FY2015 Regulatory Science Research Report: Database and Knowledge Management
Major program commitments under GDUFA include metrics for applications (e.g., a 10-month review period to act on 90% of complete ANDAs), backlogs, current good manufacturing practice inspections, efficiency enhancements, and regulatory science. DQMM is constructing a database of application-specific data that will greatly enhance generic drug review quality, efficiency, and consistency. An integrated data platform accumulated from a wide range of scientific and regulatory sources throughout the Agency should also be constructed and used to guide regulatory review, research, and decision-making activities.
To achieve information visibility, OGD needs an information technology (IT) solution where all data contained in ANDA submissions are federated into a single data source and enhanced with application-specific metadata (such as from Drugs@FDA) and product-specific BE guidance. This database could be mined for specific information related to a single ANDA or a product-specific guidance and could present an aggregate view across a specified filter (e.g., approved extended release tablets containing a specific grade of a specific rate-controlling excipient). Other important aspects of this IT solution would include:
- Standardizing the transmission of Clinical Data Interchange Standards Consortium (CDISC) Study Data Tabulation Model (SDTM) data to facilitate loading of the database.
- Developing a user interface layer. This layer will allow for improved visualizing of the data, sharing of code, reproducing results, and implementing new analytical methods.
- Incorporating the complex nature of the data into models in the internal database and developing the proper relationships among data elements from FDA and external data sources.
- Mining text from, for example, ANDAs, ANDA reviews, NDAs, NDA reviews, and drug product labels that contain a significant amount of unstructured data.
- Federating other data sources to content within the ANDA-related database. A wealth of information (e.g., adverse event reports, drug product prescription rates, drug product labels, drug approval information, drug substance physiochemical properties, and drug substance therapeutic classification) contained in databases available from FDA and the Department of Health and Human Services, as well as external databases, could be used alongside the ANDA-related database to improve the premarket and postmarket analysis of generic drugs and assist in the prioritization of reviews and surveillance efforts, respectively.
Currently, in support of other topics in this report, ORS collects data to develop the initial internal database. For example, information pertaining to modified release solid oral dosage forms is gathered and integrated with other data sources, such as Drugs@FDA, drug product labels, and therapeutic classification systems to provide a single view. This integration essentially serves as a pilot to demonstrate the utility of information visibility for knowledge creation. As more data are collected, quantitative risk assessment models will be developed to support application and postmarket surveillance prioritization. This work will support the eventual development of a production business intelligence system with federated search capability.
Table 1. Knowledge management roadmap
ORS staff facilitating research in this area
- Andrew Babiskin, Meng Hu, Hopi Lin, Sarah Dutcher, Liang Zhao, and other ORS staff members
Projects and Collaborators
- Office of Bioequivalence, OGD
- Division of FiIing Review, OGD/ORO
- Office of Computational Science, OTS
- Other stakeholder offices across FDA
Publications and Presentations
- Research projects in progress
- Research projects in progress