Using "costs states" in a semi-Markov model to estimate cost-effectiveness with an illustration for metastatic HER2+ breast cancer in the Czech Republic

This paper proposes an original method for assessing costs of medical treatment. It defines states in a semi-Markov model associated with specific costs of the treatment, and not with patients' health statuses. Costs assigning to these "costs states" is more straightforward; moreover, it allows to estimate the periods separately when no treatment is administered. This method is applied to individuals' data drawn from the Czech clinical practice in the treatment of metastatic HER2+ breast cancer. The aim is to assess the cost-effectiveness of adding pertuzumab to the combination of trastuzumab+docetaxel within first-line therapy. The Kaplan-Meier estimates of survival functions were supplemented by the Cox proportional hazard model and the accelerated failure time model that both control for patients' characteristics. Based on the employed data, the addition of pertuzumab does not result in significantly longer patients' survival. Since the treatment is associated with higher costs, adding pertuzumab is not considered to be cost-effective; however, this could be due to relatively short patients' follow-up that is available at the moment.