Pharmacovigilance Triaging

Overview

TCS' Pharmacovigilance Triaging app automates the Signal Intake process optimizing the manual effort and reducing error in the process through use of TCS’ Text and Email Classification approach providing a solution to challenge of using Natural Language and unstructured data in a case management system for decision making.

Life Science organizations regularly have to deal with multiple sources of signal data which have to be tracked, analyzed and responded to minimize the risks for their products. Pharmacovigilance triaging processes are usually manual, handled by back offices, with minimal integrations and prone to human errors with added complexity of dealing with long text and emails related to the signal. TCS' Pharmacovigilance Triaging app helps automate and optimize the signal intake process by leveraging Appian as a system for capture of the Signal data from different sources and easy integration to back-end Signal Data Repository. The app uses the TCS Email and Text Classification approach which uses the latest improvements in the Machine Learning and Natural Language processing to process text and automate case assignment.

Key Features & Functionality

Reads data from Adverse Event PDF forms for case processing and storage
Checks the system for duplicate cases and triaging the cases through task assignments
Classifys cases based on seriousness criteria in order to prioritize cases for further processing through use of TCS’ Email and Text Classification Machine Learning Algorithms
Provides the case view at the product level to understand trends and analysis based on the same
Workflow assignment based on the severity of the casesignal

Benefits & Business Impact

Operational efficiency: Improve operational efficiency through the integration and optimization of manual efforts in the process.
Flexibility: The application provides the ability to accelerate the development of the E2E PV triaging process and can be integrated into existing applications in the landscape.
Technology reusability: Uses the Text and Email Classification models, which are plug and play into any workflow that needs email/text to be classified for the workflow.
Reporting features: Shows the accuracy of the model.