Stepwise identification of prodromal dementia: Testing a practical model for primary care

Area of Science:

• Neurology
• Gerontology
• Primary Care Medicine

Background:

• Prodromal dementia remains significantly underdiagnosed in primary care settings.
• Early detection is crucial for timely intervention and management of dementia.

Purpose of the Study:

• To develop and validate a clinical model for detecting prodromal dementia within primary care.
• To improve the diagnostic accuracy of dementia screening in a primary care context.

Main Methods:

• A "functional-cognitive" step-down screening model was evaluated using the Functional Activities Questionnaire (FAQ) and Montreal Cognitive Assessment (MoCA).
• Participants were classified into healthy cognition, mild cognitive impairment (MCI) without conversion, and prodromal dementia groups.
• Gradient Boosting Trees were employed for classification and performance analysis.

Main Results:

• The proposed step-down model demonstrated superior performance compared to using the FAQ alone.
• The model showed comparable results to administering both screening measures to all participants.
• Key performance metrics included high sensitivity, specificity, and F1 scores for the step-down model.

Conclusions:

• The developed functional-cognitive model is supported by the study's findings.
• Validated algorithms are provided for primary care use, aiding in early dementia detection.
• The model's accuracy is expected to improve with data from diverse clinical settings, supporting timely pharmacological treatment initiation.