Analytical approach:
This economic evaluation was based on a micro-simulation decision model which determined changes in cognitive function over time. A lifetime horizon was considered. The authors stated that the perspective adopted was that of society.
Effectiveness data:
The clinical data were derived from published studies which were selected and justified by the authors. The key characteristics of the sources used were reported. For example, the data on disease progression came from an observational study of 180 patients with AD. The data on treatment effect for memantine plus donepezil compared with donepezil alone were derived from a recently published phase III randomised controlled trial (RCT) with a 12-month follow-up. The donepezil treatment effect was taken from a placebo-controlled clinical trial and superimposed on the natural history of AD. Some assumptions were also made, mainly on the long-term impact of the two treatments. The key clinical outcome was the impact of treatment on cognitive function, assessed by the Severe Impairment Battery (SIB).
Monetary benefit and utility valuations:
The utility valuations were derived from published data on the relationship between the Health utilities Index Mark 3 (HUI3) and the disease severity. These data came from a cross-sectional study in which the HUI3 was administered to caregivers of 679 patients with AD, dwelling in the community and in institutions.
Measure of benefit:
Quality-adjusted life-years (QALYs) were used as the summary benefit measure and a 3% annual discount rate was applied. Changes in SIB score were also reported, but were not combined with costs.
Cost data:
The analysis considered both expenditures on formal health care services (inpatient, outpatient, home health, pharmacy, and long-term care) and informal services (unpaid caregiver time) to reflect the societal perspective. The costs and quantities were presented as macro-categories related to the severity of disease. These costs were derived from published sources. Drug dosages were based on consumption patterns in the RCT and costed using average wholesale prices. The costs were in US dollars ($) and the price year was 2005. Future costs were discounted at an annual rate of 3%.
Analysis of uncertainty:
A deterministic univariate sensitivity analysis was undertaken by varying the model inputs by plus 10% or minus 10% of their baseline values, except for the baseline score for mental impairment which was varied using the published range. The best and worst case scenarios were considered.