Increasing Efficiency and Cost-Effectiveness by Automating the Authoring of the Development Safety Update Report

A Development Safety Update Report (DSUR) is a comprehensive review of safety information collected during an annual reporting period for an investigational drug. The creation of a DSUR is a resource-intensive process, requiring effective communication among, and collaboration with, various functional stakeholders.

To increase the efficiency and cost-effectiveness of generating a DSUR with the use of an electronic authoring platform. This computerized platform auto-populates the DSUR template using available source data.

Requirements for the DSUR authoring tool were developed, and available data sources were mapped to respective sections of a company DSUR template. The DSUR authoring tool, mined, scanned, and extracted the relevant data from multiple-input source documents using Natural Language Processing logic. The data were then auto-populated into the template, as required by regulatory guidance ICH-E2F, with different text options based on questions and answers. The time- and cost-savings gained through automation were analyzed for DSURs authored during 2017-2019.

The DSUR contains 31 sections (20 main sections and 11 sub-sections), of which 9 were fully automated, 12 were partially automated, and 10 were not automated in this release of the automation tool. Use of the DSUR automation tool resulted in a time-savings of approximately 25% per report, corresponding to a cost saving of US$4550 per report.

Deployment of the automation tool reduced the time and cost required to manually produce DSURs, as well as improved the quality, compliance, and consistency of company-prepared DSURs. This tool enhanced the potential of the organization to generate larger volumes of DSURs in a timely manner, without requiring additional resources or compromising quality. This or a similar authoring automation tool could be employed in the future to generate other regulatory submission documents, which also require various source documents and cross-functional contributions.