Assessment of the psychometric properties of the EQ-5D-3L and EQ-5D-5L instruments in psoriasis

The aim of this study was to assess the psychometric properties of the five-level (EQ-5D-5L) in comparison with the standard three-level (EQ-5D-3L) classification systems of the EQ-5D in a sample of psoriatic patients. Psoriatic subjects (n = 396) recruited from 16 private practicing centers from various areas of Greece self-completed the two EQ-5D versions and the Dermatology Life Quality Index, while information was also collected on socio-demographics, clinical characteristics and comorbidities. The EQ-5D-5L and EQ-5D-3L were evaluated in terms of agreement, feasibility, ceiling effects, redistribution properties, inconsistency, informativity, and convergent and known-groups validity. Missing values were negligible in both versions. The agreement between the EQ-5D-5L and the EQ-5D-3L was very high (ICC = 0.94), while the largest differences were identified for subjects with moderate health status. Ceiling effects decreased in the EQ-5D-5L system by 14.08% (p < 0.05), with "anxiety/depression" showing the highest relative reduction (-10.31%; p < 0.05). Overall inconsistency was rather low (1.7%) and respondents preferred to report milder problems in the EQ-5D-5L. Absolute informativity improved by 56.42% in the EQ-5D-5L, while relative informativity declined by 9.24%, with only "anxiety/depression" demonstrating a small increase (6.77%). Both instruments demonstrated good convergent and known-groups validity, with evidence of a slightly better convergent performance and discriminatory efficiency of the EQ-5D-5L. In conclusion, both instruments demonstrated consistency, valid redistribution and good construct validity. The EQ-5D-5L system may be preferable to the EQ-5D-3L in psoriatic patients, as it demonstrated a marginally better performance in terms of reduced ceiling effects, increased informativity, and improved convergent and known-groups validity efficiency, particularly in the domain of "anxiety/depression".