
==== Front
Sci Rep
Sci Rep
Scientific Reports
2045-2322
Nature Publishing Group UK London

39232021
71048
10.1038/s41598-024-71048-4
Article
Modern psychometric evaluation of Thai WHOQOL-BREF and its shorter versions in patients undergoing warfarin in Thailand: Rasch analysis
http://orcid.org/0000-0002-4921-8508
Kangwanrattanakul Krittaphas krittaphas@buu.ac.th

http://orcid.org/0000-0002-1590-5182
Kulthanachairojana Nattanichcha
https://ror.org/01ff74m36 grid.411825.b 0000 0000 9482 780X Division of Social and Administrative Pharmacy, Faculty of Pharmaceutical Sciences, Burapha University, 169 Long-Hard Bangsaen Rd., Mueang, Chonburi, 20131 Thailand
4 9 2024
4 9 2024
2024
14 2063917 4 2024
23 8 2024
© The Author(s) 2024
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, 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 you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. 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-nc-nd/4.0/.
Rasch analysis was employed to investigate the psychometric properties of the World Health Organization Quality of Life-BREF (WHOQOL-BREF) and its shorter versions (EUROHIS-QOL-8 and WHOQOL-5) within the context of patients undergoing warfarin in Thailand. A group of 260 patients were recruited from three public hospitals and tasked with completing the WHOQOL-BREF questionnaire. Rasch analysis showed that the WHOQOL-BREF, structured into four-domain subtests, achieved a commendable fit to the Rasch model (χ2[16] = 12.26, p = 0.73), met the criterion of unidimensionality (7.31% significant t-tests; lower bound confidence interval, 4.66), and demonstrated satisfactory reliability (PSI = 0.87). The adoption of a subtest approach facilitated an acceptable fit to the Rasch model for each domain of the WHOQOL-BREF, except for the social domain. However, the presence of local dependency of the three-item social domain was detected, so the reliability was not reported. The WHOQOL-5 proved to be unidimensional, fitting the Rasch model acceptably, and had satisfactory reliability. Conversely, the EUROHIS-QOL-8 presented local dependency; thus, reliability was not reported. Consequently, the WHOQOL-BREF in its four-domain subtests is recommended for pre- and post-HRQoL measurements, whereas the WHOQOL-5 can effectively measure HRQoL levels in between-group analyses.

Keywords

Health-related quality of life
WHOQOL-BREF
Thailand
Rasch model
Patients undergoing warfarin
Subject terms

Psychology
Health care
issue-copyright-statement© Springer Nature Limited 2024
==== Body
pmcIntroduction

In the past three decades, there has been a rapid change in health concerns from communicable diseases (HIV/AIDs, malaria, tuberculosis, influenza, etc.) to non-communicable diseases (NCDs), which are considered to be non-contagious ailments associated with multiple risk factors, extended latency period, and chronicity1,2. NCDs have become a significant global health concern, responsible for 74% of all yearly global deaths1,2.

Among NCDs, cardiovascular disease (CVD) is a critical health problem globally, being the foremost cause of death among them, resulting in 17.9 million deaths annually1,3. This trend is similar in Thailand, where mortality rates from coronary artery disease and ischemic heart disease, two prevalent CVDs, have risen from 31.36 to 33.54 between 2007 and 20214. Additionally, numerous studies have highlighted the adverse impact of CVDs on health-related quality of life (HRQoL) due to their tendency to impair physical health and necessitate lifelong treatment with prolonged anticoagulant therapy5–8.

Warfarin is one of the major anticoagulant medications used in the management of various CVDs such as atrial fibrillation, mechanical valve replacement, deep vein thrombosis, and pulmonary embolism9. However, its administration entails a complex dosage regimen, necessitates monthly blood monitoring for the international normalized ratio, and carries a notable risk factor for treatment response variability as well as internal and external bleeding events10. In addition, warfarin can interact with a wide range of foods, herbs, and supplements10, prompting alternations in eating habits and lifestyles among patients undergoing warfarin therapy. These factors are expected to have a negative impact on the HRQoL of the affected individuals. Consistent with previous research conducted in Thailand, it has shown that patients undergoing warfarin treatment tend to experience lower HRQoL levels than the general population11.

Generic HRQoL instruments are designed to measure HRQoL levels in both healthy individuals and patients with various health conditions, including those undergoing warfarin therapy. Examples of such generic HRQoL instruments are the EuroQoL-5 dimension, the World Health Organization Quality of Life Brief (WHOQOL-BREF), the 36-item Short-Form (SF-36) Health Survey, etc. However, these generic instruments may not always be sensitive enough to detect clinical changes specific to certain disease conditions. In such instances, condition-specific instruments for patients with cardiovascular disease may be more suitable for capturing clinical changes in patients undergoing warfarin treatment, such as the Duke Anticoagulant Satisfaction Scale (DASS)12,13. Currently, there is no clear consensus to recommend a specific HRQoL instrument for measuring HRQoL levels among patients undergoing warfarin treatment. This lack of consensus is partly due to mixed evidence from recent studies, which have employed different instruments to measure HRQol levels in this patient population, including DASS14,15, SF-3616, and WHOQQOL-BREF17,18.

In this study conducted in Thailand, the measurement of HRQoL among patients undergoing warfarin treatment was conducted using the WHOQOL-BREF11, an abbreviated version of the WHOQOL-100. This instrument was derived from data collected through the WHOQOL-100, a comprehensive tool collected from 23 countries globally, including Thailand. The WHOQOL-BREF is  therefore considered a cross-cultural instrument suitable for diverse sociodemographic participants19. Comprising 26 items categorized into four dimensions, including physical, psychological, social, and environmental health domains, the WHOQOL-BREF is extensively used to measure HRQoL levels across various populations, ranging from healthy individuals to those with health conditions. Additionally, to facilitate HRQoL measurement and reduce respondent burden, shorter versions of the WHOQOL-BREF have been developed, namely the EUROHIS-QOL-820 and World Health Organization Quality of Life-5 (WHOQOL-5)21. These shorter versions have been successfully employed to measure HRQoL levels in healthy individuals20,22 and therapeutic contexts21,23.

Previous psychometric evaluations supported the use of WHOQOL-BREF as a valid, reliable, and applicable tool to measure HRQoL measurements in various populations, including human immunodeficiency virus/acquired immunodeficiency syndrome (HIV/AIDs) in Thailand24, students with disabilities in Malaysia25, Japanese couples26, next of kin to elderly residents in Sweden nursing homes27, individuals with brain injuries and orthopedic conditions in New Zealand23, and general population samples22,28. In addition, the Thai version of the WHOQOL-BREF has demonstrated its validity, reliability, and practicality in measuring HRQoL in various Thai populations, including general populations22,29, elderly people30, cancer31 and HIV/AIDs patients24, and Thai college students32. However, no studies have investigated the psychometric characteristics of the WHOQOL-BREF and its shorter versions in Thai patients undergoing warfarin treatment.

Previous Thai studies have investigated the psychometric properties of the WHOQOL-BREF using classical test theory (CTT)24,29–32. However, CTT’s treatment of raw scores from the five-point Likert scale as interval data is problematic, given that the Likert scale is ordinal data33. This assumption is invalid, prompting a shift toward Rasch analysis, a contemporary statistical approach based on item-response theory. Rasch analysis holds significance because it can transform the ordinal scale, whether dichotomous or polytomous, into an interval scale without violating the data’s assumptions34. Nevertheless, it requires certain essential assumptions, namely unidimensionality and local independence35,36. Furthermore, this method provides valuable statistics such as the person separation index (PSI) to determine reliability, differential item functioning (DIF) to measure consistency across sociodemographic subgroups, item difficulty to evaluate individual item endorsements, etc37. A crucial outcome of Rasch analysis involves generating an ordinal-interval conversion table, facilitating the transformation of the ordinal scale into an interval scale for calculating the HRQoL score for each individual while upholding parametric statistical assumptions38. Therefore, Rasch analysis has gained prominence as a preferred approach for measuring the psychometric properties of HRQoL instruments, especially the WHOQOL-BREF, in various populations23,28,33,36.

To date, there has been only one previous study on the psychometric evaluation of WHOQOL-BREF, EUROHIS-QOL-8, and WHOQOL-5 within the general Thai population22. The results showed that the WHOQOL-BREF and EUROHIS-QOL-8 are valid instruments for measuring HRQoL levels, but the WHOQOL-5 lacks significant reliability22. Nevertheless, no study has investigated the properties of WHOQOL-BREF and its shorter versions in patients undergoing warfarin in Thailand. Consequently, this study aimed to investigate the psychometric properties of WHOQOL-BREF, EUROHIS-QOL-8, and WHOQOL-5 using Rasch analysis among patients undergoing warfarin and to generate an ordinal-interval conversion table for interval score calculation in HRQoL measurement.

Methods

Study design

This study used a dataset derived from a project titled “Health-related quality of life and willingness to pay measurement among patients on warfarin in Thailand11” A cross-sectional design was employed, involving 260 patients receiving warfarin treatment in outpatient departments at three public hospitals. Purposive sampling was used to recruit eligible participants who met the following criteria: being at least 18 years old, having been prescribed warfarin for clinical indications for a minimum of 2 months, being proficient in Thai, and providing informed consent for study participation. Patients with life-threatening diseases, cognitive impairment, or disabilities were excluded from the study. This Rasch analysis was conducted with 260 samples calculated from the original study11.

Data collection

Face-to-face interviews were conducted to complete the questionnaire. Interviewers were instructed not to explain the meanings of questions or response options to eligible patients, allowing them to complete the questionnaire based on their understanding. Data collection was performed among patients awaiting appointments at outpatient anticoagulant clinics in three public hospitals: Chonburi Hospital (n = 159), Burapha University Hospital (n = 35), and Bang Lamung Hospital (n = 66) between June 2022 and June 2023. Before the study commenced, patients were provided with a participant information sheet explaining the objectives and overall data collection process in plain language. Written informed consent was obtained from each patient to confirm their voluntary participation in this study. This research received approval from the Institutional Review Board of Burapha University (IRB1- 021/2566 Amendment 1) and Chonburi Hospital’s Institutional Review Board (37/65/O/q).

Instruments

WHOQOL-BREF

The Thai version of the WHOQOL-BREF was used in this study, with permission granted by the director of Suan Prung Psychiatric Hospital in Thailand29. This instrument comprises 26 items categorized into four health domains: physical health (7 items), psychological health (6 items), social relationships (3 items), and environment (8 items). Additionally, there were two items inquiring about general health and overall quality of life (QOL), bringing the total to 26 items. Patients were asked to rate their health status over the past two weeks using a five-point Likert scale, with response options ranging from 1 (not at all), 2 (not much), 3 (moderately), 4 (a great deal), and 5 (completely). Notably, three items (2, 9, and 11) were negatively worded, requiring reverse scoring before computation. A higher score on the WHOQOL-BREF indicates a better health status.

EUROHIS-QOL-8

EUROHIS-QOL-8 is a shorter version of WHOQOL-BREF20, comprising eight items extracted from the Thai WHOQOL-BREF instrument. These items include satisfaction with health (item 1), energy (item 3), self-esteem (item 7), daily life activities (item 10), personal relationships (item 13), home environment (item 16), financial support (item 17), and overall QOL (item 26). Response options and recall period align with those of the WHOQOL-BREF. Each item score contributes to the computation of an overall score, where higher scores indicate a better health status.

WHOQOL-5

The WHOQOL-5 is another short version of the WHOQOL-BREF, featuring five items sourced from the Thai overall WHOQOL-BREF instrument21. These items cover satisfaction with health (item 1), daily life activities (item 10), personal relationships (item 13), home environment (item 16), and overall QOL (item 26). Similar to EUROHIS-QOL-8 and WHOQOL-BREF, the response options and recall period remain consistent. The overall score is derived from the individual item scores, with higher scores indicating a better health status.

Data analyses

Kurtosis and skewness coefficients were used to assess the normal distribution of data for each item in the WHOQOL-BREF. Generally, acceptable values fall within the range of − 1.00 to 1.0039. Internal consistency was evaluated for the entire WHOQOL-BREF instrument, EUROHIS-QOL-8, WHOQOL-5, and individual domains of the WHOQOL-BREF using Cronbach’s alpha. An acceptable Cronbach’s alpha value is generally considered to be greater than 0.7, indicating good internal consistency40. Descriptive statistics, including frequencies and percentages, were used to report patient characteristics such as sex (male/female), age group (18–53/54–65/66 and higher), marital status (single/living as married), education level (no formal education, primary and secondary school/college), and comorbidities other than CVD (Yes/No).

Rasch analysis

The Rasch analysis was performed on the 24-item WHOQOL-BREF, excluding items related to satisfaction with health (item 1) and overall QOL (item 26) as they do not belong to the four domains of the WHOQOL-BREF. However, these two items were included in the analysis for the shorter versions of the WHOQOL-BREF (EUROHIS-QOL-8 and WHOQOL-5).

Item-trait interaction was evaluated using non-significant Chi-square statistics to determine the overall Rasch model fit37. A non-significant Chi-square suggests homogeneity in the hierarchical ordering of items (item difficulty) across the underlying trait or construct, indicating a good fit condition.

Furthermore, item-person interaction statistics were examined using standardized individual item and person fit residuals, presented as z-score values ranging between − 2.5 and 2.5, indicating an acceptable fit to the Rasch model37.

Unidimensionality refers to the measurement score assessing only one attribute at a time37. In this study, the approach proposed by Smith was used to investigate unidimensionality41. The scale was deemed unidimensional if the percentage of significant t-tests was less than 5% and/or the lower bound of the confidence interval (CI) for significant t-tests overlapped with the 5% cut-off point23. If the model did not fit the Rasch model, we explored possible sources of misfit, and unidimensionality was not a major point of concern in this case. Nonetheless, the unidimensionality test results could inform us whether the misfit to the Rasch model was possibly due to multidimensionality.

Local independence is a requirement of the model, and item dependence was examined using the Yen’s Q3 statistic in the form of the residual correlation matrix. Although no critical value for Yen’s Q3 can definitively indicate item dependency, a recent study by Christensen et al. suggested that a critical value of 0.2 above average correlation appears to be a reasonable cut-off point42. Therefore, item dependency was detected if the residual correlation exceeded 0.2 above the average correlation42. A subtest approach was adopted by merging related items to create super items, to address local dependency41. Super items were formed by merging items based on their respective WHOQOL-BREF dimensions (physical, psychological, social, and environmental domains)22. The presence of local dependency was also indicated if the domain subtest was not unidimensional, as observed in previous studies exploring the higher-order structure of the WHOQOL instrument22,43. Notably, if the models presented with local dependency, Cronbach’s alpha and PSI were not reported to show their degree of reliability44.

DIF, also known as item bias, was examined to determine whether the scale remained invariant for measuring the construct of interest (HRQoL) across sociodemographic subgroups23,41. DIF was investigated concerning item performance across sex (male/female), age group (18–53/54–65/66 and higher), marital status (single/living as married), education level (no formal education, primary and secondary school/college), and comorbidities other than CVD (Yes/No). The DIF analysis was conducted using analysis of variance followed by graphical inspection45. A significant p-value (< 0.05) for differences between sociodemographic subgroups was considered an indication of DIF detection. If DIF was detected, the DIF items were further split by personal factors anchored by non-DIF items using paired t-tests to investigate mean differences41. The magnitude of the difference was estimated as an effect size, with a value less than 0.1 considered indicative of a small DIF, requiring no further action41,46.

The PSI is a valuable statistic obtained from Rasch analysis, measuring the measure’s ability to differentiate between different levels of the construct of interest37. It can be interpreted similarly to Cronbach’s alpha, with a value exceeding 0.7 indicating acceptable reliability for the whole measure47. In the context of HRQoL measurement using WHOQOL-BREF, EUROHIS-QOL-8, and WHOQOL-5 questionnaires, a PSI of 0.7 suggests the measure’s reliability for between-group analyses, whereas a PSI of 0.85 or higher recommends the measure for pre-post HRQoL measurement23. The Rasch analysis was conducted using RUMM 2030 software, while other statistical analyses were performed using STATA 17 (StataCorp LLC, College Station, TX, USA).

Results

Sample characteristics

Table 1 presents the demographic characteristics of patients recruited for the warfarin study. The samples included an equal number of male and female participants. The average age of the samples was 58.3 years (SD = 12.6). Most participants had primary school education or no formal education (60.8%) and were living as married couples (61.9%). A significant proportion reported comorbidities other than CVD (68.5%). Notably, there were no missing data for the 26-item WHOQOL-BREF questionnaire. Table 1 Sample characteristics.

Characteristics	Frequencies (Percentage)	
Sex, n (%)	
 Male	130 (50.0)	
 Female	130 (50.0)	
Age group, n(%)	
 18–53	88 (33.8)	
 54–65	91 (35.0)	
 66 and upper	81 (31.2)	
Marital status, n(%)	
 Single	99 (38.1)	
 Living as married	161 (61.9)	
Education level, n(%)	
 No/Primary/Secondary school	229 (88.1)	
 Bachelor’s degree or higher	31 (11.9)	
Comorbidities other than CVD, n(%)	
 No	82 (31.5)	
 Yes	178 (68.5)	
CVD: Cardiovascular diseases.

The kurtosis and skewness coefficients for all 26 items of the WHOQOL-BREF were within an acceptable range of − 1 to 1. The range for skewness and kurtosis coefficients was < 0.01 for item 12 (Work capacity) to 0.88 for item 10 (Activities of daily living), and < 0.01 for item 21 (Physical environment) to 0.88 for item 24 (Mobility), respectively. Regarding internal consistency, Cronbach’s alpha values were calculated for the four-domain subtests and WHOQOL-5 were 0.84 and 0.72, respectively because the EUROHIS-QOL-8 showed local dependency. Additionally, Cronbach’s alpha values for each WHOQOL-BREF domain were 0.65, 0.77, and 0.77 for physical, psychological, and environmental domains, respectively where social domain was not reported for Cronbach’s alpha due to its presence of local dependency.

Table 2 presents Rasch model fit statistics for the 24-item WHOQOL-BREF and its four-domain subtests among patients undergoing warfarin therapy. Among the 24-item WHOQOL-BREF, items 25 (sexual activity) and 11 (dependence on medical aids) were identified as the most difficult items to endorse, whereas items 8 (bodily image) and 9 (negative feelings) were the easiest. Five items (2, 4, 11, 25, and 20) showed misfits to the Rasch model due to their item-fit residual exceeded the acceptable threshold of ± 2.5. Table 2 Rasch model fit statistics item locations, fit residuals, and Chi-square for the Thai WHOQOL-BREF and four-domain subtests for the patients with warfarin therapy.

Measures	Fit statistics	
Location	S.E.	Individual item fit	χ2	Prob	
1. Thai WHOQOL-BREF items	
1.1 Physical health	
Pain (Q2)	˗0.226	0.070	4.121	26.033	 < 0.001	
Energy (Q3)	0.528	0.083	0.302	4.032	0.672	
Sleep (Q4)	0.531	0.066	2.637	12.095	0.060	
Activities of daily living (Q10)	˗0.390	0.094	˗1.721	19.948	0.003	
Dependence on medical aids (Q11)	0.859	0.065	10.640	243.576	 < 0.001	
Work capacity (Q12)	0.508	0.081	− 0.689	13.462	0.036	
Mobility (Q24)	0.122	0.070	− 0.203	3.371	0.761	
1.2 Psychological health	
Positive feeling (Q5)	0.128	0.092	− 1.253	15.546	0.016	
Concentration (Q6)	− 0.103	0.086	− 1.795	15.685	0.016	
Self-esteem (Q7)	− 0.244	0.095	− 2.252	32.077	 < 0.001	
Bodily-image (Q8)	− 1.747	0.093	− 1.118	9.177	0.164	
Negative feeling (Q9)	− 0.918	0.071	1.242	13.020	0.043	
Personal belief (Q23)	− 0.698	0.088	− 1.729	24.027	0.001	
1.3 Social health	
Personal relationship (Q13)	− 0.003	0.086	− 0.575	13.102	0.041	
Social support (Q14)	0.294	0.078	0.053	5.246	0.513	
Sexual activity (Q25)	0.901	0.065	3.243	10.536	0.104	
1.4 Environment	
Security (Q15)	− 0.008	0.086	− 1.453	22.169	0.001	
Home environment (Q16)	− 0.467	0.096	− 0.602	2.576	0.860	
Financial support (Q17)	0.477	0.082	0.551	2.539	0.864	
Health care (Q18)	0.092	0.084	0.281	14.383	0.026	
Accessibility of needed information (Q19)	− 0.328	0.089	0.743	1.241	0.975	
Leisure activity (Q20)	0.522	0.067	6.822	41.219	 < 0.001	
Physical environment (Q21)	− 0.144	0.090	− 1.309	13.937	0.030	
Transport (Q22)	0.315	0.080	− 1.382	13.927	0.030	
2. Four domain subtests	
Physical	0.178	0.026	1.452	2.599	0.857	
Psychological	˗0.307	0.027	˗2.025	6.144	0.407	
Social	0.201	0.036	1.630	1.984	0.921	
Environmental	˗0.071	0.024	0.505	3.106	0.795	
Bolded values were fit residual ≥|2.5|.

Table 3 summarizes the overall Rasch model fit statistics. The initial model of the 24-item WHOQOL-BREF (excluding anchor items 1 and 26) showed a significant misfit to the Rasch model (χ2[96] = 540.08, p < 0.01), and unidimensionality was not achieved (13.46% significant t-tests; lower bound CI, 10.81) (Analysis 1). No DIF was detected based on sociodemographic factors. To address multidimensionality and significant misfits, four-domain subtests were created. These subtests achieved a perfect fit to the Rasch model (χ2 [16] = 12.26, p = 0.73) with high reliability (PSI = 0.87) and acceptable unidimensionality (7.31% significant t-tests; lower bound CI, 4.66) (Analysis 2). However, DIF by age group was observed for physical (F = 6.92, p < 0.01) and environmental (F = 10.23, p < 0.01) domain super items, with graphical inspection showing deviations for age group (≥ 66 years). Subsequently, the physical and environmental domain subtests were split by age group (18–65 vs 66 and upper), with negligible effect size for mean difference (< 0.01 and 0.02 for physical and environmental domain super items, respectively). Therefore, no action was deemed necessary for DIF, and the original four-domain subtests were used for Rasch analysis. As shown in Table 2, the psychological (− 0.307) and social (0.201) domains were identified as the easiest and most difficult domain super items to endorse, with no misfit observed in any of the super items to the Rasch model. Table 3 Summary of overall Rasch model fit statistics for the patients with warfarin therapy.

Analysis	Measures	Item fit residual (mean ± SD)	Person fit residual (mean ± SD)	Goodness of fit	PSI	Significant t-test (Unidimensionality)	
χ2	p-value	%	Upper bound CI	Lower bound CI	
	WHOQOL-BREF	
1	24-item WHOQOL-BREF	0.61 ± 3.04	- 0.33 ± 1.91	540.08 (96)	 < 0.01	N/A	13.46	16.11	10.81	
2	4 domains subtests	0.39 ± 1.68	- 0.50 ± 1.19	12.26 (16)	0.73	0.87	7.31	9.96	4.66	
	WHOQOL-BREF domains	
3	Physical – 7 items	0.23 ± 2.36	- 0.34 ± 1.20	150.47 (28)	 < 0.01	N/A	5.77	8.42	3.12	
4	Physical-5 super items	0.13 ± 1.92	- 0.47 ± 1.26	29.63 (20)	0.08	0.70	2.31	4.96	- 0.34	
5	Psychological – 6 items	- 0.10 ± 1.25	- 0.56 ± 1.44	52.70 (24)	 < 0.01	0.78	2.69	5.61	0.04	
6	Psychological – 5 super items	0.13 ± 0.81	- 0.55 ± 1.37	20.52 (20)	0.43	0.79	1.54	4.19	- 1.11	
7	Social – 3 items	0.35 ± 1.04	- 0.57 ± 1.12	12.16 (12)	0.43	N/A	1.92	4.57	- 0.73	
8	Environment – 8 items	0.26 ± 2.14	- 0.66 ± 1.75	70.70 (32)	 < 0.01	N/A	4.62	7.27	1.97	
9	Environment – 7 items	0.02 ± 1.00	- 0.71 ± 1.71	32.64 (28)	0.25	0.80	3.85	6.50	1.20	
	EUROHIS-QOL-8	
10	Initial/Final	- 0.18 ± 1.42	- 0.58 ± 1.50	36.26 (32)	0.28	N/A	3.85	6.50	1.20	
	WHOQOL-5	
11	Initial/Final	- 0.26 ± 0.68	- 0.64 ± 1.29	14.74 (20)	0.79	0.72	1.92	4.57	- 0.73	
N/A: Non assessment due to local dependency.

The person-item distribution plot for the initial 24-item WHOQOL-BREF indicated a minor ceiling effect, with approximately 0.77% of the total samples not covered by the item thresholds. Conversely, the four-domain subtests had item thresholds that covered participants’ ability to measure the HRQoL construct, indicating no ceiling effects (Fig. 1).Fig. 1 Person-item threshold distributions between person frequencies and item location for (a) initial 24-item WHOQOL-BREF, and (b) four-domain subtests.

As shown in Table 3, the 7-item physical domain showed significant misfit to the Rasch model (χ2 [28] = 150.47, p < 0.01) and showed local dependency, although it achieved acceptable unidimensionality (5.77% significant t-tests; lower bound CI, 3.12) (Analysis 3). To address these issues, five super items were created from items 10, 11, and 12, based on a residual correlation greater than ± 0.2, aiming to achieve an acceptable fit (χ2 [20] = 29.63, p = 0.08) and led to local independency with borderline reliability (PSI = 0.70) (Analysis 4). The five super items also demonstrated strict unidimensionality (2.31% significant t-tests; lower bound CI, − 0.34), supporting local item independence. However, DIF was detected for item 3 (energy) concerning sex. Subsequently, item 3 was split by sex (male/female), with a negligible effect size for the mean difference (0.01), indicating minimal DIF. Therefore, no action was taken for item 3 in the physical domain.

The 6-item psychological domain also showed a significant misfit to the Rasch model (χ2 [24] = 52.70, p < 0.01) despite achieving unidimensionality (2.69% significant t-tests; lower bound CI, 0.04) with good reliability (PSI = 0.78) (Table 3) (Analysis 5). Following a similar approach to the physical domain, a subtest comprising items 7 and 9 was generated. This adjustment allowed the five super items (four items and one super item by merging items 7 and 9) to achieve a satisfactory fit (χ2 [20] = 20.52, p = 0.43) and a strict unidimensionality (1.54% significant t-tests; lower bound CI, − 1.11), indicating the absence of local dependency (Analysis 6). The five super items maintained good reliability (PSI = 0.79), and no DIF was observed across the considered personal factors.

The three-item social domain demonstrated a good fit to the Rasch model (χ2 [12] = 12.16, p = 0.43) and strict unidimensionality (1.92% significant t-tests; lower bound CI, − 0.73) (Analysis 7). However, it showed local dependency, and DIF was also identified for item 25 (sexual activity) concerning the age group. Although graphical inspection provided limited insights, it indicated slight deviations in the age group ≥ 66 years. Item 25 was consequently split by age subgroups (18 to 65/ ≥ 66), revealing a small effect size for mean differences (0.01), which was considered negligible for DIF. Therefore, no further action was taken regarding the DIF of item 25.

The eight-item environmental domain did not meet the fit criteria for the Rasch model (χ2 [32] = 70.70, p < 0.01), and it showed local dependency. However, it achieved unidimensionality (4.62% significant t-tests; lower bound CI, 1.97) (Analysis 8). By creating a subtest consisting of items 17 and 20 as seven super items, the misfit issue was resolved (χ2 [28] = 32.64, p = 0.25). This adjustment also led to acceptable unidimensionality (3.85% significant t-tests; lower bound CI, 1.20) and local indepedency with good reliability (PSI = 0.80) (Analysis 9). No significant DIF was detected for personal factors.

The person-item threshold distribution for individual WHOQOL-BREF domains is presented in Fig. 2. Ceiling effects were observed for physical five super items, psychological five super items, social domains, and environmental seven super items, accounting for 0.77%, 3.17%, 3.09%, and 2.34% of total samples not covered by the item thresholds, respectively.Fig. 2 Person-item threshold distributions between person frequencies and item location for (a) physical 5-super items, (b) psychological 5- super items, (c) social domain, (d) environmental 7-super items.

Table 4 shows the Rasch item locations for EUROHIS-QOL-8 and WHOQOL-5. No misfit items were identified for the shorter versions. The EUROHIS-QOL-8 demonstrated a fit to the Rasch model (χ2 [32] = 36.26, p = 0.28) and unidimensionality (3.85% significant t-tests; lower bound CI, 1.20) (Analysis 10 in Table 3); however, it showed local dependency between items 16 and 17 where the residual correlation (0.12) > 0.2 above the average correlation (− 0.14). The WHOQOL-5 showed an acceptable fit to the Rasch model (χ2 [20] = 14.74, p = 0.79) and unidimensionality (1.92% significant t-tests; lower bound CI, − 0.73) with good reliability (PSI = 0.72) and no local dependency (Analysis 11 in Table 3). No significant DIF was observed for either shorter version. The ceiling effect in the person-item distribution for EUROHIS-QOL-8 and WHOQOL-5 was approximately 3.49% and 2.76%, respectively. (Fig. 3). Table 4 Rasch model item locations, fit residual, and Chi-square for the EUROHIS-QOL-8 and WHOQOL-5 for the patients with warfarin therapy.

Measures	Fit statistics	
Location	S.E.	Individual item fit	χ2	Prob	
1. EUROHIS-QOL-8	
Satisfied with health (Q1)	0.465	0.100	0.101	2.068	0.558	
Energy (Q3)	0.565	0.088	1.522	2.807	0.422	
Self-esteem (Q7)	− 0.272	0.100	− 2.361	8.933	0.030	
Activities for daily living (Q10)	− 0.422	0.100	− 1.946	4.390	0.222	
Personal relationship (Q13)	0.016	0.090	0.139	0.559	0.906	
Home environment (Q16)	− 0.556	0.101	− 0.228	3.052	0.384	
Financial support (Q17)	0.566	0.086	1.618	9.779	0.021	
Quality of life (Q26)	− 0.362	0.101	− 0.288	1.751	0.626	
2. WHOQOL-5	
Satisfied with health (Q1)	0.646	0.100	0.504	6.412	0.379	
Activities for daily living (Q10)	− 0.261	0.101	− 1.339	4.315	0.634	
Personal relationship (Q13)	0.187	0.090	− 0.177	2.782	0.836	
Home environment (Q16)	− 0.379	0.101	− 0.315	4.865	0.561	
Quality of life (Q26)	− 0.193	0.102	0.003	2.117	0.909	

Fig. 3 Person-item threshold distributions between person frequencies and item location for (a) EUROHIS-QOL-8, and (b) WHOQOL-5.

Conversion tables were generated for the four-domain subtests, individual WHOQOL-BREF domain, and WHOQOL-5 to transform ordinal scores into interval scores for further HRQoL computation, provided no missing values were detected. These conversion tables are shown in the supplement materials.

Discussions

This study represents the first investigation into the psychometric properties of the Thai versions of WHOQOL-BREF, EUROHIS-QOL-8, and WHOQOL-5 using Rasch analysis among Thai patients undergoing warfarin treatment.

The four-domain subtests and individual components of the WHOQOL-BREF for physical, psychological, and environmental domains achieved acceptable reliability levels (PSI = 0.70 − 0.87). Depending on the specific measurement objective, it is highly recommended to use the total scores from the four-domain subtests when assessing HRQoL levels for pre- and post-measurement in Thai patients undergoing warfarin treatment because it can produce a higher reliability with a PSI of 0.87, implying that it is a sufficiently reliable instrument (PSI > 0.85)23.

Each of the WHOQOL-BREF domains failed to meet the fit criteria for the Rasch model in item-based analysis, except for the social domain. Thus, the super-item approach was adopted, whereby items with high residual correlations were grouped. This approach improved the fit of all WHOQOL-BREF domains to the Rasch model, including physical, psychological, and environmental domains. Despite both the physical and social domains displaying significant DIF concerning personal factors, the effect size of mean differences was deemed minimal, indicating that all four WHOQOL-BREF domains can effectively measure the HRQoL construct regardless of the specific subgroups being assessed. Similar to previous Rasch analyses conducted in patients with multiple sclerosis48, all four WHOQOL-BREF domains demonstrated unidimensionality, achieved an acceptable fit to the Rasch model, and showed consistent item responses across demographic subgroups following the implementation of the super-item approach. This approach thus provides insights into how each WHOQOL-BREF domain can be used to report HRQol levels effectively in real-world practice for patients undergoing warfarin treatment. Notably, although the social domain achieved an acceptable fit to the Rasch model and no DIF was detected, it demonstrated local dependency between items 13 (personal relationship) and 14 (social support) using Yen’s Q3 statistics. Reliability indices for both Cronbach’s alpha and PSI could not be reported44. Therefore, to measure HRQoL levels among patients undergoing warfarin therapy, individual WHOQOL-BREF domains, except for the social domain, can be used.

In the present study, Rasch analysis revealed five items that did not fit to the Rasch model: pain (item 2), sleep (item 4), and dependence on medical aids (item 11) in the physical domain; leisure activities (item 20) in the environmental domain; and sexual activity (item 25) in the social domain. Notably, previous Rasch analyses in both general population samples and specific therapeutic contexts have also identified misfitting items22,33,49, some of which were negatively worded (e.g., items 2 and 11). This suggests a need for caution when using negatively worded questions to measure HRQoL levels, as respondents may not fully understand their meanings, resulting in invalid responses and potentially biased HRQol estimations. Our findings are consistent with previous Rasch studies conducted in general Thai populations22, where the sexual activity item also showed a misfit to the Rasch model. This indicates that both the general Thai population and patients undergoing warfarin treatment may struggle to understand and respond to this item. This also suggests the need for cognitive debriefing with the Thai population to validate the translation process of misfitting items as subtle differences might reflect the unique importance placed on each facet of the quality of life by Thais. However, these misfit items were only problematic in the baseline model framework. The use of a subtest approach effectively addressed the misfit problem, with all problematic items going undetected when employing this approach. This suggests that the WHOQOL-BREF instrument can be effectively used to report HRQoL scores using the four-domain WHOQOL-BREF subtests rather than focusing on individual items.

Although previous literature is mixed regarding the unconditional item-trait interaction chi-square test from RUMM2030 increasing the possibility of type I error when the sample size is larger than 20050,51, we performed the Rasch analysis with sample size of less than 200 to investigate the robustness of the Rasch properties. All Rasch properties were similar to those of the original analysis (n = 260), which may imply that our results were robust to any changes in sample size and might not increase the possibility of type I errors as shown in the supplement material. Nevertheless, it should be reanalyzed using other Rasch softwares such as DIGRAM, which provides conditional statistics, to confirm this finding in future studies.

The person-item distribution showed minor ceiling effects ranging from 0.77% to 3.17% across the four WHOQOL-BREF domains, indicating that each separate domain can effectively distinguish individual HRQoL scores, as their ceiling effects are below 15%. Moreover, no ceiling effect was observed for the four-domain subtests, indicating that item thresholds adequately covered the participants’ abilities to rate their HRQoL levels.

Similar to previous Rasch analyses conducted on general Thai samples22, the WHOQOL-5 demonstrated an acceptable fit to the Rasch model and unidimensionality. Additionally, no misfit items or significant DIF were detected across sociodemographic factors. The person-item targeting distribution for WHOQOL-5 also showed negligible ceiling effects (2.76%). Nevertheless, the EUROHIS-QOL-8 presented with local dependency, so the reliability was not reported. However, unlike previous Rasch findings22, the WHOQOL-5 had satisfactory reliability (PSI = 0.72 and Cronbach’s alpha = 0.72 for WHOQOL-5). Notably, earlier Rasch analyses indicated that WHOQOL-5 fell short of acceptable reliability (PSI = 0.66)22, possibly due to its limited five-item structure, which typically leads to lower reliability. Moreover, the diversity of health conditions among the general Thai samples in previous study contrasts with the specific inclusion criteria in this study, which focused on individuals with heart diseases undergoing warfarin therapy. This variation in health conditions might have influenced how participants perceived and rated the importance of WHOQOL-5 items differently. Regarding both the short forms, the WHOQOL-5 is highly recommended for HRQoL measurements in group analyses among patients undergoing warfarin treatment. Although the EUROHIS-QOL-8 demonstrated a fit to the Rasch model, unidimensionality, and no DIF, local dependency was observed and reliability indices for both Cronbach’s alpha and PSI were not reported44. Therefore, due to improper Rasch properties, it is not recommended for HRQoL measurements.

Certain limitations need to be addressed. First, this study investigated the psychometric properties of the shorter versions from the samples of the main dataset for WHOQOL-BREF. Second, it generated an ordinal-interval conversion table without testing its measurement properties; thus, future studies should investigate whether the HRQoL scores derived from the conversion tables can discriminate and be responsive to changes in health conditions for patients undergoing warfarin treatment. Third, this study collected data from patients at three public hospitals; consequently, it might not be representative of patients undergoing warfarin therapy in Thailand. Future studies should be conducted with patients from hospitals in each region of Thailand.

Conclusions

Rasch analysis demonstrates that the Thai WHOQOL-BREF is a valid and reliable instrument for measuring and reporting HRQoL scores, particularly in the context of pre-post measurements among patients undergoing warfarin treatment, using the four-domain super items. Certain items, such as item 2 (pain), item 4 (sleep), item 11 (dependence on medical aids), item 20 (leisure activity), and item 25 (sexual activity), did not fit well with the Rasch model. Therefore, cognitive debriefing on misfitting items with the Thai samples to ensure translation validity should be performed in future studies. Nonetheless, these misfit items could be effectively addressed within the super-item approach. In addition, Rasch analysis confirms the suitability of all individual WHOQOL-BREF domains for HRQoL measurement in between-group analyses, except for social relationships, which pose satisfactory reliability (PSI > 0.7). As the EUROHIS-QOL-8 demonstrates a local dependence between items, Rasch analysis only supports the use of WHOQOL-5 for HRQoL measurement, particularly in between-group analysis scenarios for Thai patients undergoing warfarin therapy.

Supplementary Information

Supplementary Information.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-024-71048-4.

Acknowledgements

We would like to express our sincere gratitude to the director of Suan Prung Psychiatric hospital’s director for their valuable permission to use the Thai WHOQOL-BREF for data collection in this study. We also thank our well-trained interviewers for their help in data collection at three public hospitals (Angkawara Songtarn, Ketganok Vannaputtarak, Kornkanok Chana, Pacharinton Poktang, Phatchanida Moorasee, and Puwis Sonphong). Moreover, we would like to thank Thitiya Khongmee, Nampratia Pawasan, and Sitanan Chityam, who are pharmacists at three public hospitals for facilitating data collection process. Special thanks go to all samples at those three public hospitals for their valuable time to participate in this study.  This work does not contain any individual person’s data in any form (including any individual details, images, or videos. 

Author contributions

KK: conceptualization (lead), investigation (equal), formal analysis (lead), methodology (lead), writing the fist draft of manuscript (lead), review and editing the manuscript (lead); NK: conceptualization (supporting), investigation (equal), methodology (supporting), review and editing the manuscript(supporting). All authors edited, read, and approved the final manuscript and are all in agreement with the manuscript. The content has not been published elsewhere.

Data availability

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

Competing interests

The authors declare no competing interests.

Ethical approval and consent to participate

The protocol for the research was approved by Burapha University’s Institutional Review Board (IRB1- 021/2566 Amendment 1) and Chonburi Hospital’s Institutional Review Board (37/65/O/q) and adhered to the Helsinki Declaration, and good clinical practice declaration. 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 participant is ensured that their participation is anonymous and the collected data are confidential.

Publisher's note

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

1. Budreviciute A Management and prevention strategies for non-communicable diseases (NCDs) and their risk factors Front. Public Health 2020 8 574111 10.3389/fpubh.2020.574111 33324597
Budreviciute, A. et al. Management and prevention strategies for non-communicable diseases (NCDs) and their risk factors. Front. Public Health 8, 574111. 10.3389/fpubh.2020.574111 (2020).33324597 10.3389/fpubh.2020.574111
2. Piovani D Nikolopoulos GK Bonovas S Non-communicable diseases: the invisible epidemic J. Clin. Med. 2022 11 19 5939 10.3390/jcm11195939 36233805
Piovani, D., Nikolopoulos, G. K. & Bonovas, S. Non-communicable diseases: the invisible epidemic. J. Clin. Med. 11(19), 5939. 10.3390/jcm11195939 (2022).36233805 10.3390/jcm11195939
3. Haitjema S Health-related quality of life is poor but does not vary with cardiovascular disease burden among patients operated for severe atherosclerotic disease IJC Heart Vessels 2014 4 53 58 10.1016/j.ijchv.2014.07.001
Haitjema, S. et al. Health-related quality of life is poor but does not vary with cardiovascular disease burden among patients operated for severe atherosclerotic disease. IJC Heart Vessels 4, 53–58. 10.1016/j.ijchv.2014.07.001 (2014).10.1016/j.ijchv.2014.07.001
4. Number and death rate of non-communicable Disease in Thailand 2017–2021., <https://www.thaincd.com/2016/mission/documents-detail.php?id=14480&tid=32&gid=1-020> (2021).
5. Xie J Patient-reported health status in coronary heart disease in the United States: age, sex, racial, and ethnic differences Circulation 2008 118 491 497 10.1161/circulationaha.107.7520069 18625894
Xie, J. et al. Patient-reported health status in coronary heart disease in the United States: age, sex, racial, and ethnic differences. Circulation 118, 491–497. 10.1161/circulationaha.107.7520069 (2008).18625894 10.1161/circulationaha.107.7520069
6. Ko HY Lee JK Shin JY Jo E Health-related quality of life and cardiovascular disease risk in Korean adults Korean J. Fam. Med. 2015 36 349 356 10.4082/kjfm.2015.36.6.349 26634104
Ko, H. Y., Lee, J. K., Shin, J. Y. & Jo, E. Health-related quality of life and cardiovascular disease risk in Korean adults. Korean J. Fam. Med. 36, 349–356. 10.4082/kjfm.2015.36.6.349 (2015).26634104 10.4082/kjfm.2015.36.6.349
7. Radaideh KM Matalqah LM Health-related quality of life among atrial fibrillation patients using warfarin therapy Epidemiol. Biostat. Public Health 2018 15 e12763
Radaideh, K. M. & Matalqah, L. M. Health-related quality of life among atrial fibrillation patients using warfarin therapy. Epidemiol. Biostat. Public Health 15, e12763 (2018).
8. Yusuf S Reddy S Ôunpuu S Anand S Global burden of cardiovascular diseases: Part II: variations in cardiovascular disease by specific ethnic groups and geographic regions and prevention strategies Circulation 2001 104 2855 2864 10.1161/hc4701.099488 11733407
Yusuf, S., Reddy, S., Ôunpuu, S. & Anand, S. Global burden of cardiovascular diseases: Part II: variations in cardiovascular disease by specific ethnic groups and geographic regions and prevention strategies. Circulation 104, 2855–2864 (2001).11733407 10.1161/hc4701.099488
9. Dipiro JT Pharmacotherapy: a pathophysiologic approach Connecticut Appleton Lange 2014 4 141 142
Dipiro, J. T. et al. Pharmacotherapy: a pathophysiologic approach. Connecticut Appleton Lange 4, 141–142 (2014).
10. Chu, V. L. & Maltz, H. C. in Casebook in Clinical Pharmacokinetics and Drug Dosing (ed Henry Cohen) (McGraw-Hill Education, 2015).
11. Kulthanachairojana N Health-related quality of life and willingness to pay measurement among patients on warfarin in Thailand J. Pharm. Policy Pract. 2023 16 130 10.1186/s40545-023-00632-2 37904238
Kulthanachairojana, N. et al. Health-related quality of life and willingness to pay measurement among patients on warfarin in Thailand. J. Pharm. Policy Pract. 16, 130. 10.1186/s40545-023-00632-2 (2023).37904238 10.1186/s40545-023-00632-2
12. Samsa G A new instrument for measuring anticoagulation-related quality of life: development and preliminary validation Health Qual Life Outcomes 2004 2 22 10.1186/1477-7525-2-22 15132746
Samsa, G. et al. A new instrument for measuring anticoagulation-related quality of life: development and preliminary validation. Health Qual Life Outcomes 2, 22. 10.1186/1477-7525-2-22 (2004).15132746 10.1186/1477-7525-2-22
13. Gafou A Maragos K Bellia M Digenopoulou-Andrioti E Theodosiadis G Instruments for measuring anticoagulation-related quality of life: modification, and preliminary validation Haema 2007 10 129 141
Gafou, A., Maragos, K., Bellia, M., Digenopoulou-Andrioti, E. & Theodosiadis, G. Instruments for measuring anticoagulation-related quality of life: modification, and preliminary validation. Haema 10, 129–141 (2007).
14. Mameli A Sestu A Marongiu F Barcellona D Living on oral anticoagulants: duke anticoagulation satisfaction scale results J. Clin. Med. 2023 12 24 7574 10.3390/jcm12247574 38137642
Mameli, A., Sestu, A., Marongiu, F. & Barcellona, D. Living on oral anticoagulants: duke anticoagulation satisfaction scale results. J. Clin. Med. 12(24), 7574. 10.3390/jcm12247574 (2023).38137642 10.3390/jcm12247574
15. Schoen RR Nagy MW Porter AL Margolis AR Patient satisfaction with extended international normalized ratio follow-up intervals in a veteran population Ann. Pharmacother. 2020 54 442 449 10.1177/1060028019889414 31752504
Schoen, R. R., Nagy, M. W., Porter, A. L. & Margolis, A. R. Patient satisfaction with extended international normalized ratio follow-up intervals in a veteran population. Ann. Pharmacother. 54, 442–449. 10.1177/1060028019889414 (2020).31752504 10.1177/1060028019889414
16. Fang MC Health-related quality of life associated with warfarin and direct oral anticoagulants in venous thromboembolism Thromb Res. 2022 216 97 102 10.1016/j.thromres.2022.06.008 35779378
Fang, M. C. et al. Health-related quality of life associated with warfarin and direct oral anticoagulants in venous thromboembolism. Thromb Res. 216, 97–102. 10.1016/j.thromres.2022.06.008 (2022).35779378 10.1016/j.thromres.2022.06.008
17. Iqbal MS Assessing quality of life using WHOQOL-BREF: A cross-sectional insight among patients on warfarin in Malaysia Saudi Pharm. J. 2020 28 936 942 10.1016/j.jsps.2020.06.014 32792838
Iqbal, M. S. et al. Assessing quality of life using WHOQOL-BREF: A cross-sectional insight among patients on warfarin in Malaysia. Saudi Pharm. J. 28, 936–942 (2020).32792838 10.1016/j.jsps.2020.06.014
18. Iqbal MS Determinants of health-related quality of life among warfarin patients in Pakistan PloS ONE 2020 15 e0234734 10.1371/journal.pone.0234734 32555595
Iqbal, M. S. et al. Determinants of health-related quality of life among warfarin patients in Pakistan. PloS ONE 15, e0234734 (2020).32555595 10.1371/journal.pone.0234734
19. The Whoqol Group The world health organization quality of life assessment (WHOQOL): development and general psychometric properties Soc. Sci. Med. 1998 46 1569 1585 10.1016/s0277-9536(98)00009-4 9672396
The Whoqol Group. The world health organization quality of life assessment (WHOQOL): development and general psychometric properties. Soc. Sci. Med. 46, 1569–1585. 10.1016/s0277-9536(98)00009-4 (1998).9672396 10.1016/s0277-9536(98)00009-4
20. Schmidt S Mühlan H Power M The EUROHIS-QOL 8-item index: psychometric results of a cross- cultural field study Eur. J. Public Health 2006 16 420 428 10.1093/eurpub/cki155 16141303
Schmidt, S., Mühlan, H. & Power, M. The EUROHIS-QOL 8-item index: psychometric results of a cross- cultural field study. Eur. J. Public Health 16, 420–428. 10.1093/eurpub/cki155 (2006).16141303 10.1093/eurpub/cki155
21. Geyh S Fellinghauer BA Kirchberger I Post MW Cross-cultural validity of four quality of life scales in persons with spinal cord injury Health Qual Life Outcomes 2010 8 94 10.1186/1477-7525-8-94 20815864
Geyh, S., Fellinghauer, B. A., Kirchberger, I. & Post, M. W. Cross-cultural validity of four quality of life scales in persons with spinal cord injury. Health Qual Life Outcomes 8, 94. 10.1186/1477-7525-8-94 (2010).20815864 10.1186/1477-7525-8-94
22. Kangwanrattanakul K Krägeloh CU Psychometric evaluation of the WHOQOL-BREF and its shorter versions for general Thai population: confirmatory factor analysis and Rasch analysis Qual. Life Res. 2024 33 335 348 10.1007/s11136-023-03521-y 37906345
Kangwanrattanakul, K. & Krägeloh, C. U. Psychometric evaluation of the WHOQOL-BREF and its shorter versions for general Thai population: confirmatory factor analysis and Rasch analysis. Qual. Life Res. 33, 335–348. 10.1007/s11136-023-03521-y (2024).37906345 10.1007/s11136-023-03521-y
23. Balalla SK Medvedev ON Siegert RJ Krägeloh CU Validation of the WHOQOL-BREF and shorter versions using rasch analysis in traumatic brain injury and orthopedic populations Arch. Phys. Med. Rehabil. 2019 100 1853 1862 10.1016/j.apmr.2019.05.029 31229529
Balalla, S. K., Medvedev, O. N., Siegert, R. J. & Krägeloh, C. U. Validation of the WHOQOL-BREF and shorter versions using rasch analysis in traumatic brain injury and orthopedic populations. Arch. Phys. Med. Rehabil. 100, 1853–1862. 10.1016/j.apmr.2019.05.029 (2019).31229529 10.1016/j.apmr.2019.05.029
24. Sakthong P Schommer JC Gross CR Sakulbumrungsil R Prasithsirikul W Psychometric properties of WHOQOL-BREF-THAI in patients with HIV/AIDS J. Med. Assoc. Thai. 2007 90 2449 2460 18181334
Sakthong, P., Schommer, J. C., Gross, C. R., Sakulbumrungsil, R. & Prasithsirikul, W. Psychometric properties of WHOQOL-BREF-THAI in patients with HIV/AIDS. J. Med. Assoc. Thai. 90, 2449–2460 (2007).18181334
25. Abdullah Bandar NF Jani R Karim MA Psychometric properties of the WHOQOL-BREF questionnaire among disabled students in malaysian higher learning institutions Applied Res. Quality Life 2014 9 469 478 10.1007/s11482-013-9244-8
Abdullah Bandar, N. F., Jani, R. & Karim, M. A. Psychometric properties of the WHOQOL-BREF questionnaire among disabled students in malaysian higher learning institutions. Applied Res. Quality Life 9, 469–478. 10.1007/s11482-013-9244-8 (2014).10.1007/s11482-013-9244-8
26. Yoshitake N The psychometric properties of the WHOQOL-BREF in Japanese couples Health Psychol. Open 2015 2 2055102915598089 10.1177/2055102915598089 28070365
Yoshitake, N. et al. The psychometric properties of the WHOQOL-BREF in Japanese couples. Health Psychol. Open 2, 2055102915598089. 10.1177/2055102915598089 (2015).28070365 10.1177/2055102915598089
27. Rosén H Ahlström G Lexén A Psychometric properties of the WHOQOL-BREF among next of kin to older persons in nursing homes Health Qual. Life Outcomes 2020 18 103 10.1186/s12955-020-01345-9 32306965
Rosén, H., Ahlström, G. & Lexén, A. Psychometric properties of the WHOQOL-BREF among next of kin to older persons in nursing homes. Health Qual. Life Outcomes 18, 103. 10.1186/s12955-020-01345-9 (2020).32306965 10.1186/s12955-020-01345-9
28. Krägeloh CU Validation of the WHOQOL-BREF quality of life questionnaire for general use in New Zealand: confirmatory factor analysis and Rasch analysis Qual. Life Res. 2013 22 1451 1457 10.1007/s11136-012-0265-9 22983781
Krägeloh, C. U. et al. Validation of the WHOQOL-BREF quality of life questionnaire for general use in New Zealand: confirmatory factor analysis and Rasch analysis. Qual. Life Res. 22, 1451–1457. 10.1007/s11136-012-0265-9 (2013).22983781 10.1007/s11136-012-0265-9
29. Mahatnirundkul S Comparison of the WHOQOL-100 and the WHOQOL-BREF (26 items) J. Ment. Health Thai 1998 5 4 15
Mahatnirundkul, S. Comparison of the WHOQOL-100 and the WHOQOL-BREF (26 items). J. Ment. Health Thai 5, 4–15 (1998).
30. Taboonpong S Suttharangsee W Chailangka P Evaluating psychometric properties of WHO quality of life questionnaire in Thai elderly J. Gerontol. Geriatric. Med. 2001 2 6 12
Taboonpong, S., Suttharangsee, W. & Chailangka, P. Evaluating psychometric properties of WHO quality of life questionnaire in Thai elderly. J. Gerontol. Geriatric. Med. 2, 6–12 (2001).
31. Phungrassami T Katikarn R Watanaarepornchai S Sangtawan D Quality of life assessment in radiotherapy patients by WHOQOL-BREF-THAI: a feasibility study J. Med. Assoc. Thai 2004 87 1459 1465 15822541
Phungrassami, T., Katikarn, R., Watanaarepornchai, S. & Sangtawan, D. Quality of life assessment in radiotherapy patients by WHOQOL-BREF-THAI: a feasibility study. J. Med. Assoc. Thai 87, 1459–1465 (2004).15822541
32. Li K Kay NS Nokkaew N The performance of the world health organization's WHOQOL-BREF in assessing the quality of life of thai college students Soc. Indicators Res. 2009 90 489 501 10.1007/s11205-008-9272-1
Li, K., Kay, N. S. & Nokkaew, N. The performance of the world health organization’s WHOQOL-BREF in assessing the quality of life of thai college students. Soc. Indicators Res. 90, 489–501 (2009).10.1007/s11205-008-9272-1
33. Chang K-C Wang J-D Tang H-P Cheng C-M Lin C-Y Psychometric evaluation, using Rasch analysis, of the WHOQOL-BREF in heroin-dependent people undergoing methadone maintenance treatment: further item validation Health Qual. Life Outcomes 2014 12 148 10.1186/s12955-014-0148-6 25277717
Chang, K.-C., Wang, J.-D., Tang, H.-P., Cheng, C.-M. & Lin, C.-Y. Psychometric evaluation, using Rasch analysis, of the WHOQOL-BREF in heroin-dependent people undergoing methadone maintenance treatment: further item validation. Health Qual. Life Outcomes 12, 148. 10.1186/s12955-014-0148-6 (2014).25277717 10.1186/s12955-014-0148-6
34. Wang WC Yao G Tsai YJ Wang JD Hsieh CL Validating, improving reliability, and estimating correlation of the four subscales in the WHOQOL-BREF using multidimensional Rasch analysis Qual. Life Res. 2006 15 607 620 10.1007/s11136-005-4365-7 16688494
Wang, W. C., Yao, G., Tsai, Y. J., Wang, J. D. & Hsieh, C. L. Validating, improving reliability, and estimating correlation of the four subscales in the WHOQOL-BREF using multidimensional Rasch analysis. Qual. Life Res. 15, 607–620. 10.1007/s11136-005-4365-7 (2006).16688494 10.1007/s11136-005-4365-7
35. Kaplan RM Hays RD Health-related quality of life measurement in public health Annu. Rev. Public Health 2022 43 355 373 10.1146/annurev-publhealth-052120-012811 34882431
Kaplan, R. M. & Hays, R. D. Health-related quality of life measurement in public health. Annu. Rev. Public Health 43, 355–373. 10.1146/annurev-publhealth-052120-012811 (2022).34882431 10.1146/annurev-publhealth-052120-012811
36. Uddin MN Islam FMA Psychometric evaluation of an interview-administered version of the WHOQOL-BREF questionnaire for use in a cross-sectional study of a rural district in Bangladesh: an application of Rasch analysis BMC Health Serv. Res. 2019 19 216 10.1186/s12913-019-4026-0 30953506
Uddin, M. N. & Islam, F. M. A. Psychometric evaluation of an interview-administered version of the WHOQOL-BREF questionnaire for use in a cross-sectional study of a rural district in Bangladesh: an application of Rasch analysis. BMC Health Serv. Res. 19, 216. 10.1186/s12913-019-4026-0 (2019).30953506 10.1186/s12913-019-4026-0
37. Robinson M Johnson AM Walton DM MacDermid JC A comparison of the polytomous Rasch analysis output of RUMM2030 and R (ltm/eRm/TAM/lordif) BMC Med. Res. Methodol. 2019 19 36 10.1186/s12874-019-0680-5 30786868
Robinson, M., Johnson, A. M., Walton, D. M. & MacDermid, J. C. A comparison of the polytomous Rasch analysis output of RUMM2030 and R (ltm/eRm/TAM/lordif). BMC Med. Res. Methodol. 19, 36. 10.1186/s12874-019-0680-5 (2019).30786868 10.1186/s12874-019-0680-5
38. Fan, X. & Sun, S. Item response theory. In Handbook of Quantitative Methods for Educational Research (ed Timothy Teo) 45–67 (SensePublishers, 2013).
39. Demir S Comparison of normality tests in terms of sample sizes under different skewness and kurtosis coefficients Int. J. Assessment Tools Educat. 2022 9 397 409 10.21449/ijate.1101295
Demir, S. Comparison of normality tests in terms of sample sizes under different skewness and kurtosis coefficients. Int. J. Assessment Tools Educat. 9, 397–409 (2022).10.21449/ijate.1101295
40. Terwee CB Quality criteria were proposed for measurement properties of health status questionnaires J. Clin. Epidemiol. 2007 60 34 42 10.1016/j.jclinepi.2006.1003.1012 17161752
Terwee, C. B. et al. Quality criteria were proposed for measurement properties of health status questionnaires. J. Clin. Epidemiol. 60, 34–42. 10.1016/j.jclinepi.2006.1003.1012 (2007).17161752 10.1016/j.jclinepi.2006.1003.1012
41. Tennant A Küçükdeveci AA Application of the Rasch measurement model in rehabilitation research and practice: early developments, current practice, and future challenges Front. Rehabil. Sci. 2023 4 1208670 10.3389/fresc.2023.1208670 37529206
Tennant, A. & Küçükdeveci, A. A. Application of the Rasch measurement model in rehabilitation research and practice: early developments, current practice, and future challenges. Front. Rehabil. Sci. 4, 1208670. 10.3389/fresc.2023.1208670 (2023).37529206 10.3389/fresc.2023.1208670
42. Christensen KB Makransky G Horton M Critical values for Yen’s Q3: identification of local dependence in the rasch model using residual correlations Appl. Psychol. Measure. 2017 41 178 194 10.1177/0146621616677520
Christensen, K. B., Makransky, G. & Horton, M. Critical values for Yen’s Q3: identification of local dependence in the rasch model using residual correlations. Appl. Psychol. Measure. 41, 178–194. 10.1177/0146621616677520 (2017).10.1177/0146621616677520
43. Skevington SM Rowland C Panagioti M Bower P Krägeloh C Enhancing the multi-dimensional assessment of quality of life: introducing the WHOQOL-Combi Qual. Life Res. 2021 30 891 903 10.1007/s11136-020-02661-9 33331967
Skevington, S. M., Rowland, C., Panagioti, M., Bower, P. & Krägeloh, C. Enhancing the multi-dimensional assessment of quality of life: introducing the WHOQOL-Combi. Qual. Life Res. 30, 891–903. 10.1007/s11136-020-02661-9 (2021).33331967 10.1007/s11136-020-02661-9
44. Marais, I. Local dependence. Rasch models in health, 111–130 (2012).
45. Andrich D Hagquist C Real and artificial differential item functioning in polytomous items Educ. Psychol. Meas. 2015 75 185 207 10.1177/0013164414534258 29795818
Andrich, D. & Hagquist, C. Real and artificial differential item functioning in polytomous items. Educ. Psychol. Meas. 75, 185–207. 10.1177/0013164414534258 (2015).29795818 10.1177/0013164414534258
46. Rouquette A Hardouin JB Vanhaesebrouck A Sébille V Coste J Differential item functioning (DIF) in composite health measurement scale: Recommendations for characterizing DIF with meaningful consequences within the Rasch model framework PLoS ONE 2019 14 0215073 10.1371/journal.pone.0215073
Rouquette, A., Hardouin, J. B., Vanhaesebrouck, A., Sébille, V. & Coste, J. Differential item functioning (DIF) in composite health measurement scale: Recommendations for characterizing DIF with meaningful consequences within the Rasch model framework. PLoS ONE 14, 0215073. 10.1371/journal.pone.0215073 (2019).10.1371/journal.pone.0215073
47. Romanoski J Douglas G Test scores, measurement, and the use of analysis of variance: an historical overview J. Appl. Meas. 2002 3 232 242 12147911
Romanoski, J. & Douglas, G. Test scores, measurement, and the use of analysis of variance: an historical overview. J. Appl. Meas. 3, 232–242 (2002).12147911
48. Pomeroy IM Tennant A Mills RJ Young CA The WHOQOL-BREF: a modern psychometric evaluation of its internal construct validity in people with multiple sclerosis Qual. Life Res. 2020 29 1961 1972 10.1007/s11136-020-02463-z 32193839
Pomeroy, I. M., Tennant, A., Mills, R. J. & Young, C. A. The WHOQOL-BREF: a modern psychometric evaluation of its internal construct validity in people with multiple sclerosis. Qual. Life Res. 29, 1961–1972. 10.1007/s11136-020-02463-z (2020).32193839 10.1007/s11136-020-02463-z
49. Lin C-Y Psychometric evaluation of the WHOQOL-BREF, Taiwan version, across five kinds of Taiwanese cancer survivors: Rasch analysis and confirmatory factor analysis J. Formosan Med. Assoc. 2019 118 215 222 10.1016/j.jfma.2018.03.018 29661488
Lin, C.-Y. et al. Psychometric evaluation of the WHOQOL-BREF, Taiwan version, across five kinds of Taiwanese cancer survivors: Rasch analysis and confirmatory factor analysis. J. Formosan Med. Assoc. 118, 215–222. 10.1016/j.jfma.2018.03.018 (2019).29661488 10.1016/j.jfma.2018.03.018
50. Müller M Item fit statistics for Rasch analysis: can we trust them? J. Stat. Distribut. Appl. 2020 7 5 10.1186/s40488-020-00108-7
Müller, M. Item fit statistics for Rasch analysis: can we trust them?. J. Stat. Distribut. Appl. 7, 5. 10.1186/s40488-020-00108-7 (2020).10.1186/s40488-020-00108-7
51. Hagell P Westergren A Sample size and statistical conclusions from tests of fit to the rasch model according to the rasch unidimensional measurement model (Rumm) program in health outcome measurement J. Appl. Meas. 2016 17 416 431 28009589
Hagell, P. & Westergren, A. Sample size and statistical conclusions from tests of fit to the rasch model according to the rasch unidimensional measurement model (Rumm) program in health outcome measurement. J. Appl. Meas. 17, 416–431 (2016).28009589
