Requirements and strategy for the development of a pediatric drug ontology

Clinical and epidemiological researchers across all medical special-ties need tools and knowledge representations to support the classi-fication, aggregation, and analysis of medication data. The National Drug File Reference Terminology (NDF-RT), a named standard for classifying medications, is developed by the US Department of Vet-erans Affairs (VA) as an extension to their National Drug File, which is the master list of drugs prescribed to VA patients, which are adults. NDF-RT is organized as a multi-axial hierarchy with addition-al relations between ingredients, medications, chemical structures, mechanism of action, and therapeutic indications. We describe our experience applying NDF-RT to a dataset of encoded medications that were collected from an international cohort of over 8,000 chil-dren. Our data-driven approach allows us to extract selected NDF-RT sub-classes of a researcher-provided concept of "antibiotics". We believe that a subset of concepts and relationships from NDF-RT will be sufficient to support pediatric research analyses involving classes and properties of medications, and that an NDF-RT subset relevant to pediatrics will be more easily adopted by clinical investi-gators and epidemiologists, thereby promoting standardization of drug classifications. Researchers from all domains would benefit from informatics tools utilizing ontologies to support data cleaning and analysis that is explicit, valid, and repeatable. We predict that a pediatric drug ontology view can be extracted from the NDF-RT reference ontology, and we hope for feedback from the ontology community on ways to advance this idea.