Managing Scientist Exponent, Inc. Natick, United States
Background: Electronic Health Record (EHR) data collected in near real-time will improve the timeliness, frequency, and completeness of clinical information for the long-term care population.
Objectives: To characterize the current data compiled by the Long-Term Care (LTC) Data Cooperative, a provider-run effort collecting EHR data from multiple nursing homes across the United States.
Methods: We described data from the 793 US-based nursing homes located across 28 states that joined the LTC Data Cooperative before March 31, 2022. This study population included residents admitted to facilities on or after January 1, 2019, representing 293,605 short- and long-stay individuals and 386,052 admissions. The EHR data comprised an array of values that were associated with relevant timestamps, recorded on a daily basis, or otherwise noted frequently.
Results: The population was 59% female, 86% aged ≥ 65 years, 82% White, and 15% Black or African American. Residents diagnosed with Alzheimer’s Disease or Related Dementia comprised 31% of admissions. There were 220,300,150 medication administrations, 44% of which were for residents with dementia. Vaccination records were present for 69% of residents in the dataset (n=202,060). COVID-19 vaccine refusals were documented for 6% of residents (n= 18,445). Compared to the standard MDS, assessments were recorded more frequently in the EHR: 3.6 activities of daily living assessments per quarter (n=553 facilities), 4.0 Brief Interview for Mental Status assessments per quarter (n=356 facilities), and 3.7 Patient Health Questionnaire-9 assessments per quarter (n=329 facilities). Temperature readings were documented a median of 2.0 observations per day for short stays and a median of 1.3 observations per day for long stays. Most facilities (755 of 793 facilities) documented bedside blood glucose readings (median (short stay): 2.7 times/day; median (long stay): 1.0 times/day).
Conclusions: EHR data from long-term care facilities represent a rich and comprehensive longitudinal data source for researchers to conduct effectiveness and safety research and offer an efficient and effective method for data collection for interventional research in long-term care.
