Reference-Group Adjusted Behavioural Dysfunction Questionnaire Score Discriminates Highly Behavioural-Variant...

Area of Science:

• Neurology
• Neuropsychiatry
• Diagnostic Tools

Background:

• Behavioural variant frontotemporal dementia (bvFTD) is characterized by behavioral disorders, necessitating accurate assessment for early diagnosis.
• The Behavioural Dysfunction Questionnaire (BDQ) shows promise in identifying bvFTD but requires enhanced discriminatory power.
• Adapting BDQ scoring based on reference groups can potentially improve its diagnostic utility.

Purpose of the Study:

• To enhance the diagnostic accuracy of the BDQ for mild-stage bvFTD.
• To evaluate the effectiveness of adjusting BDQ scoring based on reference groups (MDD, ADD) using machine learning.

Main Methods:

• A combined prospective and retrospective cross-sectional study analyzed data from 241 patients (50 bvFTD, 71 MDD, 120 ADD).
• Two scoring methods were used: standard average score and machine learning-adjusted scores.
• Fivefold cross-validation was employed to validate the adjusted scoring method.

Main Results:

• The adjusted BDQ score demonstrated superior discrimination between bvFTD and MDD compared to the unadjusted score.
• The adjusted BDQ score showed similar discriminatory power between bvFTD and ADD as the unadjusted score.
• Adjusted scoring resulted in a smaller difference between cut-offs for achieving at least 90% sensitivity and specificity.

Conclusions:

• Adjusted BDQ scores are recommended when differentiating bvFTD from MDD or ADD.
• This adaptive scoring approach improves diagnostic precision in clinical practice.
• Further research should explore adjusted scoring for other differential diagnoses of bvFTD.