With these previous limitations in mind, the authors have developed an economic model incorporating micro-simulation techniques that will allow modelling a single patient at a time, generating clinical events and treatment
switches according to preset probabilities and triggers. This type of model allows for the occurrence of any number of events such as exacerbations, decline in lung function, worsening of dyspnoea (as measured by clinical instruments) and improvement in QOL (as measured by diseasespecific or generic instruments). Following the taxonomy described by Brennan et al. this model may be considered an individual sampling model (ISM) that simulates each individual through discrete time periods. The major difference to previous COPD models is the absence of defined health states per se. Outcomes are estimated on the basis of the presence/absence of events and lung function levels modelled as a continuous variable. These changes increase the flexibility with which to describe the different
combinations of clinical events that can be experienced by a patient over the course of their disease. GOLD state definitions are assigned whenever estimating outcomes that are related to disease severity, but there is potential to avoid using these state definitions if outcomes and pay-offs can be estimated independently of GOLD disease severity.