Computerized decision support to optimally funnel patients through the diagnostic pathway for dementia

Area of Science:

• Neurology
• Biomarkers
• Diagnostic Tools

Background:

• Rising dementia prevalence necessitates efficient diagnostic strategies.
• Emerging disease-modifying therapies (DMTs) require precise patient selection.
• Memory clinics need optimized pathways for syndrome, etiological, and DMT eligibility diagnosis.

Purpose of the Study:

• To develop and validate a data-driven, stepwise approach for dementia diagnosis.
• To assess the efficiency and accuracy of this approach across three clinical scenarios.
• To guide clinicians in selecting tests with added value in memory clinics.

Main Methods:

• Utilized data from two memory clinic cohorts (ADC, PredictND) comprising 504 dementia patients, 191 MCI, and 188 controls.
• Employed digital cognitive screening (cCOG), neuropsychological assessment (NP), MRI, and CSF biomarkers in a sequential testing strategy.
• A clinical decision support system (CDSS) generated a disease state index (DSI) to guide diagnostic certainty and confirm diagnoses.

Main Results:

• cCOG prescreening reduced NP needs by 42% for syndrome diagnosis (accuracy 0.71).
• Stepwise etiological diagnosis achieved 80% accuracy, requiring MRI in 77% and CSF in 37% of cases.
• For DMT eligibility, stepwise testing identified 90% of potentially eligible AD dementia patients, with 60% proceeding to CSF testing.

Conclusions:

• Data-driven diagnostic pathways are accurate and efficient across different clinical scenarios.
• A CDSS tool can assist clinicians in optimizing test selection for added value.
• Efficient diagnostic pathways are crucial for maintaining accessibility and affordability of dementia care, especially with new DMTs.