German questionnaires assessing quality of life and psycho-social status in people with epilepsy: Reliable change and intercorrelations.

ABSTRACT

OBJECTIVES: People with epilepsy (PWE) not only suffer from seizures but also from various psycho-social issues containing facets such as social functioning, anxiety, depression or stigmatization, and consequently quality of life. (1) Assessing reliable change of these issues is crucial to evaluate their course and potential treatment effects. As most psycho-social self-report questionnaires have been validated in separate samples, their clinical-socio-demographic differences may limit the comparability and generalizability of the scales' internal consistency, which is important for the reliable change index (RCI). Using a co-normalized approach, we provide the internal consistency and RCIs for a large set of questionnaires targeting quality of life (QOLIE-31-P), depressive symptoms (NDDI-E), anxiety (GAD-7), seizure severity (LSSS), subjective antiseizure medication adverse events (LAEP), stigma, epilepsy-related fear, and restrictions in daily life (PESOS), and subjective cognition (FLei). As for some German versions of these measures, psychometric data is still missing, we also add important information for the German language area. (2) In addition, knowledge about intercorrelations of these constructs is needed to shape questionnaire usage and treatment approaches. We thus investigate associations of these scales and compare weighted and unweighted subscales of the QOLIE-31-P. METHODS: In our prospective study, 202 adult in-patients of the Epilepsy-Center Berlin-Brandenburg with a reliable diagnosis of epilepsy filled out a set of self-report questionnaires between 03/2018 and 03/2021. We calculated Cronbach's α, RCIs, and bivariate intercorrelations and compared the respective correlations of weighted and unweighted scales of the QOLIE-31-P. RESULTS: For most of the scales, good to excellent internal consistency was identified. Furthermore, we found intercorrelations in the expected directions with strong links between scales assessing similar constructs (e.g., QOLIE-31-P Cognition and FLei), but weak relationships between measures for different constructs (e.g., QOLIE-31-P Seizure worry and FLei). The QOLIE-31-P Total score was highly correlated with most of the other scales. Some differences regarding their correlational patterns for weighted and unweighted QOLIE-31-P scales were identified. CONCLUSIONS: Psycho-social constructs share a large amount of common variance, but still can be separated from each other. The QOLIE-31-P Total score represents an adequate measure of general psycho-social burden.