Detecting Adverse Drug Event Signals from a Clinical CaseBase

With the rapid development of medical information systems in Chinese hospitals over the last two decades, many of these organizations have accumulated astronomical amounts of structured and unstructured clinical data, including patient diagnostic data, treatment data, lab test data, etc. Secondary use of these data for research, such as Real World Evidence (RWE) studies, has the potential to improve medical quality and safety in daily clinical practice. In this study, we describe CaseBase, a Clinical Data Warehouse (CDW) that extracts structured clinical symptoms or findings and related temporal information from narrative clinical documents and integrates medication information from the CPOE system. An Adverse Drug Event (ADE) signals detection platform has also been developed based on CaseBase to analyze and visualize drug-symptom relations in clinical data. A prototype of this platform has been evaluated in a 2,000-bed hospital and some initial results are reported here.