A new scoring system for the Similarity Task of the Frontal Assessment Battery (FAB) to assess the severity of impairment in conceptualization: development and validity study

<p>Background: The Similarity Task of the Frontal Assessment Battery (FAB) is a typical task for assessing conceptualization. Our preceding study suggested that the degree of impairment in conceptualization may be different between the patients who correctly expressed common attributes of word pairs and those who could not express them in the task. Objective: To develop an extended scoring system including evaluation of whether a subject can express common attributes for 3 pairs of words in the Similarity Task of the FAB. Subjects: A total of 69 dementia patients who underwent the Mini-mental State Examination (MMSE), Alzheimer's Disease Assessment Scale (ADAS), and FAB. Methods: In each patient, we classified responses to the 3 items in the Similarity Task of the FAB into the following categories: a) correct response, b) incorrect response expressing a common attribute of the word pair, or c) other incorrect responses, and made 3 patients groups based on this classification for each word pair. The MMSE scores and the mean scores of the 5 tasks in the FAB other than the Similarity (FAB15) were compared among the 3 groups using analysis of variance. We determined a new point allocation of the Similarity Task (the extended scoring system) based on the results, and evaluated correlations between the total score of the extended scoring system and other measures. Results: We allocated 3 points to a correct response, 1 point to an incorrect response expressing a common attribute of the word pair, and 0 point to other incorrect responses as a weighting for the extended scoring system. The allocation gave a score of 0 to 9 in a total of 3 items. The total score of the extended scoring system had a significant correlation coefficient (r > 0.5) with the score of FAB15 and with the score of the Ideational Praxis in the ADAS. In multiple correlation analysis with the total score of the extended scoring system, the years of education, the MMSE score, and the score of the Ideational Praxis in the ADAS, the total score of the extended scoring system had a significant partial correlation coefficient with the score of the Ideational Praxis. Conclusion: We developed an extended scoring system in consideration of the quality of incorrect responses in the Similarity Task and showed the construct validity of the total score.</p>