Initial specialist validation of clinical decision support recommendations from a machine learning-enabled digital cognitive assessment - Summary - MDSpire
Advertisement
Initial specialist validation of clinical decision support recommendations from a machine learning-enabled digital cognitive assessment
To evaluate the clinical appropriateness of recommendations and decision pathways generated by the CDS functionality embedded in the Linus Health Core Cognitive Evaluation (CCE) for improving patient care in cognitive impairment.
Approach:
Key Findings:
All cognitive-impairment recommendations met the appropriateness threshold.
All seven borderline/impaired-DCR pathways were deemed appropriate (median 7–8).
Two pathways fell below the threshold: cognitively unimpaired individuals with Green DCR scores (median 6) and a preliminary anti-amyloid treatment referral pathway (median 5).
Moderate agreement among experts was observed (median ICC(2,k) = 0.61), with lower agreement for individual diagnostic-concern recommendations (median ICC(2,k) = 0.25), indicating variability in expert opinions.
Interpretation:
Cognitive neurologists judged CCE-derived CDS as appropriate for PCP workup and referral decisions in older adults with suspected cognitive impairment, highlighting the need for effective tools in primary care.
Limitations:
The study reflects specialist heterogeneity on borderline non-cognitive items, which may affect the reliability of the findings.
Ceiling effects on high-rated items may have influenced agreement, potentially limiting the interpretation of expert consensus.
Conclusion:
Findings support the initial content validity of assessment-linked CDS and identify refinement priorities in low-risk and emerging-therapy pathways, emphasizing the need for further studies to enhance practical application.