Today, pharmaceutical and medical technology (MedTech) safety teams spend the vast majority of their time manually handling adverse events (AE) collection and processing. Timely insights from this data are critical for managing product risks, but data is often stored in disparate databases by different team members. This creates siloes of data that are difficult and costly to aggregate and analyze.

AE reporting continues to increase 15% annually. The problem of time spent processing data vs.  managing risks with data-driven insights will only worsen as volumes increase. The dilemma is exacerbated as the sources of safety data proliferate, such as emails, social media, journal articles, audio files, handwritten documents, real-time data from devices and other unstructured formats that cannot be easily manually processed. In response, there is a growing movement to use artificial intelligence (AI) to help change the safety team’s focus from the manual handling of sources to putting data and insight into use within the organization. This shift away from manual to automated processes is commonly referred to as “Zero Touch Case Processing.”

There are three key steps in case processing: receiving, transforming and analyzing information. Today, AE automation is focused on the second step of information transformation. This stemmed from the industry’s assumption that automating the transformation of data would fix upstream and downstream issues in receiving and analyzing AE information.

Despite advances in automating data transformation, there is still a lot of work that must be performed in the first stage of receiving information before transformation can start. The growing sources of AE information increase the complexity associated with AE processing as well.

Safety teams have attempted to tackle the increase in AE volumes and complexity by simply assigning more workers to case processing operations. This strategy is unsustainable due to the limited availability of trained workers, especially amidst a multi-year pandemic when worker shortages plague every industry.

Historically, organizations have used separate disconnected systems to cover each of the three steps in AE case processing. If the systems were connected, they typically necessitated multiple costly sub-systems to manage integrations.

Disconnected systems prevent a singular view of a status within the AE processing cycle. This means that teams who handle different steps aren’t aware of updates or challenges as the case is pushed through transformation and analysis. A person who initially received the case doesn’t know what’s going on with that information after they send it off to be transformed.

In addition, siloed systems create additional steps when developing reports, as multiple sources of information must be reviewed and verified. This requires significant collaboration between multiple team members who handled the case processing and report development and is made that much more complex by the disconnected nature of these complex systems.

AI-driven automation supports and optimizes activities in each of the three steps of AE processing. Data is automatically normalized and reviewed for completeness during information reception when AI-fueled automation technology is utilized from the start of AE processing. This ensures that all data compiled during reception is prepared for transformation. AI optimizes the transformation stage by allowing the immediate extraction and translation data for analysis. This helps safety teams quickly identify any gaps in information that could negatively impact the analysis stage. In the final step, AI streamlines information analysis by providing the ability to quickly and interactively mine data. Because these activities are performed automatically, it provides a “zero touch” approach to Safety, which provides more time for human workers to focus on what is really important.

A key technology to evolve AE case processing is Natural Language Processing (NLP). NLP capabilities allow case information to be translated and normalized for smoother transformation and analysis. This is especially critical for organizations that are operating on a global scale and processing AEs in multiple languages. NLP is a major benefit to safety teams analyzing AEs to ensure nothing is missed and AEs are correctly identified and categorized.

The increase in AE reporting that the pharmaceutical safety industry has seen throughout the pandemic will not subside after the pandemic lessens. In fact, the rate is likely to continue increasing, as citizens now have a heightened awareness of their role in pharmaceutical safety thanks to digital and online tools like social media and online forums. This public awareness will drive citizens to produce more AE information, and challenges are going to get worse in coming years unless pharmaceutical safety teams leverage technology to benefit from the increase in available AE data. The future of life science AE case processing is a zero-touch approach to safety throughout the entire cycle, supported by AI and NLP.