Skip to content
DraconiticResearch
Search Sources API
197,787 studies31 sourcesUpdated 1 h ago
Peer-reviewedCC BY-NC 4.0Alzheimer's & Dementia : Diagnosis, Assessment & Disease Monitoring

Characterizing cognitive concerns in subjective cognitive decline using large language models and electronic health record notes

  • Taiyang Chen1
  • Justin Wang1,2
  • Sheril Varghese1,3
  • Jiazi Tian4
  • Rebecca E. Amariglio5,6
  • Li Zhou1,6
  • Gad A. Marshall5,6
  • Liqin Wang1,6
6 affiliations
  1. 1Division of General Internal Medicine and Primary Care Department of Medicine Brigham and Women's Hospital Boston Massachusetts USA
  2. 2Lexington High School Lexington Massachusetts USA
  3. 3Columbia University Irving Medical Center New York New York USA
  4. 4Clinical Augmented Intelligence Group Department of Medicine Massachusetts General Hospital Boston Massachusetts USA
  5. 5Department of Neurology Brigham and Women's Hospital Massachusetts General Hospital Boston Massachusetts USA
  6. 6Harvard Medical School Boston Massachusetts USA
Read full textPDFDOICite

Abstract

INTRODUCTION: Subjective cognitive decline (SCD) refers to self‐ or informant‐reported cognitive concerns despite normal objective testing. These concerns are often documented as free text in electronic health records (EHRs), limiting systematic analysis. We characterized cognitive concern domains and assessed differences across demographic subgroups.

METHODS: We studied 2175 individuals aged ≥65 years with an EHR‐derived SCD phenotype. Using large language models with expert review, we derived a taxonomy of concern themes and sub‐themes and assigned themes to note excerpts. We examined theme frequencies and compared distribution across demographic groups.

RESULTS: We identified 13 themes and 40 sub‐themes. Theme‐level classification showed strong performance (micro‐F1 = 84.8%). Memory Issues was the most common theme (68.5%), with concerns spanning multiple cognitive domains. Theme distribution varied by age, sex, marital status, and education.

DISCUSSION: EHR notes captured heterogeneous, multi‐domain cognitive concerns among adults with an EHR‐derived SCD phenotype. This approach enables scalable, clinically interpretable profiling of concerns that may support early risk stratification.

Related

  1. A Multidisciplinary Neuro-Lupus Program for the Evaluation and Management of Patients With SLE and Cognitive Concerns
    Douglas R. Wilcox, Erin E. Mathieu, Emma L. Weizenbaum +3
    Neurology: Clinical PracticePeer-reviewedLicence not statedFull textGeneral medicinePMC
  2. Characterizing Cancer Symptom Experiences and Related Psychosocial Concerns Through Chinese Social Media Narratives: Computational Text Analysis.
    Huan Yu, Yin Wang, Jiaqi Ding +4
    Journal of medical Internet researchPeer-reviewedCC BYFull textOncologyEurope PMC
  3. Effects of cognitive-motor dual-task walking exercise versus walking-only exercise on cognitive function and patient-centred outcomes in older adults with subjective brain fog: a randomised controlled trial.
    Emre Dansuk, Burak Menek
    BMJ openPeer-reviewedCC BYFull textPsychiatry & mental healthEurope PMCCrossref
  4. Age at Menopause and Subjective Cognitive Symptoms Predict Digital Cognitive Outcomes at the Gynecological Well-Woman Visit
    Teresa Facchetti, Jillian L. Joyce, Sandra Rizer +5
    Journal of clinical and experimental neuropsychologyPeer-reviewedLicence not statedFull textEndocrinology & metabolismPMC
  5. Episodic memory-associated weighted atrophy network mapping in mild cognitive impairment and subjective cognitive decline: a multi-center study
    Anxing Huang, Keke Jiang, Lei Chi +10
    Translational PsychiatryPeer-reviewedCC BY 4.0Psychiatry & mental healthCrossref

Cite

BibTeXRIS
BibTeX
@article{chen2026characterizing,
  title = {{Characterizing cognitive concerns in subjective cognitive decline using large language models and electronic health record notes}},
  author = {Taiyang Chen and Justin Wang and Sheril Varghese and Jiazi Tian and Rebecca E. Amariglio and Li Zhou and Gad A. Marshall and Liqin Wang},
  journal = {Alzheimer's \& Dementia : Diagnosis, Assessment \& Disease Monitoring},
  year = {2026},
  month = oct,
  doi = {10.1002/dad2.70525},
  url = {https://doi.org/10.1002/dad2.70525},
  note = {CC BY-NC 4.0},
}
RIS
TY  - JOUR
TI  - Characterizing cognitive concerns in subjective cognitive decline using large language models and electronic health record notes
AU  - Taiyang Chen
AU  - Justin Wang
AU  - Sheril Varghese
AU  - Jiazi Tian
AU  - Rebecca E. Amariglio
AU  - Li Zhou
AU  - Gad A. Marshall
AU  - Liqin Wang
PY  - 2026
DA  - 2026/10/01
JO  - Alzheimer's & Dementia : Diagnosis, Assessment & Disease Monitoring
DO  - 10.1002/dad2.70525
UR  - https://doi.org/10.1002/dad2.70525
AB  - INTRODUCTION: Subjective cognitive decline (SCD) refers to self‐ or informant‐reported cognitive concerns despite normal objective testing. These concerns are often documented as free text in electronic health records (EHRs), limiting systematic analysis. We characterized cognitive concern domains and assessed differences across demographic subgroups.  METHODS: We studied 2175 individuals aged ≥65 years with an EHR‐derived SCD phenotype. Using large language models with expert review, we derived a taxonomy of concern themes and sub‐themes and assigned themes to note excerpts. We examined theme frequencies and compared distribution across demographic groups.  RESULTS: We identified 13 themes and 40 sub‐themes. Theme‐level classification showed strong performance (micro‐F1 = 84.8%). Memory Issues was the most common theme (68.5%), with concerns spanning multiple cognitive domains. Theme distribution varied by age, sex, marital status, and education.  DISCUSSION: EHR notes captured heterogeneous, multi‐domain cognitive concerns among adults with an EHR‐derived SCD phenotype. This approach enables scalable, clinically interpretable profiling of concerns that may support early risk stratification.
SN  - 2352-8729
ER  -