-an-application/pharmconnect-make-a-medicine-funding-application/pharmconnect-supporting-information/#guidelines
        View in Article
        • Shuker C.
        • Bohm G.
        • Hamblin R.
        • et al.
        Progress in public reporting in New Zealand since the Ombudsman’s ruling, and an invitation.
        N Z Med J. 2017; 130: 11-22
        View in Article
        • Devlin N.J.
        • Hansen P.
        • Kind P.
        • Williams A.
        Logical inconsistencies in survey respondents’ health state valuations-a methodological challenge for estimating social tariffs.
        Health Econ. 2003; 12: 529-544
        View in Article
        • Szende A.
        • Janssen B.
        • Cabases J.
        Self-Reported Population Health: An International Perspective Based on EQ-5D.
        Springer Nature, Dordrecht, the Netherlands2014
        View in Article
        • Van den Berg B.
        SF−6D population norms.
        Health Econ. 2012; 21: 1508-1512
        View in Article
        • Herdman M.
        • Gudex C.
        • Lloyd A.
        • et al.
        Development and preliminary testing of the new five-level version of EQ-5D (EQ-5D-5L).
        Qual Life Res. 2011; 20: 1727-1736
        View in Article
        • Sullivan T.
        • Hansen P.
        • Ombler F.
        • Derrett S.
        • Devlin N.
        A new tool for creating personal and social EQ-5D-5L value sets, including valuing ‘dead’.
        Soc Sci Med. 2020; 246112707
        View in Article
        • Yang Z.
        • Busschbach J.
        • Liu G.
        • Luo N.
        EQ-5D-5L norms for the urban Chinese population in China.
        Health Qual Life Outcomes. 2018; 16: 210
        View in Article
        • Grochtdreis T.
        • Dams J.
        • Konig H.H.
        • Konnopka A.
        Health-related quality of life measured with the EQ-5D-5L: estimation of normative index values based on a representative German population sample and value set.
        Eur J Health Econ. 2019; 20: 933-944
        View in Article
        • Wong E.L.Y.
        • Cheung A.W.L.
        • Wong A.Y.K.
        • Xu R.H.
        • Ramos-Goni J.M.
        • Rivero-Arias O.
        Normative profile of health-related quality of life for Hong Kong general population using preference-based instrument EQ-5D-5L.
        Value Health. 2019; 22: 916-924
        View in Article
        • Scalone L.
        • Cortesi P.A.
        • Ciampichini R.
        • Cesana G.
        • Mantovani L.G.
        Health related quality of life norm data of the Italian general population: results using the EQ-5D-3L and EQ-5D-5L instruments.
        Epidemiol Biostat Public Health. 2015; 12
        View in Article
        • Shiroiwa T.
        • Fukuda T.
        • Ikeda S.
        • et al.
        Japanese population norms for preference-based measures: EQ-5D-3L, EQ-5D-5L, and SF-6D.
        Qual Life Res. 2016; 25: 707-719
        View in Article
        • Golicki D.
        • Niewada M.
        EQ-5D-5L Polish population norms.
        Arch Med Sci. 2017; 13: 191-200
        View in Article
        • Poder T.G.
        • Carrier N.
        • Kouakou C.R.C.
        Quebec health-related quality-of-life population norms using the EQ-5D-5L: decomposition by sociodemographic data and health problems.
        Value Health. 2020; 23: 251-259
        View in Article
        • Abdin E.
        • Subramaniam M.
        • Vaingankar J.A.
        • Luo N.
        • Chong S.A.
        Measuring health-related quality of life among adults in Singapore: population norms for the EQ-5D.
        Qual Life Res. 2013; 22: 2983-2991
        View in Article
        • McCaffrey N.
        • Kaambwa B.
        • Currow D.C.
        • Ratcliffe J.
        Health-related quality of life measured using the EQ-5D-5L: South Australian population norms.
        Health Qual Life Outcomes. 2016; 14: 133
        View in Article
        • Garcia-Gordillo M.A.
        • Adsuar J.C.
        • Olivares P.R.
        Normative values of EQ-5D-5L: in a Spanish representative population sample from Spanish health survey, 2011.
        Qual Life Res. 2016; 25: 1313-1321
        View in Article
        • Nguyen L.H.
        • Tran B.X.
        • Le Q.N.H.
        • Tran T.T.
        • Latkin C.A.
        Quality of life profile of general Vietnamese population using EQ-5D-5L.
        Health Qual Life Outcomes. 2017; 15: 199
        View in Article
        • EuroQol Research Foundation
        EQ-5D-5L User Guide.
        2019
        https://euroqol.org/publications/user-guides
        View in Article
        • Statistics New Zealand
        Disability.
        https://www.stats.govt.nz/information-releases/disability-survey-2013
        Date: 2013
        Date accessed: September 16, 2020
        View in Article
        • Ministry of Health
        Annual data Explorer 2018/19: New Zealand health survey [data file].
        http://minhealthnz.shinyapps.io/nz-health-survey-2018-19-annual-data-explorer/
        Date: 2019
        Date accessed: September 16, 2020
        View in Article
        • Stanley J.
        • Semper K.
        • Millar E.
        • Sarfati D.
        Epidemiology of multimorbidity in New Zealand: a cross-sectional study using national-level hospital and pharmaceutical data.
        BMJ Open. 2018; 8
        View in Article
        • Austin P.C.
        • Escobar M.
        • Kopec J.A.
        The use of the Tobit model for analyzing measures of health status.
        Qual Life Res. 2000; 9: 901-910
        View in Article
        • Ministry of Education
        Education counts.
        (URL:)
        https://www.educationcounts.govt.nz/indicators/main/student-engagement-participation/retention_of_students_in_senior_secondary_schools
        Date: 2020
        Date accessed: September 16, 2020
        View in Article
        • OECD. 2019
        Health at a Glance 2019: OECD Indicators.
        OECD Publishing, Paris, France2019
        View in Article
        • Hansen P.
        • Ombler F.
        A new method for scoring additive multi-attribute value models using pairwise rankings of alternatives.
        J Multi Criteria Dec Anal. 2008; 15: 87-107
        View in Article
        • Oppe M.
        • Devlin N.J.
        • van Hout B.
        • Krabbe P.F.
        • de Charro F.
        A program of methodological research to arrive at the new international EQ-5D-5L valuation protocol.
        Value Health. 2014; 17: 445-453
        View in Article
        • Ludwig K.
        • von der Schulenburg J.M.
        • Greiner W.
        German value set for the EQ-5D-5L.
        Pharmacoecon. 2018; 36: 663-674
        View in Article
        • Wong E.L.
        • Ramos-Goni J.M.
        • Cheung A.W.
        • Wong A.Y.
        • Rivero-Arias O.
        Assessing the use of a feedback module to model EQ-5D-5L health states values in Hong Kong.
        Patient Patient Centered Outcomes Res. 2018; 11: 235-247
        View in Article
        • Golicki D.
        • Jakubczyk M.
        • Graczyk K.
        • Niewada M.
        Valuation of EQ-5D-5L health states in Poland: the first EQ-VT-based study in Central and Eastern Europe.
        Pharmacoecon. 2019; 37: 1165-1176
        View in Article
        • Ramos-Goñi J.M.
        • Craig B.M.
        • Oppe M.
        • et al.
        Handling data quality issues to estimate the Spanish EQ-5D-5L value set using a hybrid interval regression approach.
        Value Health. 2018; 21: 596-604.31
        View in Article
        • Devlin N.J.
        • Brooks R.
        EQ-5D and the EuroQol Group: past, present and future.
        Appl Health Econ Health Pol. 2017; 15: 127-137
        View in Article

      Article info

      Publication history

      Published online: June 26, 2021
      Accepted: April 5, 2021

      Identification

      DOI: https://doi.org/10.1016/j.jval.2021.04.1280

      Copyright

      © 2021 ISPOR-The Professional Society for Health Economics and Outcomes Research. Published by Elsevier Inc.

      User license

      Creative Commons Attribution – NonCommercial – NoDerivs (CC BY-NC-ND 4.0) |
      How you can reuse Information Icon
      Creative Commons Attribution – NonCommercial – NoDerivs (CC BY-NC-ND 4.0)

      Permitted

      For non-commercial purposes:

      • Read, print & download
      • Redistribute or republish the final article
      • Text & data mine
      • Translate the article (private use only, not for distribution)
      • Reuse portions or extracts from the article in other works

      Not Permitted

      • Sell or re-use for commercial purposes
      • Distribute translations or adaptations of the article

      Elsevier's open access license policy

      ScienceDirect

      Access this article on ScienceDirect
      We use cookies to help provide and enhance our service and tailor content. To update your cookie settings, please visit the for this site.
      Copyright © 2023 Elsevier Inc. except certain content provided by third parties. The content on this site is intended for healthcare professionals.


      RELX