Real-World Healthcare Resource Utilization and Costs of Patients with Alzheimer’s Disease By Stage: A Retrospective Observational Study

Speaker(s)

Fazio-Eynullayeva E1, Mystkowski P1, Desai R2, Yee CW2, Liu CL3, Cunnington M3, Mattke S4, Lv L5
1Novo Nordisk Inc., Plainsboro, NJ, USA, 2Analysis Group, Inc, Boston, MA, USA, 3Analysis Group Ltd, London, UK, 4University of Southern California, Center for Economic and Social Research, Los Angeles, CA, USA, 5Novo Nordisk, Inc., Plainsboro Township, NJ, USA

Presentation Documents

OBJECTIVES: This study aimed to compare the Alzheimer’s disease (AD)-related healthcare resource utilization (HCRU) and costs by disease stage and identify the predictors of high AD-related healthcare costs.

METHODS: This retrospective observational study analysed insurance claims data linked to electronic medical records (EMR) from 01/01/2015 to 12/31/2021. Patients with ≥1 cognitive assessment (index date) and with ≥1 medical or pharmacy claim for AD diagnosis or AD medication within 120 days of an index date were selected. Potential confounding was addressed by adjusting for baseline characteristics via inverse probability of treatment weighting (IPTW). Generalized linear regression models compared AD-related HCRU and healthcare costs between early AD (EAD: mild cognitive impairment due to AD/mild AD dementia) and advanced AD (AAD: moderate/severe AD dementia). Least Absolute Shrinkage and Selection Operator (LASSO) regression with three-fold cross-validation identified predictors of high AD-related healthcare costs (in 80th percentile of costs). Multivariable logistic regression models were used to estimate the likelihood of high AD-related costs based on identified predictors.

RESULTS: A total of 193 patients were included (median age: 82 years; 63.2% female), 108 with EAD and 85 with AAD. Following IPTW adjustment, the AAD cohort had significantly higher overall AD-related HCRU (incidence rate ratio [IRR]=3.64, p<0.001) and outpatient visits (IRR=2.76, p<0.001) during follow-‑up compared to the EAD cohort. Similarly, total AD-related healthcare costs (cost ratio=3.26, p=0.003), medical costs (cost ratio=3.24, p=0.004) and outpatient costs (cost ratio=3.42, p=0.002) were higher in the AAD cohort versus the EAD cohort. Patients that were frail (odds ratio [OR]=2.12, p=0.047) or with AAD (OR=2.43, p=0.017) were significantly more likely to incur high AD-related healthcare costs.

CONCLUSIONS: Patients with AAD incurred higher AD-related HCRU and costs compared to those with EAD. Further research is necessary to determine whether interventions earlier in disease progression can mitigate these costs for AAD patients.

Code

EE340

Topic

Economic Evaluation

Disease

Neurological Disorders