Over the last few decades, we have seen a spurt of medicines approved by regulatory authorities which have necessitated an increase in the need for monitoring drugs. Pharmacovigilance, is an arm of medical science which deals with gathering, monitoring and assessing the safety of drugs. It has gained increased focus and prominence, over the last decade. It’s thus not a surprise, that with changing times, regulatory teams are increasing channels for monitoring and collecting adverse event reports (ADRs) from social media and websites as well.
Processing medical literature, CIOMS and patient reports for ICSR, involves understanding the complete text and entering into the case, the details of the adverse event, drugs, patient and his/her history and so on, for further processing steps namely, Quality Review and Medical Review. Due to high data content, extracting ICSR specific information is one of the biggest challenge facing the industry.
Artificial Intelligence for ICSR processing
With the strides taken in technology, we are at a cusp of a revolution in Artificial Intelligence, which will have an impact on our everyday life. We are already seeing its increased usage in fraud detection, recommendation engine, crop classification etc. AI methods and systems such as Natural Language Processing (NLP), Machine Learning, and Neural Networks are now at our disposal to not only synergize but also improve with the current workflow.
Following are the broad two categories of usage AI technologies in ICSR Processing:
- Ingestion of structured and unstructured content:- Comprises of components for reading incoming case intake information via XML, Docx, images including PDF and PDF text including forms-tables. Here OCR/ICR along with NLP/machine learning is used to extract ICSR information from information sources in a regulatory compliant manner.
- AI for decision making:- Sometimes, the quality of information available in ICSR is poor. In such scenarios, Semi-supervised or unsupervised learnings play a major role in devising hypothesis. For example, building Unlisted Events and Drugs Correlation, Causality Classifiers etc., specific type of Neural Networks are built and improvised with training over a period. These are faster and more accurate compared to other methods.
Key considerations of Artificial Intelligence in ICSR processing
The complete system, has multiple complex sub-systems and driven by accuracy and user feedback. One of the challenges and goals of such implementation will be high degree of Configurability and adaptability to different formats.
The efficiency of such systems can be observed at long time intervals and data richness, typically thousands of live case processing and many years of data which leads to arriving at appropriate configuration of the system parameters.
One of the strategy for arriving at the model configuration as an initial upstart is, using historically processed case data, as training samples for the AI system.
As with any Artificial Intelligence system implementations as described above, doesn’t intend to completely replace human element, but complements the process, and helps in identifying and bringing out seemingly hidden relationships for ensuring accurate ICSR processing in Pharmacovigilance.
Technology day by day is reaching the zenith and bringing new things in front of us for convenience. The advent of artificial intelligence in ICSR processing brings about the following benefits to the industry:
- Reduced cycle times: It significantly reduces cycle time achieved by processing cases faster through automation.
- Improved quality and accuracy: Achieved through standardized inputs and automated case intake and processing.
- A complementary solution of existing safety databases: Artificial Intelligence can be implemented without much disruption to the systems and processes currently managed within pharma ecosystem due to its adherence to standards like E2B natively.
- Scalable and futuristic solution: AI allows to handle volume growth by managing the growing Adverse Events (AE) volume and diverse types of incoming data formats.
- Foundation roadmap: It lays a roadmap for use cases within and beyond Pharmacovigilance e.g. landscape analysis, Real World Evidence, Enterprise Knowledge Management and Quantitative Sciences.