Skip to main content
Download PDF

EQ-5D-3L and EQ-5D-5L population norms for Thailand

BMC Public Health volume 24, Article number: 1108 (2024) Cite this article

Abstract

Background

The previous Thai norm-based scores for the EQ-5D-5L were established with Thai general population samples aged 20–70 years in 2019. Nevertheless, these values need to be updated after the COVID-19 pandemic because of its effects on both physical and mental health. This study therefore aimed to establish population norms of the Thai EQ-5D-3L, EQ-5D-5L and EQ-VAS scores as well as to estimate disutility values associated with self-reported main diseases.

Methods

Individual face-to-face interviews were conducted with 2000 adult (age ≥ 18 years) members of the general Thai population to estimate norm-based scores. Each participant completed the EQ-5D-3L and EQ-5D-5L as well as questions related to their sociodemographic factors and self-reported main diseases. Responses to the two instruments were converted to health utility (HU) scores on the basis of available value sets. Descriptive statistics were used to report the norm-based scores stratified by age and sex categories. Response redistribution determining the response consistency between EQ-5D versions was investigated. The HU score agreement from those two instruments was investigated using intraclass correlation coefficient (ICC). Tobit regression models were employed to investigate the relationships between sociodemographic factors and HU and EQ-VAS scores. Moreover, it was used to estimate the disutility values associated with self-reported main diseases.

Results

The means (percentage of ceiling effects) of EQ-5D-3L, EQ-5D-5L, and EQ-VAS scores were 0.845 (57.80%), 0.923 (49.05%), and 79.83 (3.20%), respectively. The average percentage of inconsistent response was 1.09%. A good agreement level was found between both EQ-5D versions with the ICCs of 0.789 (95% CI: 0.558–0.878). Female, older, and unemployed participants and those with BMI ≥ 30 reported lower EQ-5D-3L and EQ-5D-5L than their counterparts. Bone/Joint disorder and stroke contributed to the largest disutility value for those two instruments.

Conclusions

The Thai norm-based scores from those two instruments were diminished when advancing age and among female, unemployed, and obese (BMI ≥ 30) participants. It is expected to provide information to policy makers to better allocate health care resources to those with diminished norm-based scores.

Peer Review reports

Introduction

The EQ-5D is a brief self-completion instrument developed by the EuroQoL group in the 1980s [1]. It has been proven to be a valid and reliable tool for HRQoL measurement in both general population samples and in clinical contexts [2, 3]. As a result, the EQ-5D is commonly used to compute HU scores using country-specific value sets for economic analyses [4, 5]. The first version, EQ-5D-3L, has five questions, one for each of the following dimensions: mobility (MO), self-care (SC), usual activities (UA), pain/discomfort (PD), and anxiety/depression (AD). Each dimension has three response options including no problem, some/moderate problem, and extreme problem [6]. A newer version, EQ-5D-5L, still has five questions by retaining the original five dimensions, but expanding the response options from three to five levels for each dimension as follows: no problem, slight problem, moderate problem, severe problem, and extreme problem/unable to perform. These modifications resulted in the EQ-5D-5L being able to cover a wider range of health states and having enhanced discriminatory power for use in both the general public and in therapeutic settings [2, 3].

EQ-5D-3L population norms have been reported for Japan [7], six European countries (Belgium, France, Germany, Italy, the Netherlands, and Spain) [8], while EQ-5D-5L population norms have been also reported for a number of Asian, European, American, and African countries including South Australia [9], Barbados and Jamaica [10], Belgium [11] Bulgaria [12], Canada [13], China [14], Hong Kong [15], Columbia [16], Denmark [17], England [18], France [19], Germany [20], Indonesia [21], Iran [22], Ireland [23], Japan [24], New Zealand [25], Norway [26], Poland [27], Russia [28], Slovenia [29], Spain [30], Sweden [31], Trinidad and Tobago [32], the United States [33], and Vietnam [34].

In Thailand, the EQ-5D has been widely used to assess and evaluate health interventions, and it has also been strongly recommended for economic analyses by the recent Thai Health Technology Assessment Guidelines [35]. The EQ-5D-5L has been proven to be a practical, reliable, and valid instrument, with better discriminatory power than the EQ-5D-3L in the general population [3] and in therapeutic settings [36, 37]. However, both versions have been used to generate HU scores in patients with chronic diseases, especially for cancer patients [38], because both versions have a Thai-specific value set to compute the HU scores [39, 40]. As a result, there is clear need to establish population norms for the EQ-5D-3L and EQ-5D-5Lto benchmark the HU scores for evaluating population health care, health equity, and health care interventions among the Thai population.

To the best of our knowledge, only one study [41] has been conducted to estimate and report EQ-5D-5L population norms for the general Thai population. However, these population norms were developed based on a sample with a limited age range (20–70 years of age) to be considered representative for a national survey because the samples with some age groups were not included especially for the aged >70 years of age considered a vulnerable group of having diminished HRQoL level for both physical and mental health. Moreover, the previous Thai norms only sampled participants from five provinces of Thailand, which does not adequately represent the whole country, particularly since provinces from metropolitan area were not recruited. Previous evidence revealed that people living in rural and metropolitan areas could have some variations of HRQoL levels due to their different lifestyles, activities, and personal and social characteristics [42]. This is also in contrast to the original valuation study of the EQ-5D-5L in Thailand, where 12 provinces were randomly sampled [39].

Notably, the data for the previous norms [41] were collected during the pre-COVID-19 pandemic. Many studies have investigated the HRQoL measurement during and after the COVID-19 pandemic, and they revealed that people often have experienced diminished physical and mental health after the pandemic particularly due to the prevalence of depressive symptoms post COVID-19 [43,44,45,46]. There is a need to update the norms after the COVID-19 pandemic for the Thai population. Therefore, the objectives of this study were to develop and compare the HU norm-based scores obtained from the EQ-5D-3L and EQ-5D-5L with a wider age range (≥ 18 years old) and larger sample size (2,000 respondents) from 12 provinces to improve generalizability of the results and to find the associations between the sociodemographic factors and HU scores. We also aimed to determine the consistency of response redistribution between EQ-5D versions and estimate the disutility (1-HU scores) associated with self-reported main diseases to facilitate QALYs calculation for economic analyses.

Method

Study design

A cross-sectional survey study was conducted with participants (n = 2,000) from the general Thai population. We specified the region and metropolitan areas to be selected, and 12 provinces were randomly selected specific to each specified regions and area as follows:1) Bangkok Metropolitan (Bangkok, Samut Prakan and Nonthaburi), 2) Central (Chonburi and Nakhon Pathom), 3) North (Chiang Mai and Nakhon Sawan), 4) North-east (Nakhon Ratchasima, Khon Kaen and Buriram), and 5) South (Nakhon Si Thammarat and Phatthalung). A quota sampling method was employed to select the study participants in proportion to age, sex, and area of residence (metropolitan and rural areas) to represent the general Thai population. Face-to-face interviews were conducted with individuals at their home residences from May to June 2023.

Instruments

EQ-5D

The EQ-5D is a questionnaire for measuring current health state to generate HU scores for economic analyses. Each participant is asked to rate their health on the day that the questionnaire is administered. It is comprised of two parts: (1) descriptive system containing five questions for each of the five dimensions: MO, SC, UA, PD, and AD, and (2) EQ-VAS that requires respondents to rate their current health status on a 20-cm vertical line where its endpoints are labeled as worst imaginable health state at 0 and best imaginable health state at 100, resulting in a score range of 0 to 100 [1]. The responses to the descriptive system are used to compute the HU scores using a country-specific value set. At present, both Thai EQ-5D-3L and EQ-5D-5L have their own value set where the maximum score is 1.000 representing perfect health, while the minimum scores are ˗0.4540 and ˗0.4212 for the worst rates health states of the EQ-5D-3L [40] and EQ-5D-5L [39], respectively. Notably, a negative HU score indicates a health state worse than death. The EuroQoL group officially granted to use both versions in this study.

Procedure

Individuals who met the pre-defined eligibility criteria were asked to participate in the face-to-face interviews with well-trained interviewers. The eligibility criteria included (1) age ≥ 18 years, (2) understanding of Thai language and data collection process evaluated by the primary researcher or the interviewers. However, we did not recruit participants if they had been diagnosed with an acute or life-threatening illness or had cognitive impairment. The interviewers were instructed to read questions and response options without explaining the meanings to the study participants to ensure that they completed the questionnaires based on their own understanding.

Before the interviews commenced, the participant information sheet was given to each study participant, and written informed consent was obtained to document voluntary participation from each study participant. However, they could withdraw from the study at any time if they felt uncomfortable. Each participant was asked to complete the 23-item questionnaire in the following order: (1) demographic information (12 items), (2) EQ-5D-3L (5 items), (3) EQ-5D-5L (5 items) and (4) EQ-VAS (1 item). The demographic section inquired about the following aspects: sex, age, education level, health insurance scheme, occupation, average monthly household income, health conditions, main disease affecting their health, smoking, alcohol consumption, weight, and height.

Data analyses

Descriptive statistics (mean, standard deviations [SD], frequencies and percentage) were used to report the sociodemographic characteristics of the study participants where appropriate, and to compare with the sample characteristics from the original valuation EQ-5D-5L study [39]. Frequencies and percentages were used to report the responses for both EQ-5D versions. Furthermore, frequencies and percentages were employed to report the response redistribution determining the response consistency between EQ-5D versions. In order to quantify consistency, the response to EQ-5D-3L was recorded to the EQ-5D-5L response (3L5L) as follows: 1 = 1, 2 = 3, and 3 = 5, and inconsistency size was calculated from the responses to both EQ-5D versions as| 3L5L– 5L| -1, so zero or less than zero can be determined as consistency for the responses between EQ-5D versions [3, 47]. The mean EQ-VAS scores were also presented for each pair to ensure the validity of response redistribution.

The responses to the EQ-5D-3L and EQ-5D-5Lquestionnaires were converted to HU scores using the Thai value sets for both EQ-5D-3L [40] and EQ-5D-5L [39], while the EQ-VAS scores were reported in the form of 0-100. Due to high ceiling effects of the HU and EQ-VAS scores, non-parametric statistics were employed to test the differences of HU scores derived from those two instruments stratified by demographic factors including Mann-Whitney U test and Kruskal Wallis H tests for two and more than two groups categorical variables, respectively. The HU and EQ-VAS scores were also presented and reported by six age bands (< 30, 30–39, 40–49, 50–59, 60–69, and ≥ 70) and stratified by sex. Moreover, the intraclass correlation coefficient (ICC) was used to determine the agreement level of the HU scores from those two instruments. The ICC was computed with the two-way mixed-effects model based on absolute agreement, yielding an ICC ranging from 0.00 to 1.00 [48]. The ICC was classified into four levels of agreements as poor agreement (ICC < 0.50), moderate agreement (0.50 ≤ ICCs < 0.75), good agreement (0.75 ≤ ICCs < 0.90), and excellent agreement (ICCs ≥ 0.90) [48].

Due to highly skewed data for the HU and EQ-VAS scores, the associations between significant factors from the univariate analysis as independent variables and the HU and EQ-VAS scores as dependent variables were investigated using multivariable Tobit regression models [49]. Prior to running the regression analyses, Spearman correlation’s rho was employed to generate a correlation matrix of sociodemographic factors among each other to test the influence of multicollinearity [24]. Any sociodemographic variable with an absolute value of Spearman’s rho > 0.30 was considered as exhibiting collinearity with another variable [24] and were not therefore entered into the Tobit regression model. Furthermore, another Tobit regression analysis was performed to determine the size of the association of disutility value (disutility = 1-HU scores) from two instruments with the self-reported main disease. Moreover, the EQ-VAS scores were converted to HU scores where the EQ-VAS scores were divided by 100 and were used to estimate the size of disutility value similar to other HU scores. All analyses were performed using STATA 17 (StataCorp LLC, College Station, TX, USA), with a p-value of < 0.05 being considered statistically significant.

Results

Sample characteristics

Table 1 illustrates the sample characteristics of this study. The majority of participants were female (53.0%) with an average age was 46.2 ± 16.9 years, and 67.2% living in a rural area. Compared to the previous EQ-5D-5L valuation study [39], our samples shared the similar proportions of some characteristics in term of sex and some age bands.

Table 1 The sample characteristics as compared to those from the original EQ-5D-5L valuation study
Full size table

Response distribution and redistribution for both EQ-5D versions

Table 2 shows the response distribution to both EQ-5D versions. The participants rated themselves as no problems with the SC dimension with the highest percentage for both EQ-5D versions (95.50% for the EQ-5D-3L vs. 94.95% for the EQ-5D-5L) followed by MO (84.05% vs. 83.55%), UA (83.40% vs. 81.00%), AD (76.10% vs. 67.55%) and PD (68.95% vs. 61.00%) for both versions. Of note, it also showed that the percentages of reporting “no problems” for all EQ-5D-5L dimensions were lower than those of the EQ-5D-3L dimensions.

Table 2 Response distribution to both EQ-5D versions
Full size table

As shown in Table 3, most of the samples reporting level 1-EQ-5D-3L remained at the level 1-EQ-5D-5L for all EQ-5D dimensions ranging from 86.73% for AD to 98.95% for SC. Similarly, our samples rating level 2- EQ-5D-3L shifted their answers to level 2- EQ-5D-5L from 58.69% for PD to 67.71% for SC, whereas the proportion of samples ranging from 21.18% for SC to 36.72% for PD of the samples upgraded their answers to level 3-EQ-5D-5L. Our samples rated themselves as level 3-EQ-5D-3L redistributed their answers to level 4-EQ-5D-5L ranging from 11.11% for AD to 45.45% for PD, while approximately 27.78% for AD ˗ 71.43% for MO of our samples redistributed their response to level 5-EQ-5D-5L. Of the 10,000 redistribution pairs, some inconsistent pairs were observed in five dimensions: MO (n = 16, 0.16%) SC (n = 12, 0.12%), UA (n = 14, 0.14%), PD (n = 19, 0.19%), and AD (n = 48, 0.48%).

Table 3 Response redistribution from EQ-5D-3L to EQ-5D-5L
Full size table

Population norms of the EQ-5D-3L, EQ-5D-5L, and EQ-VAS

Table 4 shows the Thai population norms of the HU scores from two instruments classified by sex and age bands. As expected, the HU scores and the percentage of respondents with full health were generally higher for the younger age band (< 30 years) than other age bands. Moreover, male participants had higher EQ-5D-3L, EQ-5D-5L, and EQ-VAS scores than females for most age groups except for the participants aged < 30 years. For example, for participants aged < 30 years, the mean HU score for the EQ-5D-3L was 0.949 for males and 0.953 for females, for the EQ-5D-5L 0.972 (male) and 0.976 (female). For respondents aged ≥ 70 years, in contrast, those scores were 0.589 (male) and 0.560 (female) for the EQ-5D-3L, 0.774 (male) and 0.764 (female) for the EQ-5D-5L.

Table 4 Population norms of the EQ-5D-3L, EQ-5D-5L and EQ-VAS
Full size table

Regarding the HU scores from the two instruments, we found that the EQ-5D-5L yielded the highest HU scores for all sexes and age bands. Nevertheless, the highest percentage of participants with perfect health was observed for the EQ-5D-3L followed by the EQ-5D-5L. The ICC between the HU scores was 0.789 (95% CI: 0.558–0.878). for EQ-5D-3L/EQ-5D-5L pair.

As shown in Fig. 1, the mean EQ-5D-5L scores were higher than those of the EQ-5D-3L scores for all sexes and age bands (p < 0.05). The responses to those two instruments are presented in the Supplementary materials.

Fig. 1
figure 1

Comparison of population norms between EQ-5D-3L and EQ-5D-5L

Full size image

Relationships between norms-based scores and sociodemographic information

Table 5 conveys the results of univariate analyses between norms-based scores and sociodemographic factors. The mean EQ-5D-3L and EQ-5D-5L were 0.845 (SD = 0.204) and 0.923 (SD = 0.113). The univariate analysis showed that female and older respondents had significantly (p < 0.05) lower HU scores for those two instruments compared to males and younger participants. Moreover, the HU scores from the two instruments were affected by other sociodemographic factors including marital status, education level, health insurance scheme, monthly household income, employment status, self-reported health conditions, smoking, alcohol consumptions, and BMI (all p < 0.05). The mean EQ-VAS score was 79.83 (SD = 11.75) as also shown in Table 5. Similar to the HU scores, it was generally reduced for female and older participants, and it was affected by other sociodemographic factors, as well (all p < 0.05).

Table 5 EQ-5D-3L, EQ-5D-5L and EQ-VAS population norms by sample characteristics
Full size table

Table 6 depicts the relationships between the HU scores and sociodemographic factors using Tobit regression models. Analysis 1 included sex, age bands, monthly household income, and smoking as predictor variables which had significant association with the HU and EQ-VAS scores in the previous Thai population norms study [41]. HU scores for all two instruments were associated with age band (all p < 0.05). However, for sex, only EQ-5D-3L and EQ-5D-5L scores were significant predictors (both p < 0.05). Analysis 2 included all significant demographic factors identified from the univariate analysis. Seven sociodemographic factors, including sex, age band, average monthly household income, occupations, residence of origin, health insurance, and BMI, were employed to investigate the associations with HU scores because their absolute values of Spearman correlation were less than 0.3. Similar to the analysis 1, only age band was associated with HU and EQ-VAS scores. Compared with the samples aged < 30 years, older samples had lower mean HU and EQ-VAS scores where the greatest difference was found among the oldest age group (≥ 70 years). Unlike the analysis 1, female samples reported lower mean HU and EQ-VAS scores than male samples; however, significant difference was only found with EQ-5D-3L scores (β = -0.045, p = 0.010). Regression also showed that unemployed and obese (BMI ≥ 30) participants reported lower HU and EQ-VAS scores than their counterparts (all p < 0.05).

Table 6 Relationship between norm-based scores from EQ-5D-3L, EQ-5D-5L, EQ-VAS and sample characteristics
Full size table

Disutility values associated with self-reported main diseases

Table 7 presents disutility values obtained from two instruments associated with self-reported main diseases. Results showed that the pattern of disutility values obtained from two instruments was similar across diseases where bone/joint, stroke, renal, asthma/COPD, produced higher disutility than other diseases. However, differences in disutility values from the two instruments were observed. The disutility values from the EQ-VAS and those two instruments also shared similar patterns across diseases; however, participants with allergic rhinitis did not report significant differences in disutility value as compared to the healthy state.

Table 7 Self-reported main disease associated with disutility derived from EQ-5D-3L, EQ-5D-5L and EQ-VAS
Full size table

Discussion

The present study generated updated post-COVID Thai norms-based HU scores for the EQ-5D-3L and the EQ-5D-5Lwith a larger general Thai population to improve generalizability of the norm-based scores. We also established the data set of disutility value specific to main disease obtained from those two instruments and EQ-VAS to facilitate HU score calculations for participants with particular diseases in Thailand. However, this study did not report the disutility value for co-morbidities. Therefore, for participants reporting several chronic diseases such as those with hypertension, diabetes, and dyslipidemia, only the main reported disease was analyzed for its effect on the participants’ overall health. Moreover, the disutility value in this study is based on participants residing at their own residence, so generalizability to participants with other conditions is limited such as hospital-based populations.

The percentage of participants reporting “no problem” in each EQ-5D dimension for both EQ-5D-3L and EQ-5D-5L can be ranked as SC (95.5% vs. 95.0%) having the highest percentage, followed by MO (84.1% vs. 83.6%), UA (83.4% vs. 81.0%), AD (76.1% vs. 67.6%), and PD (69.0% vs. 61.0%). This finding implies that the participants were more likely to report problems in mental health for those two instruments (AD for both EQ-5D versions) than physical health (MO and UA for both EQ-5D versions). Similar to earlier work in Mainland China [50] and Japan [7], results of the present study indicated that participants had more impaired mental health than physical health according to both EQ-5D versions. Of note, we found that younger participants (aged < 30 years) reported themselves as having more problems in mental health than physical health from the two instruments as shown in the Supplementary Material. This result is also consistent with those from the previous Thai norms-based study where samples aged < 25 years reported more health problems in mental health (AD) than physical health (MO) of the EQ-5D-5L questionnaire [41]. We reasoned that younger individuals might experience stress due to their work life and social environment [51]. Additionally, if they are experiencing fewer physical health issues due to their younger age, then mental health issues are a more serious consideration in relative terms. Similar to the previous Thai study [3], most of the samples rating themselves at level 1-EQ-5D-3L retained their answers at level 1- EQ-5D-5L for all five dimensions, yielding the presence of high ceiling effects for both EQ-5D versions. These findings indicate that they were relatively healthy which might be due to the exclusion of the samples with acute or life-threatening disease from our study. Nevertheless, some inconsistent responses for both EQ-5D versions were observed with an average proportion at 1.09%, highest and lowest proportions in AD (0.48%) and SC (0.12%), respectively. Therefore, our samples responded to both EQ-5D versions consistently congruent with the reports from previous studies [3, 52].

Regarding the norm-based scores, our results showed that the EQ-5D-5L yielded the mean HU scores of 0.923, while the EQ-5D-3L produced the lower mean HU scores of 0.845 (mean difference = 0.078). Similar to earlier work on head to head comparison for both EQ-5D versions among general Thai population [3, 41] and patients with locally advanced cervical cancer [53], it reported mean difference of 0.08 (0.93 for the EQ-5D-5L and 0.85 for the EQ-5D-3L) and 0.12 (0.755 for the EQ-5D-3L and 0.875 for the EQ-5D-5L) which were close values to our study. We reasoned that the mean differences might be due to different techniques employed to generate the algorithm to calculate the HU scores in that the EQ-5D-3L employed time-trade off [40], while the EQ-5D-5L used both time-trade off and discrete choice experiment techniques (Hybrid method) [39].

As expected, the HU scores from EQ-5D-3L and EQ-5D-5L had a good agreement with the highest ICC of 0.789. Our study showed that the highest ceiling effect was for the EQ-5D-3L scores (57.80%) followed by the EQ-5D-5L (49.05%). These ceiling effects for both EQ-5D versions were similar to those of Asian countries including Japan (68% for the EQ-5D-3L and 55% for the EQ-5D-5L) [7] and Mainland China (urban population 54% for the EQ-5D-5L) [14]; however, those values were higher than those of the western countries such as Switzerland (French speaking population 41.80%) [54] for the EQ-5D-3L and Poland (38.5%) [27], Norway (32.2%) [26], the USA (31.2%) [33] and Germany (30.6%) for the EQ-5D-5L [20]. We reasoned that variations between Asian and Western ethnicities towards EQ-5D items might be accounted for this discrepancy. Nevertheless, this present study and previous studies can confirm that the ceiling effect of the EQ-5D-3L could be reduced when adding two more levels of the response options to the EQ-5D questionnaire. Of note, our ceiling effects for both EQ-5D versions were also close to those of previous Thai study [3] (57.17% for the EQ-5D-3L and 49.08% for the EQ-5D-5L). No floor effects were observed for the EQ-5D-3L and EQ-5D-5L.

Although the HU scores were different across those two instruments, the regression results showed that they consistently declined when increasing age in both analysis 1 and 2. This pattern is similar to the previous norm-based scores from Thailand [41], Iran [22], Vietnam [34] for the EQ-5D-5L and Japan for those two instruments [7] although the age bands were classified differently across studies. Moreover, the present study revealed that unemployed participants and those with BMI ≥ 30 reported lower HU scores from those two instruments than their counterparts. In line with the previous Japanese norms-based study [7] and Hong Kong [15], both analysis 1 and 2 showed that females in the present study reported lower HU scores from those two instruments than male participants although there was no statistical significance for the HU scores from the EQ-5D-5L in analysis 2. Our regression results showed that the HU score difference between both sexes for the EQ-5D-5L (β = 0.009) was also close to the those of previous work for the EQ-5D-5L in Japan (β = 0.015) [7] and Hong Kong (β = 0.005) [15]. Therefore, it could imply that females generally report lower HU scores than males from both EQ-5D versions across populations from different countries.

In relation to the EQ-VAS, our findings indicate that it exhibited a minimal ceiling effect at 3.20%. This figure is notably less than that reported in prior Thai studies, where it was 12.6% [41]. One plausible explanation for this reduced percentage could be the impact of the post-COVID-19 pandemic, which has the potential to deteriorate levels of HRQoL [45]. Similar to other HU scores, the EQ-VAS scores declined with the participants with advancing age. However, some variations for other sociodemographic factors affecting both EQ-VAS and other HU scores were observed. Similar to the previous studies in Thailand [41] and Vietnam [34], the mean EQ-VAS score was 86.06 for the participants achieving the maximum scores of EQ-5D-5L scores (1.000). These findings imply that the participants may consider other health aspects beyond the EQ-5D dimensions when rating their current health with the EQ-VAS, which deserves further investigation in future studies.

Regarding the EQ-VAS scores, we found that the overall difference for mean EQ-VAS scores between this current study and the original valuation EQ-5D-5L valuation study was 3.25 (79.83 vs. 83.08) [39]. Similar to the previous study [55], it reported that the EQ-VAS scores in the post-COVID-19 pandemic were diminished by 8.4% from the pre-pandemic. Furthermore, the response distribution patterns in the EQ-5D dimensions showed changes between the current and original valuation EQ-5D-5L. In the current study, there was a higher percentage of respondents reporting moderate or greater problems than participants in the original valuation EQ-5D-5L study for most of the EQ-5D dimensions. These findings of altered HRQoL levels confirm the rationale of the present study to establish updated population norms for both EQ-5D versions after the COVID-19 pandemic. Nevertheless, the different response patterns to each EQ-5D dimension before and after the COVID pandemic should be further investigated using Rasch analysis based on the item response theory (IRT) in future studies. This could explore the presence of different interpretation of items by different groups and also to what extent items may have shifted in level of difficulty such that they may have become easier or more difficult to endorse.

Similar to previous studies of Japanese population norms for the EQ-5D-3L and EQ-5D-5L [7, 24], bone/joint disorder and stroke yielded the large disutility value (> 0.25) from the two instruments because both diseases could affect the functional capacity and limitation which could contribute to impaired HRQoL [56, 57].

There are some limitations that should be noted. First, the response rate for the interviews was not counted; however, we kept it as maximum as possible due to performing the data collection at the participant’s residence. Second, since this population norm was conducted at the participant’s residence, the results might not be generalizable to other populations such as hospital-based populations. Third, the majority of participants were healthy (63.35%) because we excluded those with acute or life-threatening diseases from this study. Fourth, the disutility value was obtained from those who reported having one main disease, which means that the values of individuals with multiple health conditions were not estimated. Fifth, data validity could be affected by the outline of the survey where the EQ-5D-3L was presented before the EQ-5D-5L. The latter questionnaire inquired about the same health dimensions with different levels of health impairments. Therefore, participants may have been more likely to rate the EQ-5D-5L with the same response of what they had rated themselves on the EQ-5D-3L. Future research needs to explore the possibility of such context effects. Sixth, this study did not report on data for other instruments such as SF-6D because the SF-6D does not have a Thai-specific value set. Future studies could estimate population norms for the SF-6D when the Thai specific-value set is available.

Conclusion

The present study established updated, post-COVID Thai population norm-based scores from EQ-5D-3L, EQ-5D-5L, and EQ-VAS using a population survey that was larger than previous work for these instruments in Thailand and that also used a larger age range. This population norms are presented the HU scores across sexes and age bands. Results showed that the mean EQ-5D-3L, EQ-5D-5L and EQ-VAS scores were 0.845, 0.923 and 79.83, respectively. Although there were differences in HU scores derived from those instruments, all HU scores have similar trend in terms of decreasing with advancing age and being lower among female, unemployed, and obese (BMI ≥ 30) participants. Regression analyses also showed that bone/joint disorder and stroke contributed to the largest disutility value (> 0.25). In addition, these population norms imply that the mental health among younger adults in Thailand (age < 30 years) deteriorated after the COVID-19 pandemic. The reason for this shift deserves further investigation, such as whether it may be due to the disease and disease control measures during the pandemic. Therefore, it is important to consider the long-term health policy implemented to enhance mental health especially for the younger adults in Thailand after the COVID-19 pandemic.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author upon reasonable request.

Abbreviations

AD:

Anxiety or depression

EQ-5D:

EuroQoL-5 dimensions

EQ-VAS:

EuroQol visual analog scale

HRQoL:

Health-related quality of life

HU:

Health utility

IQR:

Interquartile range

MO:

Mobility

PD:

Pain or discomfort

SC:

Self-care

UA:

Usual activities

References

  1. Gusi N, Olivares PR, Rajendram R. The EQ-5D health-related quality of Life Questionnaire. In: Preedy VR, Watson RR, editors. Handbook of Disease burdens and Quality of Life measures. New York, NY: Springer New York; 2010. pp. 87–99.

    Chapter  Google Scholar 

  2. Feng YS, Kohlmann T, Janssen MF, Buchholz I. Psychometric properties of the EQ-5D-5L: a systematic review of the literature. Qual Life Res. 2021;30(3):647–73.

    Article  PubMed  Google Scholar 

  3. Kangwanrattanakul K, Parmontree P. Psychometric properties comparison between EQ-5D-5L and EQ-5D-3L in the general Thai population. Qual Life Res. 2020;29(12):3407–17.

    Article  PubMed  Google Scholar 

  4. Yang F, Devlin N, Luo N. Cost-utility analysis using EQ-5D-5L data: does how the utilities are derived Matter? Value Health. 2019;22(1):45–9.

    Article  PubMed  Google Scholar 

  5. Raftery J, Young A, Stanton L, Milne R, Cook A, Turner D, et al. Clinical trial metadata: defining and extracting metadata on the design, conduct, results and costs of 125 randomised clinical trials funded by the National Institute for Health Research Health Technology Assessment programme. Health Technol Assess. 2015;19(11):1–138.

    Article  PubMed  PubMed Central  Google Scholar 

  6. Devlin NJ, Brooks R. EQ-5D and the EuroQol Group: past, Present and Future. Appl Health Econ Health Policy. 2017;15(2):127–37.

    Article  PubMed  PubMed Central  Google Scholar 

  7. Shiroiwa T, Fukuda T, Ikeda S, Igarashi A, Noto S, Saito S, et al. Japanese population norms for preference-based measures: EQ-5D-3L, EQ-5D-5L, and SF-6D. Qual Life Res. 2016;25(3):707–19.

    Article  PubMed  Google Scholar 

  8. Konig HH, Bernert S, Angermeyer MC, Matschinger H, Martinez M, Vilagut G, et al. Comparison of population health status in six European countries: results of a representative survey using the EQ-5D questionnaire. Med Care. 2009;47(2):255–61.

    Article  PubMed  Google Scholar 

  9. McCaffrey N, Kaambwa B, Currow DC, Ratcliffe J. Health-related quality of life measured using the EQ-5D-5L: South Australian population norms. Health Qual Life Outcomes. 2016;14(1):133.

    Article  PubMed  PubMed Central  Google Scholar 

  10. Bailey H, Janssen MF, La Foucade A, Boodraj G, Wharton M, Castillo P. EQ-5D self-reported health in Barbados and Jamaica with EQ-5D-5L population norms for the English-speaking Caribbean. Health Qual Life Outcomes. 2021;19(1):97.

    Article  PubMed  PubMed Central  Google Scholar 

  11. Van Wilder L, Charafeddine R, Beutels P, Bruyndonckx R, Cleemput I, Demarest S, et al. Belgian population norms for the EQ-5D-5L, 2018. Qual Life Res. 2022;31(2):527–37.

    Article  PubMed  Google Scholar 

  12. Encheva M, Djambazov S, Vekov T, Golicki D. EQ-5D-5L Bulgarian population norms. Eur J Health Econ. 2020;21(8):1169–78.

    Article  PubMed  Google Scholar 

  13. Poder TG, Carrier N, Kouakou CRC. Quebec Health-Related Quality-of-life Population norms using the EQ-5D-5L: decomposition by Sociodemographic Data and health problems. Value Health. 2020;23(2):251–9.

    Article  PubMed  Google Scholar 

  14. 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(1):210.

    Article  PubMed  PubMed Central  Google Scholar 

  15. Wong EL, Cheung AW, Wong AY, Xu RH, Ramos-Goñi JM, 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(8):916–24.

    Article  PubMed  Google Scholar 

  16. Bailey HH, Janssen MF, Varela RO, Moreno JA. EQ-5D-5L Population norms and Health Inequality in Colombia. Value Health Reg Issues. 2021;26:24–32.

    Article  PubMed  Google Scholar 

  17. Jensen MB, Jensen CE, Gudex C, Pedersen KM, Sørensen SS, Ehlers LH. Danish population health measured by the EQ-5D-5L. Scand J Public Health. 2021:14034948211058060.

  18. McNamara S, Schneider PP, Love-Koh J, Doran T, Gutacker N. Quality-adjusted life expectancy norms for the English Population. Value Health. 2023;26(2):163–9.

    Article  PubMed  Google Scholar 

  19. Gautier L, Azzi J, Saba G, Bonnelye G, de Pouvourville G. Population norms in France with EQ-5D-5L: health states, value indexes, and VAS. Eur J Health Econ. 2023;24(9):1517–30.

    Article  PubMed  Google Scholar 

  20. Grochtdreis T, Dams J, König HH, 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(6):933–44.

    Article  PubMed  Google Scholar 

  21. Purba FD, Hunfeld JAM, Iskandarsyah A, Fitriana TS, Sadarjoen SS, Passchier J, et al. Quality of life of the Indonesian general population: test-retest reliability and population norms of the EQ-5D-5L and WHOQOL-BREF. PLoS ONE. 2018;13(5):e0197098.

    Article  PubMed  PubMed Central  Google Scholar 

  22. Emrani Z, Akbari Sari A, Zeraati H, Olyaeemanesh A, Daroudi R. Health-related quality of life measured using the EQ-5D-5 L: population norms for the capital of Iran. Health Qual Life Outcomes. 2020;18(1):108.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Hobbins A, Barry L, Kelleher D, O’Neill C. The health of the residents of Ireland: Population norms for Ireland based on the EQ-5D-5L descriptive system - a cross sectional study. HRB Open Res. 2018;1:22.

    Article  PubMed  PubMed Central  Google Scholar 

  24. Shiroiwa T, Noto S, Fukuda T. Japanese Population norms of EQ-5D-5L and Health Utilities Index Mark 3: disutility catalog by Disease and Symptom in Community settings. Value Health. 2021;24(8):1193–202.

    Article  PubMed  Google Scholar 

  25. Sullivan T, Turner RM, Derrett S, Hansen P. New Zealand Population norms for the EQ-5D-5L constructed from the personal value sets of participants in a National Survey. Value Health. 2021;24(9):1308–18.

    Article  PubMed  Google Scholar 

  26. Garratt AM, Hansen TM, Augestad LA, Rand K, Stavem K. Norwegian population norms for the EQ-5D-5L: results from a general population survey. Qual Life Res. 2022;31(2):517–26.

    Article  PubMed  Google Scholar 

  27. Golicki D, Niewada M. EQ-5D-5L Polish population norms. Arch Med Sci. 2017;13(1):191–200.

    Article  PubMed  Google Scholar 

  28. Hołownia-Voloskova M, Tarbastaev A, Golicki D. Population norms of health-related quality of life in Moscow, Russia: the EQ-5D-5L-based survey. Qual Life Res. 2021;30(3):831–40.

    Article  PubMed  Google Scholar 

  29. Prevolnik Rupel V, Ogorevc M. EQ-5D-5L Slovenian population norms. Health Qual Life Outcomes. 2020;18(1):333.

    Article  PubMed  PubMed Central  Google Scholar 

  30. Hernandez G, Garin O, Pardo Y, Vilagut G, Pont À, Suárez M, et al. Validity of the EQ-5D-5L and reference norms for the Spanish population. Qual Life Res. 2018;27(9):2337–48.

    Article  PubMed  Google Scholar 

  31. Teni FS, Gerdtham UG, Leidl R, Henriksson M, Åström M, Sun S, et al. Inequality and heterogeneity in health-related quality of life: findings based on a large sample of cross-sectional EQ-5D-5L data from the Swedish general population. Qual Life Res. 2022;31(3):697–712.

    Article  PubMed  Google Scholar 

  32. Bailey H, Janssen MF, La Foucade A, Kind P. EQ-5D-5L population norms and health inequalities for Trinidad and Tobago. PLoS ONE. 2019;14(4):e0214283.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Jiang R, Janssen MFB, Pickard AS. US population norms for the EQ-5D-5L and comparison of norms from face-to-face and online samples. Qual Life Res. 2021;30(3):803–16.

    Article  PubMed  Google Scholar 

  34. Nguyen LH, Tran BX, Hoang Le QN, Tran TT, Latkin CA. Quality of life profile of general Vietnamese population using EQ-5D-5L. Health Qual Life Outcomes. 2017;15(1):199.

    Article  PubMed  PubMed Central  Google Scholar 

  35. Thavorncharoensap M, Sakthong P. Health utility. In: Pannarunothai S, Pilasant S, Saengsri W, Kingkaew P, editors. The Guideline of Health Technology Assessment in Thailand. Nonthaburi: Health Systems Research Institute; 2019. pp. 77–94.

    Google Scholar 

  36. Pattanaphesaj J, Thavorncharoensap M. Measurement properties of the EQ-5D-5L compared to EQ-5D-3L in the Thai diabetes patients. Health Qual Life Outcomes. 2015;13:14.

    Article  PubMed  PubMed Central  Google Scholar 

  37. Sakthong P, Sonsa-Ardjit N, Sukarnjanaset P, Munpan W. Psychometric properties of the EQ-5D-5L in Thai patients with chronic diseases. Qual Life Res. 2015;24(12):3015–22.

    Article  PubMed  Google Scholar 

  38. Kangwanrattanakul K. A systematic review of health state utility values in Thai cancer patients. Expert Rev Pharmacoecon Outcomes Res. 2022;22(8):1171–86.

    Article  PubMed  Google Scholar 

  39. Pattanaphesaj J, Thavorncharoensap M, Ramos-Goni JM, Tongsiri S, Ingsrisawang L, Teerawattananon Y. The EQ-5D-5L valuation study in Thailand. Expert Rev Pharmacoecon Outcomes Res. 2018;18(5):551–8.

    Article  PubMed  Google Scholar 

  40. Tongsiri S, Cairns J. Estimating population-based values for EQ-5D health states in Thailand. Value Health. 2011;14(8):1142–5.

    Article  PubMed  Google Scholar 

  41. Kangwanrattanakul K. Normative profile of the EQ-5D-5L dimensions, EQ-5D-5L index and EQ-VAS scores for the general Thai population. Qual Life Res. 2023;32(9):2489–502.

    Article  PubMed  Google Scholar 

  42. Gheasi M, Ishikawa N, Kourtit K, Nijkamp P. A meta-analysis of human health differences in urban and rural environments. Lett Spat Resource Sci. 2019;12(3):167–86.

    Article  Google Scholar 

  43. Ishikawa H, Kato M, Kiuchi T. Declines in health literacy and health-related quality of life during the COVID-19 pandemic: a longitudinal study of the Japanese general population. BMC Public Health. 2021;21(1):2180.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  44. Pieh C, Budimir S, Probst T. The effect of age, gender, income, work, and physical activity on mental health during coronavirus disease (COVID-19) lockdown in Austria. J Psychosom Res. 2020;136:110186.

    Article  PubMed  PubMed Central  Google Scholar 

  45. Malik P, Patel K, Pinto C, Jaiswal R, Tirupathi R, Pillai S, et al. Post-acute COVID-19 syndrome (PCS) and health-related quality of life (HRQoL)-A systematic review and meta-analysis. J Med Virol. 2022;94(1):253–62.

    Article  CAS  PubMed  Google Scholar 

  46. Smith P, De Pauw R, Van Cauteren D, Demarest S, Drieskens S, Cornelissen L, et al. Post COVID-19 condition and health-related quality of life: a longitudinal cohort study in the Belgian adult population. BMC Public Health. 2023;23(1):1433.

    Article  PubMed  PubMed Central  Google Scholar 

  47. Janssen MF, Birnie E, Haagsma JA, Bonsel GJ. Comparing the standard EQ-5D three-level system with a five-level version. Value Health. 2008;11(2):275–84.

    Article  PubMed  Google Scholar 

  48. Koo TK, Li MY. A Guideline of selecting and reporting Intraclass correlation coefficients for Reliability Research. J Chiropr Med. 2016;15(2):155–63.

    Article  PubMed  PubMed Central  Google Scholar 

  49. Austin PC, Escobar M, Kopec JA. The use of the Tobit model for analyzing measures of health status. Qual Life Res. 2000;9(8):901–10.

    Article  CAS  PubMed  Google Scholar 

  50. Xie S, Wu J, Xie F. Population norms for SF-6Dv2 and EQ-5D-5L in China. Appl Health Econ Health Policy. 2022;20(4):573–85.

    Article  PubMed  Google Scholar 

  51. Law PCF, Too LS, Butterworth P, Witt K, Reavley N, Milner AJ. A systematic review on the effect of work-related stressors on mental health of young workers. Int Arch Occup Environ Health. 2020;93(5):611–22.

    Article  CAS  PubMed  Google Scholar 

  52. Buchholz I, Janssen MF, Kohlmann T, Feng YS. A systematic review of studies comparing the Measurement Properties of the three-level and five-level versions of the EQ-5D. PharmacoEconomics. 2018;36(6):645–61.

    Article  PubMed  PubMed Central  Google Scholar 

  53. Katanyoo K, Thavorncharoensap M, Chaikledkaew U, Riewpaiboon A. A comparison of six approaches for measuring utility values among patients with locally advanced cervical cancer. Expert Rev Pharmacoecon Outcomes Res. 2022;22(1):107–17.

    Article  PubMed  Google Scholar 

  54. Perneger TV, Combescure C, Courvoisier DS. General population reference values for the French version of the EuroQol EQ-5D health utility instrument. Value Health. 2010;13(5):631–5.

    Article  PubMed  Google Scholar 

  55. Persson U, Olofsson S, Gu NY, Gong CL, Jiao X, Hay JW. Quality of Life in the Swedish General Population During COVID-19-Based on pre-and post-pandemic outbreak measurement. Nord J Health Econ. 2021 Sep 21;9(1):56–73.

  56. Ramos-Lima MJM, Brasileiro IC, Lima TL, Braga-Neto P. Quality of life after stroke: impact of clinical and sociodemographic factors. Clin (Sao Paulo). 2018;73:e418.

    Article  Google Scholar 

  57. Kawano MM, Araújo IL, Castro MC, Matos MA. Assessment of quality of life in patients with knee osteoarthritis. Acta Ortop Bras. 2015;23(6):307–10.

    Article  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We would like to express sincere gratitude to the EuroQoL group for their permission to use EQ-5D-3L and EQ-5D-5L questionnaires for data collection in this study. We thank our well-trained interviewers for their help in data collection. We also thank to all participants from Bangkok, Samut Prakan, Nonthaburi, Chonburi, Nakhon Pathom, Chaing Mai, Nakhon Sawan, Nakhon Ratchasima, Khon Kaen, Buriram, Nakhon Si Thammarat, and Phatthalung for their valuable time to participate in this project.

Funding

This work was financially supported by the Research Grant of Faculty of Pharmaceutical Sciences, Burapha University (Grant no. Rx6/2566).

Author information

Authors and Affiliations

  1. Division of Social and Administrative Pharmacy, Faculty of Pharmaceutical Sciences, Burapha University, 169 Long-Hard Bangsaen Rd.,, Chonburi, Mueang, 20131, Thailand

    Krittaphas Kangwanrattanakul

  2. Department of Psychology and Neuroscience, Auckland University of Technology, Auckland, New Zealand

    Christian U. Krägeloh

Authors
  1. Krittaphas Kangwanrattanakul
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Christian U. Krägeloh
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Krittaphas Kangwanrattanakul was involved in conception, study design, data collection, data analyses and interpretations and wrote the first draft of the main manuscript text. Christian U. Krägeloh rechecked the data intepretations and supervised writing the manuscript. All authors edited, read, and approved the final manuscript and are all in agreement with the manuscript. The content has not been published elsewhere.

Corresponding author

Correspondence to Krittaphas Kangwanrattanakul.

Ethics declarations

Ethics approval and consent to participate

The protocol for the research was approved by Burapha University Institutional Review Board (IRB1-031/2566) and adhered to the principles outlined in the Declaration of Helsinki. Before signing this consent form, the participant is explained the aim, methodology, and details that are provided in Participant Information Sheet which is given to each of participant. The participant understands that their participation is voluntary and that they are free to withdraw at any time, without giving a reason, without cost, and without any consequences. The participants are ensured that their participation is anonymous, and the collected data is confidential. Informed consent was obtained from all subjects or their legal guardian(s).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Kangwanrattanakul, K., Krägeloh, C.U. EQ-5D-3L and EQ-5D-5L population norms for Thailand. BMC Public Health 24, 1108 (2024). https://doi.org/10.1186/s12889-024-18391-3

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12889-024-18391-3

Keywords

BMC Public Health

ISSN: 1471-2458

Contact us