Reliability and validity of a short form of the Korean dementia screening questionnaire-cognition

Background and Purpose We aimed to determine the reliability and validity of a short form of the Korean Dementia Screening Questionnaire-Cognition (KDSQ-C) as a screening tool for cognitive dysfunction. Methods This study recruited 420 patients older than 65 years and their informants from 11 hospitals, and categorized the patients into normal cognition, mild cognitive impairment, and dementia subgroups. The KDSQ-C was completed separately by the patients and their informants. We abstracted three components of the KDSQ-C and combined these components into the following four subscales: KDSQ-C-I (items 1–5, memory domain), KDSQ-C-II (items 1–5 & 11–15, memory domain+activities of daily living), KDSQ-C-III (items 1–5 & 6–10, memory domain+other cognitive domains), and KDSQ-C-IV (items 6–10 & 11–15, other cognitive domains+activities of daily living). The reliability and validity were compared between these four subscales. Results A receiver operating characteristic (ROC) analysis of questionnaire scores provided by the patients showed that the areas under the ROC curves (AUCs) for the KDSQ-C, KDSQ-C-I, and KDSQ-C-II for diagnosing dementia were 0.75, 0.72, and 0.76, respectively; the corresponding AUCs for informant-completed questionnaires were 0.92, 0.89, and 0.92, indicating good discriminability for dementia. Conclusions A short form of the patient-and informant-rated versions of the KDSQ-C (KD-SQ-C-II) is as capable as the 15-item KDSQ-C in screening for dementia.