The Future of Regulatory Automation: Implementation Considerations for AI in Regulatory Operations
Shutterstock

The Future of Regulatory Automation: Implementation Considerations for AI in Regulatory Operations

New tools using new artificial intelligence (AI) are?almost ready for prime time?and despite the industry’s initial reluctance to embrace it, regulatory teams are being inundated with new, AI-assisted tools for improving regulatory compliance. But with a machine-first mindset comes the development team’s commitment to real-time evaluation of data.?AI and machine learning tools?are poised to bring better outcomes to patients, payers and stakeholders.

Despite an initially low uptake, some major pharmaceutical companies have already invested in?AI tools?to automate the analysis of regulatory documents. These programs range from simple automation tasks to the use of AI technology leading a fully supervised review system. These applications can streamline various processes, improve efficiency and accuracy, and provide valuable insight and analysis, including:

  • Automated Study Protocols:?The system can automatically populate the study protocols with text from the Investigator's Brochure and protocol concept documents.
  • Hyperlinked New Drug Applications:?Automation can create hyperlinks throughout the document, making it easier to navigate and access relevant information.
  • Clinical Study Report (CSR) Safety Narratives: A time-consuming process can be simplified through automation, especially when the data is large or commonly populated. A hybrid approach with medical writer oversight can provide both efficiency and medical usefulness.
  • Structured Authoring:?This approach allows writers to focus on content and creates a set of rules to standardize document organization.
  • Clinical and Scientific Perspective:?Automated systems can generate clinical and scientific information, but expert assessment is necessary to ensure correct scientific judgment is applied.
  • Automated Review Tools:?These tools streamline the tedious process of sending out sequential versions for review and allow for simultaneous commenting and tracking.
  • Translation and Localization:?The use of AI/ML for writing in another language is especially useful for lay summaries required by regulatory agencies such as the EMA.
  • Robotic Process Automation (RPA) Cases:?Automation can assist with tasks such as data extraction from Excel, data upload to servers, archiving, document import, data consistency and quality checks, bookmarking, and data arrangement.

Read our full blog to learn about key considerations for implementing regulatory automation to ensure success.

No alt text provided for this image
The Syneos Health Insights Hub generates future-focused, actionable insights to help biopharmaceutical companies better execute and succeed in a constantly evolving environment. Driven by dynamic research, our perspectives are informed by our insights-driven product development model and crafted by subject matter experts focused on real answers to customer challenges to help guide decision making and investment.?To learn more, visit?insightshub.health.
Yogitha.S. Arun (yogi)

Clinical trials professional

1 年

I am still skeptical , as it is we are trying to go paperless at clinical trial sites which means control is handed over to IT department, at Phase 1 , where the regulations needs to get stricter , if You condition AI properly you will end up in a situation where all the private funded sectors can get through the loop holes and leave the federally funded genuine public health research penniless at the mercy of partnering with big pharmas. This is not good for federal funded projects trying get a product out on market. Atleast it’s not going to work in Australia where phase 1 regulations need human intelligence not automation , may work on phase 2 or 3, not on phase 1 after HERC approval the tap is in phase 1 and so is the control on which pharma products can go to phase 2 , and Australia’s ???? brand is Safety first . Not for Australia for phase 1 , may work in America , because we don’t have FDA here to rely on safety, wherever there interest lies. Thankyou ??

回复
Milica Dragi?evi?

Localisation specialist | IKEA

1 年

I'd argue that expert assessment is also necessary for the translation of lay summaries. It is important to underline that LLMs are biased toward English and high-resource language pairs. Moreover, the task of writing or translating lay summaries is complex and nuanced, as it involves adapting content to suit the cultural context of the target audience. This means that the information contained in lay summaries can vary greatly from culture to culture.

Michael Mendes

Project Manager at ICON Strategic Solutions

1 年

AI is a power tool but must never replace human intelligence.

AI and Machine learning is the Future

It's very interesting!

要查看或添加评论,请登录

Syneos Health的更多文章

社区洞察

其他会员也浏览了