Linking data to determine risk for 30-day readmissions in dementia

The demand and the landscape of options for dementia care are growing. Standardization of care for persons with Alzheimer disease and related dementias (ADRD) lacks infrastructure across episodes of care. Use of electronic health records (EHRs) in practice settings yields valuable information that can enhance continuity of patient care. The objective of this study was to use EHR-derived variables to identify risk factors for 30-day readmissions in the ADRD population across episodes of care.

Cross-sectional, retrospective study of older adults (aged ≥ 65 years) with ADRD discharged from a large urban academic medical center between October 1, 2018, and March 31, 2022.

Data extracted across episodes of care from the EHR included demographic characteristics, medical variables, and encounter variables.

A total of 14,101 patients diagnosed with ADRD were included in the study. Factors associated with patients being more likely to experience 30-day hospital readmissions included advanced age, male sex, being a non-English speaker, having more severe comorbidities, staying in the hospital for more than 5 days, having had more than 1 surgical procedure in the prior 6 months, having had 3 or more inpatient admissions in the 6 months prior to index admission, having had more than 3 physician consultations in the prior 6 months, and having been discharged to settings other than home (all P < .05).

By utilizing the EHR to connect medical and encounter data across episodes of care, health care providers and administrators can gain valuable insight into identifying factors contributing to readmissions, which could be used to improve continuity of care for patients and caregivers, ultimately leading to better outcomes and reduced health care costs.