John Libbey Eurotext

Psychometric validation of the French version of the quality of life in epilepsy inventory (QOLIE-31): Comparison with a generic health-related quality of life questionnaire Volume 6, issue 4, December 2004

Unit of Clinical Research and Epidemiology, University Hospital of Montpellier, Department of Epileptology, University Hospital of Montpellier, Biostatistics and Epidemiology, Unit 2415, University Montpellier I, Direction économie de la santé, Laboratoire GlaxoSmithKline, Marly-le-Roi, France

A psychometric evaluation of a French transcultural version of the quality of life in epilepsy inventory-31 (QOLIE-31) was carried out. QOLIE-31 was compared to a generic health-related quality of life questionnaire, the Nottingham health profile (NHP). The psychometric properties of QOLIE-31, assessed in 190 adults with epilepsy, included: acceptability, test-retest reliability and validity (multi-trait analysis including internal consistency and item-to-scale correlations, construct validity using factor analysis, discriminative validity using relationship with disease characteristics, treatment effects, divergent and convergent validity using correlations with NHP scores). Both acceptability and reproducibility were good and internal consistency was high (Cronbach’s α coefficient = 0.86). Factor analysis with varimax rotation identified seven factors with eigenvalues > 1, with two factors, related to cognitive function and mood, accounting for 46.5% of the variance. However, goodness of fit indices revealed that a model with four factors best fitted the data. The first factor corresponds to a generic mental dimension, the second is equivalent to the cognitive functioning dimension, the third to medication effects including social functioning, and the fourth to seizure worry. Discriminative validity was good for seizure control and treatment tolerability. Hight correlations between QOLIE-31 and pertinent NHP scales (emotional reactions, energy and social isolation) were observed. The French version of QOLIE-31 thus meets established psychometric criteria for reliability and validity.