PP09-66: Otsuka eWriter: An Automated Authoring Tool for Patient Narratives
Poster Presenter
Boris Reznichenko
Director, Head of Global Regulatory Operations
OPDC United States
Objectives
To compress narrative writing time by developing a validated tool with capabilities that facilitate data retrieval from multiple sources to assist in compilation, evaluation, and interpretation of information that supports the content development.
Method
Otsuka Global Regulatory Affairs has developed validated software (eWriter) using proprietary patent pending technology to apply logic with the desired formatting to generate templates that can be auto-populated with document-specific content in text that follows user-selected style guidelines.
Results
Otsuka eWriter is a user-friendly web-based application designed with multiple functionalities to prepare narratives using significantly less time with more reliable quality and consistency than manual transcription. eWriter enables automated authoring through mapping clinical data to fit various formats tailored to specific document types. It also allows sharing common data selection and formatting between document types and simplifies reuse of common rules for complex variables (eg, adverse events of special interest [AESI]), qualifiers for events, and style guidelines. eWriter does not restrict content/formatting or data flow. It provides functionality to perform aggregated processing of all data domains and to support various data presentations (eg, incremental, percentages). eWriter also supports customization for run-time searches of multifaceted information on clinical abnormalities (including joint rules for clinical domains); sorting categories; simplified methods of data correlation; and multiple logic branches to accommodate complex document content/formatting scenarios. Use of eWriter as an authoring tool has been applied to writing safety narratives, which are detailed summaries of important medical events experienced by subjects in a clinical trial. Narrative writing requires extensive time and effort as it involves compilation of different types of data from multiple sources into a single document. With eWriter, a narrative template can be generated with both standard (for reusability) and trial-specific text and fields. Clinical data required for narratives is automatically extracted by eWriter from programmed tables and listings and other databases while the remaining safety data is manually appended. eWriter supports multiple languages within the prepared output. Lastly, eWriter has been shown in actual practice to effectively support narrative writing in clinical trials with =10 narratives.
Conclusion
Otsuka eWriter is a user-friendly tool with applications that effectively supports narrative writing. It can also be used to assess resources required to complete a narrative. eWriter creates a source document that has the ability to pinpoint changes must be made to completed narratives. Use of eWriter improves document quality and accuracy with considerable time savings. With eWriter, timelines can be shortened without compromising quality because time needed for quality reviews is substantially reduced by auto-extraction of data. Other advantages of eWriter include consistency in content and format across multiple writers, fewer errors, and less rework compared with transcribed documents. A test of productivity gains achieved by use of the eWriter application for narrative writing showed that it reduced completion time to at least half the time required for transcribed narratives. With this time savings, twice the number of narratives could be completed over the same period or resources needed to complete the same number of narratives could be reduced by half ? allowing medical writers to focus on strategic components. The next versions should further improve user experience by using a natural language/visual data interface for the expected content. Lastly, additional improvements in data interface with safety data will move the entire content creation process closer to full automation.
The authors wish to acknowledge Dr. Henrietta Ukwu for her support in making this system a reality.