In the 2020s it is expected that any authorised drug or medicinal therapy designed for human use will be safe. Pharmacovigilance processes are designed to uphold that position over time, once a product has been authorised for patient consumption. Approaches to post-market Safety monitoring have changed little in decades however, despite soaring volumes of available information – submitted in an increasing array of formats, via a proliferating range of channels.

Current approaches don’t take into account, for instance, the actual contents of a case, which means there is no discretion to allocate teams’ bandwidth according to a potential case’s complexity or risk. In particular, adverse event (AE) case intake is costing companies dearly to deliver without being sufficiently timely or accurate to reliably contain emerging safety issues or stave off costly recalls.

Mimicking human discretion

Technology-enabled process automation has long promised to transform the speed, efficiency, and accuracy of AE case intake and triage, by capturing and assessing relevant Safety signals wherever and however these come to light.

Up to now, machine intelligence has not come close to mimicking human powers of data extraction, filtering, inference, or deduction. Early automation systems had to be highly structured and painstakingly trained to recognise every possible format and variant of how important data might show up – from basic information such as the patient’s date of birth to richer detail such as the combination of possible contributors to the adverse event (from the individual’s age and general state of health to the other drugs they may be taking). This compromised the systems’ ability even just to recognise and extract the right data. This in turn limited the potential to transform process efficiency and Safety delivery.

Now, however, Generative AI (GenAI) and Large Language Models (LLMs) are beginning to positively disrupt adverse event data collection with powerful results. In early pilots data extraction accuracy and quality have exceeded 90 percent, and overall efficiency gains related to the intake process have topped 65 percent. And this is all to build on.

Tackling an increasingly unsustainable PV burden in adverse event intake

Advanced automation solutions, which transform the data collection part of the AE case intake process, are being enthusiastically received in an industry desperate for a modern, more efficient way to execute case intake/safety data collection, as volumes of case data soar and pressure mounts to accelerate analysis times.

A recent industry survey[1] shone a light on the biopharma industry’s growing appetite for AI-powered automation, revealing that over 75% of biopharma R&D organisations already use some form of advanced automation within daily processes today, and more than 70% plan to expand business process automation over the next 18 months.

This growing appetite for viable solutions has intensified in line with a maturation of AI-powered process automation technology – from early robotic process automation aligned to regimented processes, toward a less inhibited approach where the technology understands much more about what it is looking for (irrespective of format), and what to do with it.

GenAI technology, using LLMs, can quickly identify and infer what’s relevant and important and reliably summarise key findings for the user – and even extrapolate from them to make predictions about future scenarios. It can do this without the need for protracted ‘training’ and system validation. Rather, specialised applications can now be developed that can apply GenAI-type techniques, contextually, to data they haven’t seen before – learning from and processing the contents on the fly.

This is a huge leap forward that has seen pharma companies start to put GenAI AE intake solutions to the test in their operations, under the watchful eye of their Safety professionals. The ability to simply instruct a system to “Scan X document for Y contents” paves the way to faster, higher-quality extraction of more relevant data, reducing the risk of something significant being missed, and improving downstream efficiency.

Taking the strain, removing unconscious bias

A big part of the business case for harnessing GenAI in AE case intake management comes from the scope for handling first-line capture and processing of very high volumes of data – freeing up more of Safety professionals’ time to analyse the findings.

On top of this, GenAI removes human limitations such as fatigue, mental overload, distraction, data blindness, and unconscious bias. An AI-powered tool can draw on the findings of millions of prior cases and assessments, to make credible predictions and unbiased assessments regarding causality (the likelihood of a direct link between a product and a reported adverse event), based on probability.

The early output is proving very encouraging, and confidence in the technology is growing rapidly as a result. It helps that the links back to the sources are readily traceable for checking.

Extending the transformation potential

Beyond AE case intake, there will be other powerful use cases across Life Sciences R&D Safety and Regulatory operations, so it is a good idea to allow for additional applications in the future, for instance by deploying an enabling ‘platform’ rather than a single-purpose application.

Strong next contenders for GenAI/LLM treatment include real-time pharmacovigilance assessments and associated decision-making (e.g. the earlier identification of unexpected benefits/discovery of new indications); harnessing international Regulatory intelligence to transform marketing authorisation applications and maintenance; and clinical trial modeling, reducing the reliance on traditional clinical studies.

A recommended first step toward advanced automation would be to break down how processes are currently managed, the core requirements driving those processes, and where the pain points are. The next priority should be to review and rewrite standard operation procedures so that they can evolve with and be improved by advanced technology, both now and in the coming years.

About the author

Emmanuel Belabe is Senior Vice-President for Customer Success, Global Customer Support, and Solution Consulting at ArisGlobal. He is an experienced Safety director with a strong record in applying the latest IT innovations in healthcare and life sciences.

References

[1] ArisGlobal’s 2024 Industry Survey Report, Life Sciences R&D Transformation: Ambitions for Intelligent Automation & Today’s Reality,

The post Generative AI in Action in Life Sciences: Transforming Adverse Event Case Intake appeared first on .

Biopharma is an ideal target for Generative AI (GenAI)-based process transformation, with its highly regulated, templated repeat activities including medical writing in the context of license application; lifecycle maintenance; and ongoing post-market clinical and safety report generation. Even if GenAI-based applications could distil the right content to craft a first draft of these typically hefty documents (for attentive checking and amendment by skilled humans), that intervention alone would go a long way in alleviating the strain on overwrought regulatory professionals and adjacent teams.

Yet, even with the latest evolution of large language models (deep learning models trained on vast amounts of data), building a level of system ‘knowledge’ requires nuance and tailoring to be of trusted value. Even as AI models progress, they need to be guided in how to repurpose information and data correctly, with several examples of what ‘good’ looks like.

To yield tangible improvements to efficiency, AI skills need to be combined with life science industry fluency in specialist language and vocabularies, required templates and the nuanced demands of each market. That includes in-depth knowledge of what will be accepted by regulatory agencies and how that differs from region to region, and country to country; and a strong feel for specific medical writing best practices linked to each use case.

It takes vast reams of successful example content to fine-tune a GenAI model on what’s the best way to repurpose or analyse content, meanwhile – and the ideal output to aim for. For pharma companies to bring their own capabilities up to speed, and stay ahead, seems an impossible feat.

A pronounced capability gap

A survey conducted recently with the Regulatory Affairs Professionals Society (RAPS)[1] highlighted medical writing as a critical area requiring support, due to growing pressure on Regulatory professionals’ time. Over half (57 percent) of surveyed companies were planning to invest in technology to improve medical writing over the year ahead, almost on a par with eCTD v4.0 spending, the two categories dominating immediate Regulatory IT spending plans.

Clinical study protocol/report writing and drafting of regulatory documents were the primary medical writing needs identified by Regulatory professionals, requiring additional support. More than half of respondents in the 2024 study identified a need to harness AI in data extraction (56 percent) and information summarisation (53 percent), where just 9-10 percent are using AI for those purposes today – specifically within the context of medical writing. Twelve percent said they were actively in the process of incorporating AI into automated report generation from multiple sources, which is the ultimate opportunity on offer.

Despite these ambitions, more than half (53 percent) of survey respondents in pharma conceded that their organisations did not possess sufficient knowledge internally to implement AI technologies themselves.

Compliance expectations

Other potential barriers to AI uptake revolve around confidentiality and the scope for inadvertent breaches, linked in part to how the technology is ‘prompted’ to call up information. Because the technology is so complex below surface level, there are concerns that applying public cloud-based GenAI models could compromise internal data security.

All of these concerns are readily addressable in an appropriate ‘closed’ processing environment, which can be purposefully tailored to life sciences use cases. This ensures that all data interactions remain within a secure and controlled ‘space’, for data confidentiality and integrity purposes.

Other considerations include traceability and auditability: the ability to see where extracted data has come from or from which source content summarisation has been achieved (for instance, clear links in the final report to original documents). This is essential to build confidence and trust in solutions, so that they are seen to add significant value and save time; over-reliance on painstaking checks could undermine the return on investment.

Knowing that there are solutions that can be validated for priority pharma medical writing use cases will be an important facilitator for companies with a growing need for smarter support, as medical writing workloads continue to soar rather than diminish over time.

A wealth of potential benefits, including leaner & cleaner output

In looking for the value from an effective (and ideally fully-managed) GenAI medical writing automation capability, pharma companies need to look not just at the efficiency gains associated with smart data extraction, information summarisation, and narrative authoring – albeit that these are promising areas to start focusing on.

Further gains will come from improvements to consistency, and to leaner, tighter output once specified regulatory documents are being drafted according to training (from extensive exposure to approved documents) on what ‘good’ looks like.

Investing in R&D or partnership with external technology solution (and service) providers on a variety of use cases promises to make better use of scientific experts’ time, as use of their time shifts from initial drafting to the strategic thinking in collaboration with clinical development professionals.  In a more strategic context, in early clinical evaluation, Gen-AI based tools could help summarise the vast wealth of existing information on the internet to hone the focus of planned research (such as alternative therapies) in order to avoid wasted time investment. AI-assisted search across multiple sources can also summarise headline findings, while also highlighting what emerges as the best path to follow.

An urgency to act on GenAI in medical writing

Although biopharma companies themselves may lack sufficient R&D resources internally to play around with possibilities, experimentation is a powerful way forward in determining where GenAI offers maximum value in transforming medical writing. The danger otherwise is of falling behind as innovators like Moderna (which already has more than 750 active use cases of AI to improve process efficiency[2]) continue to blaze a new trail.

It will be important, then, to find a way to at least co-design a pilot solution (for example with an appropriate technology or service partner), that can be tested with example documents – to determine how powerful and accurate the technology can be and how quickly it learns and adapts, to hone its output. By trying out the technology, companies will get more of a feel for what’s possible, and the scale of difference this could make to everyday Regulatory workflow.

About the author

Punya Abbhi is Chief Operating Officer & Co-Founder at Celegence. She was previously Life Sciences Consultant at Capgemini Consulting and before that spent time at Kates Kesler Organization Consulting and Gartner, and working for a specialist life sciences service provider. Punya holds an MBA from INSEAD and a BA in Politics, Philosophy, Economics (PPE) with a minor in Cognitive Science from the University of Pennsylvania.

References

[1] Pharmaceutical Industry Regulatory Readiness & Resources 2024 Survey Report, RAPS-Celegence – a survey of 139 pharma companies headquartered chiefly in the US, Europe or Asia-Pac regions

[2] https://investors.modernatx.com/news/news-details/2024/Moderna-and-OpenAI-Collaborate-To-Advance-mRNA-Medicine/default.aspx

The post Could GenAI Revolutionise Pharma Medical Writing? appeared first on .