University of Texas Rio Grande Valley
Measuring Sleep Disturbance in Dementia Using EHR Data in Long-Term Care
Real World Data Source
LTC Data Cooperative
Dr. Ye is an assistant professor in the School of Social Work and the Department of Sociology at the University
of Texas Rio Grande Valley. Her research focuses on how social, structural, and environmental contexts shape
health outcomes in vulnerable aging populations, with a particular interest in sleep disturbance and dementia
care. Dr. Ye’s work integrates interdisciplinary approaches from sociology, health services research, and data
science to identify modifiable determinants of health and inform evidence-based interventions for older
adults living in community and long-term care settings.
Sleep disturbance is highly prevalent among nursing home residents living with dementia, yet it is frequently
under-recognized in structured diagnostic fields within electronic health records (EHR). Despite its
associations with agitation, medication use, cognitive decline, and fall risk, sleep disturbance is often
documented in narrative clinical notes rather than standardized data elements. This project addresses a key
measurement gap by integrating structured EHR data with natural language processing (NLP) analysis of
clinical notes to develop an enhanced measure of sleep disturbance in long-term care EHR data. This award
will provide Dr. Ye with the training and experience using the Long-Term Care (LTC) Data Cooperative EHR
and linked Medicare data to: (1) Develop an NLP-enhanced EHR-based measurement tool for sleep
disturbance among nursing home residents with dementia; (2) Validate this measure using medication use,
behavioral symptoms, and claims-based indicators; and (3) Examine how individual, clinical, and facility-level
factors relate to the completeness and structural patterning of sleep disturbance measurement in long-term
care data. This project will generate a scalable and validated approach for identifying sleep disturbance in
nursing home residents with dementia using real-world data. By improving detection of sleep problems in
long-term care settings, this project will enable research on its association with health outcomes and care
processes. This work will lay the foundation for future efforts leveraging real-world data and NLP for research
in aging populations.