TESTING ERADAR: A NEW EHR-BASED ALGORITHM TO INCREASE DEMENTIA RECOGNITION IN HEALTH CARE SYSTEMS
Bibliographic record
Abstract
Abstract Electronic health records (EHRs) hold potential for identifying individuals with undiagnosed dementia. Using machine learning, our team developed and externally validated a predictive algorithm called the EHR Risk of Alzheimers and Dementia Assessment Rule (eRADAR) that estimates the likelihood an individual has undiagnosed dementia with high accuracy (C statistics of 0.79 to 0.84). Through two embedded pragmatic trials, we are implementing eRADAR in 11 primary care clinics within Kaiser Permanente Washington and University of California, San Francisco. The target population is older adults age 65+ without a documented dementia diagnosis or medication. Primary care providers (PCPs) are randomized to intervention or usual care. The eRADAR algorithm is used to identify individuals with eRADAR scores in the top 15-20%, who are invited to a “brain health” visit that includes assessment of instrumental activities of daily living, depressive symptoms, and cognitive function (Montreal Cognitive Assessment). Results are shared with the patient and PCP, and the PCP is responsible for making the final diagnosis (with decision support provided through the EHR). To date, we have implemented the intervention in 9 clinics and conducted over 590 brain health visits. About 31% (658/2137) of high-risk individuals accept a brain health visit. Of these, about 18% have results suggesting dementia and another 30% mild cognitive impairment. Post-visit surveys show high acceptance of and satisfaction with the intervention. The primary study outcome is rate of dementia diagnosis over 12-month follow-up (completed by April 2025). We are also conducting semi-structured interviews to illuminate benefits and harms.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.002 | 0.007 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".