Validation and Comparison of the Informant-Rated and Self-Rated Versions of the Mild Behavioral Impairment Checklist

Abstract

Background: The Mild Behavioral Impairment Checklist (MBI-C) has been developed to assess mild behavioral impairment (MBI). However, no study has validated the use of MBI-C using a promising translation method in Taiwan. Thus, consistency and discrepancy between informant-rated and self-rated scores have not been extensively researched. Objective: This study validated and compared the informant- and self-rated versions of the MBI-C among community-dwelling people in Taiwan. Method: We recruited 202 pairs of individuals without dementia aged ≥50 years and their cohabitating informants. The participants completed the MBI-C (MBI-C-self), and the informants completed the MBI-C (MBI-C-informant) and the Neuropsychiatric Inventory Questionnaire (NPI-Q) independently. Internal consistency, inter-rater reliability, and convergent validity were examined. Results: Both MBI-C-self and MBI-C-informant exhibited satisfactory Cronbach’s α values (0.92 and 0.88, respectively). The MBI-C-informant total scorewas correlated with the NPI-Q total score (r = 0.83, p < 0.001). Inter-rater reliability between the two versions, as represented by the inter-rater correlation coefficient, was 0.57 (p < 0.001). The prevalence of MBI based on the MBI-C-informant scores was 1.5% higher than that based on the MBI-C-self scores according to the suggested cut-off score of 8.5. The affective dysregulation domain score of MBI-C-informant was significantly lower than that of MBI-C-self. Conclusion: MBI-C-informant exhibited both high reliability and validity. Discrepancies between MBI-C-informant and MBI-C-self related to the detection rates and affective dysregulation domain scores were noted. The level of consistency and discrepancy between these two versions provide implications for the use of MBI-C in clinical practice and future research.