Towards the economic evaluation of two mini-invasive surgical techniques for head&neck cancer: A customizable model for different populations

Abstract

Sometimes representing a decision process involves building multi-stage models, where some outputs of stagei become the inputs of stagei+1. This was the case for an economic evaluation study of two mini-invasive techniques for head&neck cancer, namely Transoral Robotic Surgery and Transoral Laser Microsurgery. We built a two-stage model, composed by a first decision tree ac-counting for the surgical complications and need for additional treatments, which in turn are used as initial conditions for a second decision tree that models long-term outcomes through a Markov process. To allow the automatic concatenation of decision trees, we developed a Java extension to the TreeAgePro software, a well-known tool for decision analysis. Moreover, we integrated the resulting model into UceWeb, a framework we developed over the last two years, which allows personalizing costs and preferences for different target populations. In this way the same model may be re-used to perform cost/utility and cost/effectiveness analysis in different settings.