Automated Tools for Clinical Research Data Quality Control using NCI Common Data Elements

Clinical research data generated by a federation of collection mechanisms and systems often produces highly dissimilar data with varying quality. Poor data quality can result in the inefficient use of research data or can even require the repetition of the performed studies, a costly process. This work presents two tools for improving data quality of clinical research data relying on the National Cancer Institute’s Common Data Elements as a standard representation of possible questions and data elements to A: automatically suggest CDE annotations for already collected data based on semantic and syntactic analysis utilizing the Unified Medical Language System (UMLS) Terminology Services’ Metathesaurus and B: annotate and constrain new clinical research questions though a simple-to-use “CDE Browser.” In this work, these tools are built and tested on the open-source LimeSurvey software and research data analyzed and identified to contain various data quality issues captured by the Comprehensive Research Informatics Suite (CRIS) at the University of Arkansas for Medical Sciences.