Nikolaos Matskanis, Joseph Roumier, Fabrice Estiévenart, Using Open Source software and Open data to support Clinical Trial Protocol design, Med-e-Tel Conference, April 2014.
Clinical research, including investigation of the test of hypothesis and clinical trial design, comprises an expensive and very important part of the R&D activities of the pharmaceutical companies. Trials for drug repositioning, which aim at testing established compounds to new medical conditions, appear to gain ground as these compounds have already been found safe in clinical trials and sometimes are already present in the market. Although the uncertainty of safety is lower in such trials, still problematic situations - for example an adverse effect of a drug compound - present important risks.
In many cases evidence of such issues already exists in published scientific articles and other clinical data sources but is often difficult to discover and correlate given the volume, variety and distributed nature of the information required. In addition structuring this information into a valid trial protocol can be a challenging and time-consuming task.
Scientific results from medical research, including those from clinical trials, are gradually being published as Open Data and a significant proportion of them is becoming available also as Linked Open Data (LOD).
This rapidly growing collection of well-structured (using RDF, OWL) and easily accessible (HTTP) data contains clinical information about drugs and their side effects, clinical trials, diseases and their mechanisms (pathophysiology) and other aspects of the domain. These semantically linked data sources are especially valuable as they allow semantic search tools to discover information that is otherwise hard to come across.