An Investigation of Age-Related Differential Item Functioning in the EQ-5D-5L Using Item Response Theory and Logistic Regression

The performance relationship between the EQ-5D-5L composite "Anxiety/Depression" dimension and anxiety and depression symptoms in a large, general population sample

PURPOSE

This cross-sectional study aims to understand the relationship between responses on the Anxiety/Depression (A/D) dimension of the EQ-5D-5L and symptoms of anxiety and depression on the GAD-7 and PHQ-9 instruments. In doing so, we investigate the comparative performance of the dimension between diagnostic groups (i.e. anxiety (GAD-7); depression (PHQ-9); anxiety & depression versus none). We additionally investigate the discriminatory performance between sub-populations based on gender, age, education and self-reported chronic conditions.

METHODS

19,902 general population participants completed a health survey in May/June 2020, from five European countries and the United States. Performance of A/D was calculated using the Area Under the Receiver Operating Characteristic curve (AUROC), and was compared to having anxiety (GAD-7 ≥ 8), depression (PHQ-9 ≥ 10) and both versus none for the total population and sub-populations. Several additional sensitivity analyses were conducted, including calculations of the optimal A/D cut-off.

RESULTS

The performance in the total sample was good (AUROC > 0.8) and did not differ significantly between diagnostic groups. The performance differed significantly between the age groups, with worse performance in the younger groups, and differed between those with a singular chronic condition, with worse performance in those indicating having an anxiety or depression disorder. The performance did not differ significantly by gender, education, nor total chronic conditions.

CONCLUSION

The A/D dimension captures symptoms of anxiety, depression or both equally well. Performance is worse in the younger population. Interpretation in those with a self-reported anxiety or depression disorder should be further investigated. This is the first-of-its-kind large population sample performance analysis, where we present evidence that the performance of the A/D dimension differs between ages, and thus intra-age comparative results may be flawed.

Outcome measures in hemophilia: current and future perspectives

INTRODUCTION

Hemophilia can detrimentally affect patients' quality of life and likelihood of survival. In the evolving landscape of therapies, the therapeutic gain of each treatment must be understood to accurately position it in the therapeutic armamentarium. Accordingly, appropriate outcomes must be measured with appropriate tools.

AREAS COVERED

Our narrative review (PubMed search for 'hemophilia AND outcome' until June 2023), provides a compendium of outcome measures used in hemophilia clinical research. To define each outcome measure's relative value and applicability, several characteristics are critically discussed.

EXPERT OPINION

Bleeding assessment, first annual/annualized bleeding rate, remains central in evaluating the efficacy and safety of hemophilia treatments. As modern therapies improve clinical outcomes toward zero bleeding events, this endpoint may become less sensitive to detect differences between therapeutic approaches. Technological advancements necessitate the adaptation of outcome measures to address infrequent bleeding events, age-related comorbidities, and laboratory parameters with limited comparability after different treatments. Considerable effort has been dedicated to the development of tools that comprehensively assess coagulation, such as thrombin generation assays. Patient-reported outcome measures are gaining importance although limited by their subjectivity. A definitive set of research outcome measures remains elusive. Outcomes may need to be tailored to different therapeutic interventions.

The EORTC QLU-C10D is a valid cancer-specific preference-based measure for cost-utility and health technology assessment in the Netherlands

BACKGROUND

Cost-utility analysis typically relies on preference-based measures (PBMs). While generic PBMs are widely used, disease-specific PBMs can capture aspects relevant for certain patient populations. Here the EORTC QLU-C10D, a cancer-specific PBM based on the QLQ-C30, is validated using Dutch trial data with the EQ-5D-3L as a generic comparator measure.

METHODS

We retrospectively analysed data from four Dutch randomised controlled trials (RCTs) comprising the EORTC QLQ-C30 and the EQ-5D-3L. Respective Dutch value sets were applied. Correlations between the instruments were calculated for domains and index scores. Bland-Altman plots and intra-class correlations (ICC) displayed agreement between the measures. Independent and paired t-tests, effect sizes and relative validity indices were used to determine the instruments' performance in detecting clinically known-group differences and health changes over time.

RESULTS

We analysed data from 602 cancer patients from four different trials. In overall, the EORTC QLU-C10D showed good relative validity with the EQ-5D-3L as a comparator (correlations of index scores r = 0.53-0.75, ICCs 0.686-0.808, conceptually similar domains showed higher correlations than dissimilar domains). Most importantly, it detected 63% of expected clinical group differences and 50% of changes over time in patients undergoing treatment. Both instruments showed poor performance in survivors. Detection rate and measurement efficiency were clearly higher for the QLU-C10D than for the EQ-5D-3L.

CONCLUSIONS

The Dutch EORTC QLU-C10D showed good comparative validity in patients undergoing treatment. Our results underline the benefit that can be achieved by using a cancer-specific PBM for generating health utilities for cancer patients from a measurement perspective.

Time perspective profile and self-reported health on the EQ-5D

OBJECTIVES

Time perspective (TP) is a psychological construct that is associated with several health-related behaviours, including healthy eating, smoking and adherence to medications. In this study, we aimed to examine the associations of TP profile with self-reported health on the EQ-5D-5L and to detect which domains display response heterogeneity (cut-point shift) for TP.

METHODS

We conducted a secondary analysis of EQ-5D-5L data from a representative general population sample in Hungary (n = 996). The 17-item Zimbardo Time Perspective Inventory was used to measure individuals' TP on five subscales: past-negative, past-positive, present-fatalist, present-hedonist and future. The associations between TP subscales and EQ-5D-5L domain scores, EQ VAS and EQ-5D-5L index values were analysed by using partial proportional odds models and multivariate linear regressions.

RESULTS

Respondents that scored higher on the past-negative and present-fatalist and lower on the present-hedonist and future subscales were more likely to report more health problems in at least one EQ-5D-5L domain (p < 0.05). Adjusting for socio-economic and health status, three EQ-5D-5L domains exhibited significant associations with various TP subscales (usual activities: present-fatalist and future, pain/discomfort: past-negative and future, anxiety/depression: past-negative, present-fatalist, present-hedonist and future). The anxiety/depression domain showed evidence of cut-point shift.

CONCLUSIONS

This study identified response heterogeneity stemming from psychological characteristics in self-reported health on the EQ-5D-5L. TP seems to play a double role in self-reported health, firstly as affecting underlying health and secondly as a factor influencing one's response behavior. These findings increase our understanding of the non-health-related factors that affect self-reported health on standardized health status measures.

Measurement invariance and adapted preferences: evidence for the ICECAP-A and WeRFree instruments

BACKGROUND

Self-report instruments are used to evaluate the effect of interventions. However, individuals adapt to adversity. This could result in individuals reporting higher levels of well-being than one would expect. It is possible to test for the influence of adapted preferences on instrument responses using measurement invariance testing. This study conducts such a test with the Wellbeing Related option-Freedom (WeRFree) and ICECAP-A instruments.

METHODS

A multi-group confirmatory factor analysis was conducted to iteratively test four increasingly stringent types of measurement invariance: (1) configural invariance, (2) metric invariance, (3) scalar invariance, and (4) residual invariance. Data from the Multi Instrument Comparison study were divided into subsamples that reflect groups of participants that differ by age, gender, education, or health condition. Measurement invariance was assessed with (changes in) the Comparative Fit Index (CFI), Root Mean Square Error of Approximation (RMSEA), and Root Mean Square Residual (SRMR) fit indices.

RESULTS

For the WeRFree instrument, full measurement invariance could be established in the gender and education subsamples. Scalar invariance, but not residual invariance, was established in the health condition and age group subsamples. For the ICECAP-A, full measurement invariance could be established in the gender, education, and age group subsamples. Scalar invariance could be established in the health group subsample.

CONCLUSIONS

This study tests the measurement invariance properties of the WeRFree and ICECAP-A instruments. The results indicate that these instruments were scalar invariant in all subsamples, which means that group means can be compared across different subpopulations. We suggest that measurement invariance of capability instruments should routinely be tested with a reference group that does not experience a disadvantage to study whether responses could be affected by adapted preferences.

Sex-related differential item functioning of neck disability index

PURPOSE

To investigate if the responses to the Neck Disability Index (NDI) may produce some differential item functioning (DIF) comparing men and women.

MATERIALS AND METHODS

Register-based study among patients undergoing cervical surgery. Item response theory (IRT) analysis including a model for detecting a DIF.

RESULTS

Of 338 patients, 171 (51%) were women and 167 (49%) were men. The mean age was 54.0 years. For most of the items, the average level of disability in a studied sample was associated with the middle point of the scale. The ability to distinguish people with different levels of disability was high or perfect for seven out of 10 items. While the DIF could be seen for all 10 items, only three items demonstrated statistically significant DIF - "pain intensity", "headaches" and "recreation". While the other seven items did not show statistically significant DIFs, better discrimination (steeper curves) for women could be graphically observed for "personal care", "lifting", "work", "driving" and "sleeping".

CONCLUSIONS

It seemed that the NDI may behave differently depending on the sex of respondents. Several items of the NDI may be more precise and more sensitive when detecting restrictions in functioning among women compared to men. This finding should be taken into account when using the NDI in research and clinical practice.Implications for RehabilitationWhile the Neck Disability Index have been found to be a reliable and valid scale, potential differences in its properties across different sexes have mostly remained uninvestigated.This study showed that the Neck Disability Index may behave differently depending on the sex of respondents.Several items of the Neck Disability Index were more precise and sensitive when detecting restrictions in functioning among women compared to men.This difference should be taken into account when using the NDI in research and clinical practice.