We’ve spoken before about using automation and artificial intelligence (AI) for pharmacovigilance literature monitoring. We’ve also discussed the benefits automation adds to our end-to-end medical literature monitoring solution – such as reducing Aggregate Report and Safety Signal review workload by 70% and eliminating the wasteful re-reviewing of duplicate references.
But what we want to do in this blog is introduce you to the technology behind our approach to intelligent pharmacovigilance automation – DialogML.
The DialogML relevancy ranking engine is part of the Dialog system, and is designed to make it easier for literature review teams to identify and process references containing Individual Case Safety Reports (ICSRs) and drug safety issues that are important for Aggregate Reports, and Safety Signals
For example, DialogML can give review teams the potential to find the ICSRs in a batch of articles 5x faster than without DialogML and bulk review up to 50% of the references from the ICSR workstream.
How does DialogML deliver AI for pharmacovigilance literature monitoring?
First, you create a pharmacovigilance search for your drug in Dialog and set it up to automatically run at the scheduled time using the Dialog alerting engine. Dialog outputs the results in a qualitative XML format and removes duplicate references automatically, reducing the initial volume of references to be reviewed.
DialogML then uses artificial intelligence to apply a patient safety relevancy ranking to each reference in the XML output.
The relevancy ranking is applied independently for ICSR and non-ICSR patient/product safety issues. This allows the search results to be prioritized both by their likelihood of containing an ICSR and by their likelihood of containing non-ICSR patient safety issues.
The XML output from Dialog, enhanced by DialogML, can then be ingested by most literature review tools. As a result, Dialog and DialogML can be used to easily add machine learning assistance to nearly any literature monitoring workflow.
DialogML and Drug Safety Triager
But the real magic of DialogML appears when it is used in tandem with our literature review system, Drug Safety Triager. DialogML adds intelligent automation to the automation already provided by the Drug Safety Triager. The relevancy ranking from DialogML and highlighting of key safety concepts in Drug Safety Triager are effective in making the literature monitoring process even more efficient and provide a workload reduction without introducing additional risk.
What’s more, when DialogML is used with Drug Safety Triager a feedback loop is enabled, allowing for continuous improvement and fine tuning of the DialogML machine learning model, based on comparison with the users’ manual reviews.
We’re working with some of the biggest pharmaceutical companies in the world to use DialogML to deliver AI for pharmacovigilance literature monitoring. We can do the same for your company – use the contact us form or email us directly on email@example.com.
Artificial intelligence in pharmacovigilance: Using AI to detect ICSRs in scientific literature
The question many Market Authorisation Holders (MAH) are asking themselves at the moment is how...
Seeing double: eliminating duplicate references in drug safety literature screening
Drug safety literature screening for patient safety issues is not easy an easy process. As we saw...
Staying alert: how to easily update search alerts for drug safety literature monitoring
If it isn’t documented, it doesn’t exist. As far as regulators are concerned, this is true for any...