Testing the psychometric properties of 9 bolt-ons for the EQ-5D-5L in a general population sample

Abstract

Objectives We aimed to assess the psychometric performance and added value of 9 existing boltons (breathing problems, cognition, hearing, self-confidence, skin irritation, sleep, social relationships, tiredness, andvision) for the EQ-5D-5L in a general population sample. Methods The EQ-5D-5L, 9 bolt-ons, SF-6Dv1, Patient-Reported Outcomes Measurement Information System(PROMIS)-2912, PROMIS Global Health, and Satisfaction with Life Scale were completed in an online cross-sectional survey among a general adult population sample in Hungary (n=1587). The following psychometric properties were tested for the EQ-5D-5L1 bolt-on(s): ceiling, divergent and convergent validity, structural validity, known-groupvalidity, and explanatory power. Results Adding sleep (30%),tiredness (24%), or vision (21%) substantially reduced the ceiling of the EQ-5D-5L (41%). Cognition, sleep, social relationships, and tiredness correlated with corresponding PROMIS and SF-6D items (rs=ǀ0.32ǀ2ǀ0.73ǀ). All bolt-ons, except cognition and self-confidence, loaded on a different factor from the EQ-5D-5L dimensions. Breathing, hearing, skin irritation, and vision significantly improved known-group validity in relevant health conditiongroups. The sleep bolt-on improved known-group validity in 9 of 13 chronic health conditions. Tiredness had the largest impact on explaining EQ VAS score variance in 8 of 13 conditions. Hearing and vision improved the ability of the EQ-5D-5L to capture declining health with age, whereas selfconfidence and social relationships were valuable for mental health assessment. Conclusion This study established the validity of multiple bolt-ons for the EQ-5D-5L and highlights the usefulness of including relevant bolt-ons in population-based and patient surveys. Our findings inform the further development of these bolt-ons and the bolt-on item selection for clinical studies.