Master thesis

Jorge Pérez Martín.
Cost-effectiveness analysis in OpenMarkov. Application to bilateral paediatric cochlear implantation in Spain. Dept. Artificial Intelligence. UNED, Madrid, Spain, 2015.
Supervisor: Prof. Francisco Javier Díez Vegas.

86 pages. PDF (1.913 KB), BibTeX entry.


Cochlear implantation is nowadays the most effective therapy for the treatment of severe to profound deafness. Different studies have found that unilateral cochlear implantation is cost-effective in both children and adults. Recent research has shown that bilateral cochlear implantation (an implant in each ear) is cost-effective in children, but the degree of uncertainty is still high. The goal of this study is to determine as precisely as possible whether pediatric bilateral cochlear implantation is cost-effective. In order to accomplish it, we have built a probabilistic graphical model that represents the events that may occur during the life of a cochlear implant user, their impact on quality of life and their associated costs, distinguishing those covered by the national health system from those that in Spain are covered by users and their families.

We have built and evaluated the model with the open source software OpenMarkov, developed by the Research Centre for Intelligent Decision-Support Systems (CISIAD) at UNED, and for this purpose we have created new functionalities and extended some of those already existing.

Several studies have estimated that the willingness to pay (also called “cost-effectiveness threshold”) implicitly used by the Spanish health system lies between 30.000 and 35.000 € per quality-adjusted life year (QALY). According to the results of our model, the incremental cost-effectiveness ratio of simultaneous bilateral implantation with respect to unilateral implantation for one-year-old children is 12.133 €/QALY), which implies that all those needing two implants should receive them. The sensitivity analysis confirms this conclusion, since for a threshold of 30.000 €/AVAC the probability that bilateral implantation is cost-effective reaches 98 %. Even if the national health system covered the costs that in our country are currently paid by users and their families, the intervention would still be cost-effective, with an ICER of 17.263 €/QALY and a probability of 94 %. Implantation should be done simultaneously, i.e., in one surgery, because sequential implantation has the same effectiveness but is more expensive. Our model is not able to estimate the ICER for children older than one year because the effectiveness depends on many factors, mainly on the current age, the age at the onset of deafness, when it was diagnosed and, if the child already has an implant, when he/she received it.

Our model will be publicly available so that other researchers can reproduce the results and adapt it to other contexts. In particular, it would be interesting to analyse the ICER for adults and in other countries.