Automated generation of decision-tree models for the economic assessment of interventions for rare diseases using the RaDiOS ontology

Abstract Objective: The development of decision models to assess interventions for rare diseases require huge efforts from research groups, especially regarding collecting and synthesizing the knowledge to parameterize the model. This article presents a method to reuse the knowledge collected in an ontology to automatically generate decision tree models for different contexts and interventions. Material and methods: We updated the reference ontology (RaDiOS) to include more knowledge required to generate a model. We implemented a transformation tool (RaDiOS-MTT) that uses the knowledge stored in RaDiOS to automatically generate decision trees for the economic assessment of interventions on rare diseases. Results: We used a case study to illustrate the potential of the tool, and automatically generate a decision tree that reproduces an actual study on newborn screening for profound biotinidase deficiency. Conclusions: RaDiOS-MTT allows research groups to reuse the evidence collected, and thus speeding up the development of health economics assessments for interventions on rare diseases.