Predictors of hospitalization and institutionalization in Medicaid patient populations with Alzheimer’s Disease

Objectives: Determine predictors of hospitalization and institutionalization in Medicaid populations with Alzheimer’s Disease (AD). Methods: Data were obtained from the Centers for Medicareand Medicaid Services (CMS). Individuals enrolled in Florida,New Jersey, and New York Medicaid programs on January 1, 2004, remained in that program for 1 year and exposed to an AD medication were included. AD diagnosis was based on the ICD-9-CM code 331.0. Outcomes of interest were hospitalization and institutionalization. Multivariate logistic regression models were used to test for the association between outcomes of interest and demographics, resource utilization factors, and type of AD pharmacotherapy exposure. Results: A total of 65,442 individuals qualified for the study. Age was positively and significantly associated with hospitalization (p < 0.001) and institutionalization (p < 0.001). Exposure tomemantine was significantly associated with less chance of hospitalization (OR = 0.88; 99% CI: 0.77 - 0.99). Non-Hispanic Blacks were significantly (p < 0.001) more likely to be hospitalized than Non-Hispanic Whites (OR = 1.60; 99% CI: 1.41 - 1.81). Compared to Non-Hispanic Whites, Non-Hispanic Blacks (OR = 0.73; 99% CI: 0.60 - 0.88), Hispanics (OR = 0.36; 99% CI: 0.27 - 0.47), and Non-Hispanic Others (OR = 0.42; 99% CI: 0.21 - 0.82) were significantly less likely to be institutionalized. Individuals prescribed donepezil (OR = 1.29, 99% CI, 1.08 - 1.54) and galantamine (OR = 1.46, 99% CI: 1.19 - 1.79) were significantly more likely to have an institutionalization claim (p < 0.001 for both medications). Residents of New York were significantly morelikely to be hospitalized than Florida residents (OR = 1.30; 99% CI: 1.17 - 1.44), where as New Jersey residents were significantly less likely to be hospitalized (OR = 0.75; 99% CI: 0.66 - 0.85). Finally, compared toFloridaresidents, residents of New Jerseywere significantly more likely to be institutionalized (OR = 4.61; 99% CI: 3.98 - 5.33). Conclusion: Demographics, state of residence and pharmacotherapy exposure weresignificant predictors of health care service utilization. Further pharmacoeconomic studies in AD medication therapy are warranted.