Objective:

To evaluate the efficacy of the Digital Clock and Recall (DCR™) as a rapid digital cognitive assessment to predict Mini-Mental State Examination (MMSE) scores, facilitating patient triage for anti-amyloid disease-modifying therapies (DMTs) and addressing potential biases in traditional assessments.

Approach:

• Study Design: Retrospective analysis using data from the multi-site Bio-Hermes-001 study (NCT04733989, N = 945) to train a Poisson elastic net regression model predicting MMSE scores based on age and multimodal digital features derived from the DCR.

Key Findings:

• The machine learning model predicted MMSE scores with a root-mean-squared error (RMSE) of 2.43 in the Bio-Hermes-001 test set, which is within the established test-retest reliability range of the manual MMSE.
• External validation in the Apheleia cohort showed RMSE of 2.62, indicating robust generalizability.
• The model demonstrated comparable performance across different racial and ethnic groups, suggesting fair predictions.

Interpretation:

The DCR can accurately and equitably predict MMSE scores, potentially transforming the assessment process for DMT eligibility and addressing biases present in traditional methods.

Limitations:

• The study relies on retrospective data, which may limit the generalizability of findings and introduce biases related to the original data collection methods.

Conclusion:

The DCR offers a potential alternative to traditional MMSE assessments, enabling faster identification of patients eligible for DMTs.