Healthcare Utilization Among Veterans with Alzheimer's Disease Identified from Clinical Notes

OBJECTIVES: To assess healthcare utilization among Veterans with Alzheimer's disease (AD) in the Department of Veterans Affairs (VA) healthcare system.

METHODS: Veterans with AD were identified from clinical notes in the VA electronic health record. A control group was propensity score matched on age, sex, and race. Healthcare utilization was assessed one year before and two years after AD index date. A linear mixed-effects model was used to compare hospitalizations and outpatient visits, adjusting for patient demographics and select comorbidities.

RESULTS: The AD and control group included 21,442 and 85,684 Veterans, respectively. Annual outpatient visits were roughly 2.3-2.8 times greater in the AD group than the control group during the 3-year study period. Annual hospitalization rates for AD and control groups were comparable. However, the average length of stay for AD patients was 1.8-2.1 times longer than control. The mixed-effects model showed significantly increased hospitalizations and outpatient visits (p<0.001) in the AD group compared to control, after controlling for age, sex, race, ethnicity, and select comorbidities.

CONCLUSIONS: Healthcare utilization of AD patients identified via clinical notes is higher than the matched control group. Identification of AD patients via clinical notes is a viable method for patient identification. Incorporation of clinical notes with diagnostic codes such as International Classification of Diseases may address issues related to undercoding and produce more comprehensive patient registries.