Health utilities in Alzheimer's disease and implications for cost-effectiveness analysis.

The National Institute for Health and Clinical Excellence (NICE) provisional decision against memantine and other medications for Alzheimer's disease (AD) has generated much discussion. In its decision, NICE expressed concern about the data source used for the utility scores in the industry submission of the memantine model. NICE therefore turned to an alternative data source. However, in doing so, it made the key assumption that moderate-to-severe AD patients living in the community were independent. Furthermore, the NICE data source also had its limitations. There are numerous limitations in inferring from available data how utilities vary between dependent and non-dependent patients with AD. Most importantly, we lack direct evidence from primary data. Nonetheless, it seems reasonable to assume that patients with severe AD, and likely those with moderate AD as well, have full-time care needs, regardless of their setting. The NICE assumption that they do not results in a difference in utilities between dependent and non-dependent AD individuals of only 0.06. This seems to be at the low end of what one would consider a reasonable estimate.