Supplementary Material for: The ?Protocole d��Examen Cognitif de la Personne Ag��e �C Lausanne? (PECPA-L): A Cognitive Assessment Tool for the French-Speaking Elderly in Switzerland

Objective: The main objective of this study was to validate and normalize, in the French-speaking population of Switzerland, an empirically derived structured cognitive assessment scale for the elderly. Method: 237 healthy elderly and 115 elderly with mild-to-moderate dementia as well as a preliminary sample of 27 subjects with MCI, all community-dwelling, were assessed using the PECPA-L for its validation and normalization using Crohnbach��s alpha, rank comparisons, and AUROC measures. Results: The internal consistency of the PECPA-L was good (Crohnbach��s Alpha = 0.79). The discriminating power of the PECPA-L between the normal elderly and the mildly to moderately demented (AUROC = 0.940) and those with MCI (AUROC = 0.925) was high. Normal aging does not influence all cognitive domains equally; therefore, the subscores of the PECPA-L vary in their discriminating power between the normal and the demented elderly, with gnosis abilities having the least (AUROC = 0.719) and delayed memory performance having the highest (AUROC = 0.927) discriminating power. Normative data according to age and education are provided. Conclusions: The PECPA-L is a highly appropriate tool for the detection and documentation of early cognitive impairment in the French-speaking population in Switzerland while accounting for age and education.