OBJECTIVES: To synthesize the available evidence considering full economic evaluation that used memantine alone or combined with donepezil for moderate to severe Alzheimer's disease focusing on the analytical decision models.

METHODS: The electronic databases MEDLINE, EMBASE, NHS EED, CEA Registry and LILACS were used to search for studies. Duplicate references have been removed. Abstracts and titles were been evaluated by two independent reviewers. The complete texts were evaluated for two reviewers too and disagreements were resolved by a third reviewer. After the application of the eligibility criteria, fifteen complete economic evaluations were included. Drummond et al's tool was used to evaluate the quality of the studies.

RESULTS: Regarding the study designs, one evaluation was conducted throughout the clinical trial, eleven references used Markov models and three using microssimulation model. The main measure of outcome adopted was cost per QALY and the majory of articles had considered the societal perspective on the analysis. The use of memantine was considered cost-effective and dominant in eleven studies; in a single study, its use was dominated when compared to donepezil for moderate Alzheimer's Disease. Sensitivity analyzes were systematically performed, showing robust results. The quality assessment indicated good methodological quality of the work.

CONCLUSIONS: Although there is controversy regarding the benefits derived from the use of memantine associated or not with donepezil, evidence collection suggests that it is cost-effective in the countries where the studies were done. However, local economic studies need to be performed considering possible changes on the parameters, perspectives and policy that were be adopted in countries where these evaluations were conducted.