Analytical approach:
The analysis was based on a published Markov model, with a five-year time horizon. The authors stated that a societal perspective was adopted.
Effectiveness data:
Most of the clinical inputs were from the published decision model. Some more recent and relevant data were from Dutch studies. For example, the initial distribution of patients between model health states was from a subsample of 339 patients in a study of disorders associated with dementia in 7,528 residents of a Rotterdam suburb. A key input was the efficacy of memantine over usual care, which was from the pooled analysis of four six-month, double-blind, randomised, placebo-controlled trials. The treatment effect after six months was based on an open-label extension study. Mortality was from Dutch published studies.
Monetary benefit and utility valuations:
The utility values for the level of dependency and residential status were from a UK study of 224 patients with Alzheimer’s disease.
Measure of benefit:
Quality-adjusted life-years (QALYs), time in an independent state, and time in a moderate state were the summary benefit measures. They were discounted at an annual rate of 1.5%.
Cost data:
The economic analysis included the costs of visits, family care, district nursing, hospital admission, regional institutes for mental health, day-care centres (including travel), and informal care. The resource quantities were from Dutch sources, including official guidelines and a model of the cost of care for patients with dementia. The unit costs were reported and were from official Dutch prices. The costs were in Euros (EUR) and were discounted at an annual rate of 4%. The price year was 2008.
Analysis of uncertainty:
A probabilistic sensitivity analysis was carried out using a Monte Carlo simulation, with 10,000 iterations, and using predefined probability distributions for the model inputs. An alternative worst-case scenario assumed that patients received treatment for only two years and its efficacy lasted for only six months. Univariate sensitivity analyses were performed on the odds ratios for the impact of memantine on dependency and severity, on the discount rates, and on the distributions between nursing homes and homes for the elderly.