Data collection
A cross-sectional online survey was conducted in May–June 2020 among the general adult population (age ≥ 18) of Hungary. Recruitment of respondents and data collection were performed by a survey company (New Land Media Kft. – Századvég) involving members of an online access panel. Details on participation rate were confidential. The company was responsible also for data protection and compliance with GDPR. Anonymized data were handled for the analyses. Quota sampling (a non-probability sampling method) was used in order to obtain a representative sample of the normal adult population. Quotas were set for age, sex, educational level and type of settlement according to the 2011 Population Census [25]. The targeted sample size was 2000, which is the double of the usually needed sample size (N = 1000) to achieve representativeness defined by the above detailed criteria. Nonetheless, we aimed to achieve a larger than a minimum sample size in all age-groups and a remarkable number of participants with MSK problems. Ethical approval was obtained from the Hungarian Medical Research Council (no. IV/565–5/2020/EKU). Respondents were informed that participation in the survey was voluntary, their data would remain anonymous, impersonal and would be used solely for scientific purposes. Before starting the survey, informed consent was obtained from all individual participants included in the study.
The questionnaire
The survey questionnaire consisted of two modules. The first module (reported in this paper) focused on MSK health, the second one assessed respondents’ subjective expectations towards the length and quality of their life. Socio-demographic characteristics of the sample was recorded. Participants completed the Hungarian version of MSK-HQ along with standard outcome measures of health status, HRQoL, physical functioning and well-being. MSK health problems and diagnoses were explored with relevant questions of the European Health Interview Survey (EHIS) [22]. Respondents were asked to indicate if they used healthcare services (hospital admission, specialist care, general practitioner) or received informal care due to MSK problems. Informal caregivers were also identified by self-reports regardless of MSK health status. To assess test–retest reliability, the completion of MSK-HQ was repeated in the end of the survey in a subgroup of 50 respondents who were selected randomly from the study sample.
Outcome measures
MSK-HQ and its translation and validation for Hungary
The MSK-HQ was developed in 2016 as a specific outcome measure to be used by people with different MSK conditions [8]. The questionnaire assesses how much respondents’ MSK problems have affected their life in the last two weeks. It consists of 14 items, from which eleven measure symptoms and HRQoL, while three focus on patients’ attitude towards their condition and the overall impact of symptoms. Response options are operationalized on a five-point Likert scale, ranging from 4 to 0 (ʻnot at all’ and ʻextremely’, respectively), except items 12 and 13, which are given in reverse order. The final score, ranging from 0 to 56, is calculated by summing up the scores given for each item. Higher scores indicate better MSK health status. An additional, fifteenth item assesses the number of days with significant physical activity in the past week but this item is not included in the final score.
The validation of the Hungarian version was carried out in accordance with the protocol provided by the developer of the original MSK-HQ (Oxford University Innovation). In brief, forward translations were performed independently by three researchers. A reconciled Hungarian version was produced via discussions which was then back translated independently by two experts. Results were reviewed by the developer and items with uncertanities went through a second round forward-back translation cycle. Cognitive debriefing was performed involving five patients with diverse MSK problems. After proofreading and quality check, the final Hungarian version was accepted as the best available language version by the developer.
MEHM
The Minimum European Health Module (MEHM) is composed of three questions concerning three different aspects of health: self-percieved health (response options: very good, good, fair, bad, very bad), long standing illness or chronic morbidity (response options: yes or no) and the presence of long-standing activity limitations measured by the Global Activity Limitation Indicator (GALI) (reponse options: severely limited, limited but not severely, not limited at all) [26].
EQ-5D-5L
The EQ-5D-5L is a generic questionnaire that evaluates HRQoL by assessing the following five domains: mobility, self-care, usual activities, pain/discomfort, anxiety/depression [27]. Respondents are asked to indicate their current health state on a five-level Likert-scale (ranging from 1 –’no problems’ to 5 –’unable to’). To calculate EQ-5D-5L index score, we used the Hungarian tariffs (range: -0.848 – 1.000) [28]. EQ VAS is an additional item that measures respondents self-reported health on a visual analogue scale, ranging from 0 to 100, where 0 indicates the worst and 100 indicates the best health status that the respondent can imagine. Higher scores indicate better health status.
HAQ-DI
Health Assessment Questionnaire – Disability Index (HAQ-DI) [29] measures functional ability over the past week with 20 items across the following 8 domains: dressing, arising, eating, walking, hygiene, reach, grip and common activities. Each domain contains 2 or 3 items. Possible response options for each item range from 0 to 3 (0—ʻwithout difficulty’ and 3—ʻunable to do’, respectively). Eighteen additional items are provided to indicate if aids or assisstive devices were used for the activities listed in the domains. To calculate the HAQ-DI score, we applied the alternative scoring method (no correction for aid or devices was made) thus the highest item score within each domain was considered to calculate the total score (range 0–3, higher score indicates worse status).
ICECAP-A and ICECAP-O
The ICECAP-A and ICECAP-O tools were developed to evaluate capability well-being of adults (18 +) and older people (65 +), respectively [30, 31]. Each version contains five items. For ICECAP-A these are attachment, stability, achievement, enjoyment and autonomy, while ICECAP-O items are attachment, security, role, enjoyment and control. Respondents are asked to indicate their current capability well-being on a 4-level scale (ranging from 4—ʻfull capability’ to 1—ʻno capability’). Index scores for each state can be calculated by using tariffs obtained from population level valuation studies. We used the validated Hungarian version of the questionnaire [32]. At the time of our analysis tariffs were available only for the United Kingdom so we used those for the calculations [31, 33].
WHO-5
The World Health Organisation-Five Well-Being Index (WHO-5) evaluates self-reported mental well-being in relation to the last two weeks [34]. The questionnaire comprises five items. Response options are available on a six-level scale (ranging from 0—ʻat no time’ to 5—ʻall of the time’). The final score is calculated by the multiplication of the sum of item scores by 4. Higher scores indicate better well-being.
Happiness VAS
A visual analogue scale (VAS) of happiness was applied to measure the current subjective degree of happiness of respondents (where 0 indicates ʻcompletely unhappy’ and 10 indicates ʻcompletely happy’).
Statistical analysis
Socio-demographic characteristics and health status of the sample were analyzed using descriptive statistics. The psychometric characteristics of MSK-HQ were assessed in relation to socio-demographic characteristics, self-reported MSK status and diagnosis, healthcare and informal care utilisation, general health status (MEHM), HRQoL (EQ-5D-5L), physical functioning (HAQ-DI) and well-being (ICECAP-A/-O, WHO-5, Happiness VAS). The COnsensus-based Standards for the selection of health status Measurement INstruments (COSMIN) was applied so that we investigated construct validity (clinical, convergent, discriminant validity), internal consistency and reliability [35].
Validity
To assess clinical validity, we investigated whether the MSK-HQ can differentiate between subgroups using Mann–Whitney-U and Kruskal–Wallis tests for comparisons. The association between MSK-HQ score and socio-demographic characteristics were further investigated using ordinary least square (OLS) multiple regression analysis.
To evaluate convergent validity, Spearman’s correlation was calculated between MSK-HQ, EQ-5D-5L, EQ VAS, HAQ-DI, ICECAP-A/-O, WHO-5 and Happiness VAS scores. We also investigated correlations between MSK-HQ items, EQ-5D-5L domains and HAQ-DI domains. We expected strong correlation with the ʻMobility’ and ʻPain/discomfort’ domains of the EQ-5D-5L. The correlations were considered to be strong if the coefficient was over 0.5, moderate between 0.5 and 0.3, and weak under 0.3 [36]. In the multiple regression analysis we also explored associations between MSK-HQ, HRQoL (EQ-5D-5L) and functional status (HAQ-DI) controlling for socio-demographic characteristics. Three models were developed. In the first one, we included the EQ-5D-5L score, in the second one, the HAQ-DI, and in the third model, we included both the EQ-5D-5L and HAQ-DI score to examine how these two scales influence each other’s effect. All models were controlled for basic sociodemographic variables (sex, age, education, residency, married/having a partner, living with someone, paid job, income).
EQ-5D-5L index, EQ VAS, ICECAP-A/-O and HAQ-DI scores were calculated for response options of each MSK-HQ items and differences were compared to determine discriminant validity.
Reliability
Internal consistency was examined by Cronbach’s alpha (0.7–0.8: acceptable, 0.8–0.9: good > 0.9: excellent) [37]. To assess test–retest reliability in the subsample of 50 participants who have completed the MSK-HQ repeatedly, intraclass correlation coefficient (ICC) was calculated. The ICC can range between 0 (no agreement) and 1 (perfect agreement), indicating the level of agreement (< 0.5 –’poor’, 0.50—0.749 –’moderate’, 0.75—0.900 –’good’, > 0.90 –’excellent’) [38].
Significance level for all tests was accepted as p < 0.05. Response option of ʻI do not know / I do not want to answer’ was provided for some items (paid work, income level, MEHM, happinness, EHIS and health care utilisation questions). Respondents who indicated this answer were excluded from the respective analysis of that specific item but their share is provided as ʻnot reported’ among the results. The analysis was performed using IBM SPSS version 25.0 software (IBM Corp., Armonk, NY, USA).