Background/Rationale: To determine the suitability of published estimates of the US cost of Alzheimer’s disease (AD) for use in cost-effectiveness models for new AD treatments. Methods: A systematic literature review of published information on direct medical, direct nonmedical, indirect, and informal care costs for different levels of disease severity. Results: Nineteen studies were included in the review. In studies presenting mean costs by disease severity, the change in different types of costs with increasing disease severity varied, depending on the data sources and characteristics of patients with AD. In studies presenting the results of regression analyses, costs were shown to be independently associated with cognition, functional status, behavioral symptoms, and dependence. Conclusions: Published US studies (1) did not include all the types of costs and AD populations, and (2) generally did not include all the measures of disease severity that are needed for cost-effectiveness models.
