Development of informant-reported or self-reported MCI-J questionnaire

Several studies have reported a high prevalence of missed and delayed mild cognitive impairment (MCI) or mild dementia diagnosis, which could lead to delayed treatment and increased patient and caregiver burden. Objectives: This study aimed to develop a new questionnaire for nonprofessionals to help detect early signs of MCI and dementia. Respondents included patients, family caregivers, or health professionals. Scores are calculated based on the respondent type and age of subject. Methods: This study consisted of four steps and included 461 respondents. Steps 1-3 were conducted by a working group, and step 4, by 67 specialist members of the Japanese Society of Geriatric Psychiatry. A scoring algorithm was created and predictive diagnostic probability was analyzed using misdiscrimination rate and cross-validation after item selection to establish a cut-off value for MCI or dementia symptoms. Alzheimer's disease, Lewy body dementia, and frontotemporal dementia were diagnosed. Results: The prediction error rate for patient or informant respondents was confirmed from the evaluation results of 13 items. Sensitivity and specificity were 90.6% and 56.6%, respectively, with a cut-off score of 2. Overall, 82% (61 pairs) of respondents received a definitive diagnosis following a diagnosis from the questionnaire. Conclusions: This questionnaire could promote earlier presentation to clinical settings for treatment. The high sensitivity indicates the utility of this instrument, but it is not meant as a definitive diagnostic tool and should be followed with a professional assessment.