Depression is a widespread condition associated with significant functional and social deterioration as well as extensive direct and indirect health care costs. A recent review of epidemiological studies estimates the annual prevalence rate of major depression at approximately 5% in Europe . Within the next 20 years, depression is predicted to become one of the leading causes of disability worldwide .
In 2001, the National Institute of Mental Health authorized additional research on preventing relapse in major depression as a part of a larger effort to find effective treatments capable of producing long-term durable recovery . Depression is a recurrent, potentially chronic and disabling condition. Acute treatments for depression, although effective, are often not sufficient enough for a large percentage of patients in preventing either subsequent functional impairment due to residual symptoms, or recurrent episodes. The primary objective of an intervention to prevent relapse is sustained remission of depressive symptoms. However, it is increasingly accepted that economic considerations need to be taken into account. Rising costs of interventions along with newer and more expensive antidepressant treatments bring up to questions about the cost-effectiveness of therapeutic interventions. Economic evaluation can assist decision-makers by providing additional support in making informed judgments concerning the allocation of increasingly scarce healthcare resources .
Quantifying the economic implications of a healthcare intervention requires precisely defining the target population, the characteristics of the disease and the therapeutic intervention. It also requires structuring the possible trajectory of patients in a logical, realistic order over time by considering the events that may occur, together with their health and economic implications. Providing a computational framework to illustrate disease progression over time as accurately as possible is necessary. Decision trees have been used successfully despite general recognition that they have severe limitations when applied to medical conditions . Markov models provided an alternative that allowed analysts to picture the course of a disease in terms of mutually exclusive health states and the transitions among them. While this technique considers time more explicitly and can be analyzed very efficiently, Markov models are considered highly rigid, mainly because of the lack of 'memory' imposed by the stochastic process.
Discrete Event Simulation (DES) models might offer a natural way of adequately depicting patient disease course throughout the health system  by making it possible to take into account important (baseline) prognostic factors together with life events interactions.
In this case study, our intention was to identify and compare the strengths and limits of discrete event simulation models with those of Markov models in portraying depression dynamics. To achieve this goal we employed a three-step process:
1. We described the clinical features specific to unipolar major depression.
2. We conducted a conceptual implementation of a Markov model and a DES model to detect possible abilities to address disease-specific issues relevant to major depression.
3. We discussed and compared the ability of each type of model to adequately reflect disease progression over time.