Automation in Routine Use for Data Collection and Processing for Scalable Faster RWE Generation
Author(s)
Kassekert R1, Easwar M1, Glaser M1, Ventham R1, Bate A2
1GSK, Collegeville, PA, USA, 2GSK, Dorking, Surrey, UK
Presentation Documents
Introduction Rapid and effective data collection and analysis for RWE generation has never been more important. Data collected at scale with near-real time ingestion including transparent audit trails will increasingly require automations that reliably process and make data available for analysis quickly. Safety data are collected internationally across heterogenous settings at scale. We describe the impact on routine processes of two Robotic Process Automations (RPAs) and insights for furthering zero-touch processing capability. Methods A duplicate check assistant (DCA), utilizes a list of safety case ids (input spreadsheet) to be checked for duplicates. DCA automatically logs into the safety database, navigates menu and selection items to open each case, extracts relevant data elements, searches the database for potentially duplicate cases using pre-defined sequences, captures resulting information into the input spreadsheet, and attaches the spreadsheet to the respective case in the safety database. A second data quality reviewer (DQR) extracts relevant field contents from the database into the input spreadsheet and field discrepancies are highlighted based on pre-defined rules. Results For a week in March 2020, DCA executed on 1510 cases, reducing human effort from 3 mins to 1 min 42 secs per case; DQR on 458 cases from 38 mins 27 secs to 28 mins 25 secs per case. Both automations executed without error or human intervention. Since routine use, the automations have resulted in estimated savings of 1202.3 person-hours. Conclusions Routine production-use automation for safety data has provided tangible benefits through fast and consistent workflow execution with minimal manual intervention. Implementing automations requires close examination of existing workflows and extensive end-user involvement in requirements development and testing. Understanding how to minimize human intervention while fulfilling business goals is critical. Expanding rule-based automation while incorporating AI and ML capabilities will provide even further benefits for Safety and RWD collection.
Conference/Value in Health Info
2020-11, ISPOR Europe 2020, Milan, Italy
Value in Health, Volume 23, Issue S2 (December 2020)
Code
PNS271
Topic
Economic Evaluation, Health Technology Assessment, Real World Data & Information Systems
Topic Subcategory
Cost-comparison, Effectiveness, Utility, Benefit Analysis, Data Protection, Integrity, & Quality Assurance, Decision & Deliberative Processes, Systems & Structure
Disease
No Specific Disease