339418
Developing Informatics Tools for the Evaluation of Large Health Data Sets
As more health data is collected, analyses become increasingly complex and arduous. The United States Food and Drug Administration (FDA) conducts routine surveillance of medical products to assure continued safety after approval. Medical officer case-series analysis is a mainstay of traditional pharmacovigilance, and FDA is developing informatics tools to stream-line this activity by deconstructing and organizing clinical information in adverse event (AE) reports.
methods:
Event-based Text-Mining of Health Electronic Records (ETHER) and Pattern-based and Advanced Network Analyzer for Clinical Evaluation and Assessment (PANACEA) use natural language processing and network analysis to facilitate the analysis of clinical data by deconstructing text into components such as products, diagnoses, symptoms, and time stamps; by automating the translation of terms into standard medical dictionaries such as the Medical Dictionary for Regulatory Activities and the International Classification of Diseases; and by visualizing data and aiding in pattern recognition.
ETHER uses natural language processing to summarize pertinent information in an AE report, reducing the amount of time medical officers spend reviewing AE reports, and allowing them to quickly identify and concentrate on reports of concern. The tool can also visualize the evolution of clinical information supplied within or across AE reports; thereby facilitating the identification of time patterns and symptom clusters.
PANACEA uses network analysis to construct networks made of “element nodes” such as exposures (drugs and vaccines) and outcomes (adverse events), or networks made of nodes representing cases (AE reports). Varied algorithms and visualizations permit the identification of patterns and clusters within the data.
outcomes:
When applied to AE report data, the tools allow reviewers to rapidly evaluate potential associations between products and AEs. FDA continues to use, evaluate, and iteratively improve tool functions and algorithms.
conclusions:
The value of these tools extends beyond AE data; general clinical data could be deconstructed and collated to identify patterns using these tools and methodologies. FDA is exploring the possible application of these tools to claims data and electronic medical records.
FDA envisions making these tools publicly available to aid in evaluation of health data, identification of new pharmacovigilance or public health applications, and enhancement of tool performance.
Learning Areas:
Communication and informaticsEpidemiology
Other professions or practice related to public health
Learning Objectives:
Discuss the utility of new informatics approaches in evaluating large amounts of clinical and health data
Demonstrate use of the ETHER and PANACEA tools
Keyword(s): Information Technology, Network Analysis
Qualified on the content I am responsible for because: I have been one of the key medical personnel working on the development of the ETHER and PANACEA informatics tools within FDA CBER. I am a medical officer in FDA CBER trained in preventive medicine, and posess a bachelors degree in mathematics.
Any relevant financial relationships? No
I agree to comply with the American Public Health Association Conflict of Interest and Commercial Support Guidelines, and to disclose to the participants any off-label or experimental uses of a commercial product or service discussed in my presentation.