Toward an Ontology-Based system for the Automatic Detection of Biases and Weaknesses in Drug Clinical Trial Results

摘要

New drugs can improve patient health, but require careful evaluation and comparison with older drugs in clinical trials. Clinical trial results are used as the basis for prescription in evidence-based medicine (EBM), but such trials can be impaired by various biases. We present a system for automatically detecting some biases, design flaws and other weaknesses impairing the results of a clinical trial, to help physicians to come to their own conclusions about evaluations of new drugs.