Healthcare Resource Utilization Among Veterans with Alzheimer’s Disease

Abstract Background We evaluated healthcare utilization patterns to understand the care pathways for Veterans with Alzheimer’s disease (AD) from the Department of Veterans Affairs (VA) healthcare system. Method VA electronic health records (EHR) were used to extract data from clinical notes and identify patients with AD. The patient cohort was merged with the VA Informatics and Computing Infrastructure (VINCI) administrative database from October 2008–to October 2021 to obtain systematic healthcare utilization records including total inpatient admissions and outpatient visits. A control group was extracted for comparison and matched to the patient cohort using the propensity score on age, sex, and race. A linear mixed‐effects model was used to compare the total outpatient and inpatient utilization, adjusting for patient demographics and comorbidities. Result We extracted 21,442 AD patients from the EHR clinical notes. A cohort of 85,684 non‐AD and non‐dementia Veterans was constructed as a control group. In the AD vs control groups, the mean age was 82 vs 79 years, 83 vs 84% was White, and 97% were male in both groups. The number of annual outpatient visits 1 year before, and at 1 and 2 years after the first AD diagnosis code was 30, 34 and 27, respectively. The control group maintained 12 annual outpatient visits throughout the three‐year period. The average annual hospital admission rate was 1.8 vs 1.7 in the AD vs control group. The average length of inpatient stay in the year prior to the first AD diagnostic code was 23.1 days, increased to 27.4 days at year 1 after AD diagnosis, and declined back to 20.5 days at year 2, as compared to 13.1, 13.0, and 11.9 days, respectively, in the control group. The mixed‐effects analysis showed significantly higher healthcare utilization trends across VA outpatient and inpatient care settings (p<0.001) in the AD vs control groups, after controlling for age, gender, race, ethnicity, and select comorbidities. Conclusion Healthcare resource utilization in patients with AD identified from the medical notes is more than doubled that in outpatient care and nearly doubled in inpatient length of stay relative to the matched controls in the VA healthcare system.