Image

Pharmaceutical Market Europe • March 2025 • 18

THOUGHT LEADER

Harnessing AI and big data in clinical research

By Sheila Kelly

Image

Technological innovations and digital disruptions have long shaped clinical research. The industry continuously seeks to push boundaries for faster drug development, but patient safety, data integrity and regulatory compliance remain paramount.

ICON’s latest survey and white paper, Digital Disruption: Surveying the Industry’s Evolving Landscape, builds upon a 2019 study. Then, blockchain and organ-on-a-chip generated industry buzz. Today, AI enables advancements like remote monitoring and big data analytics. This survey provides insights into attitudes and challenges in digital innovation across the clinical research industry.

ICON’s survey findings

  • 49% of respondents use AI and big data analytics in drug development
  • Organisations piloting these technologies increased from 29% in 2019 to 35% in 2024
  • Full-scale digital adoption remains limited, with only 13% reporting a comprehensive AI-driven programme – just a 1% increase from 2019
  • In 2019, 29% believed digital technologies would have no impact on operations; by 2024, this dropped to 11%
  • 33% view digital tools as key to breaking down silos and reorganising functions, up from 21% in 2019
  • 77% of respondents expect digital transformation to drive mid-to-high double-digit improvements in R&D productivity
  • AI solutions in development include predictive modelling, biomarker discovery, early detection of safety and efficacy signals, preclinical drug candidate assessment, patient recruitment and protocol optimisation.

Key findings – industry optimism

The survey highlights a 5% increase in sponsors’ optimism about AI and digital technology’s impact on R&D productivity since 2019. However, tempered expectations reflect implementation challenges.

AI adoption has progressed but remains slower than anticipated. The survey shows 49% of respondents indicating their organisations use AI and big data analytics, a 10% increase.

Additionally, organisations piloting these technologies grew from 29% in 2019 to 35% in 2024. Given the typical innovation adoption cycle, this progress is notable.

Barriers to full-scale implementation persist. Only 13% of organisations have fully integrated AI into clinical development, showing scalability – ensuring digital tools work across trial designs and therapeutic areas – is an obstacle. 88% of respondents plan to increase investment in digital technologies, yet many remain cautious due to regulatory uncertainty, high licensing costs, cybersecurity concerns and interoperability challenges.

How AI is improving clinical trials

Clinical trials are one of the longest, riskiest and most expensive phase of drug development.

ICON’s AI solutions are transforming our customers’ clinical trials by enhancing decision-making, accelerating timelines, forecasting risks, optimising resources and improving operational efficiency.

  • OneSearch leverages real-world data to identify optimal trial locations and improve trial designs by modelling diseases more effectively
  • iSubmit uses AI to simplify the eTMF document management with automated filing and compliance checks, reducing the burden on clinical project teams
  • FORWARD+ enhances visibility into resource demands, forecasting and allocating resources more efficiently
  • Mapi Research Trust provides comprehensive intelligence to support optimal clinical trial protocols/designs, leveraging AI to remain current with the latest clinical outcome assessments from public sources in near real time.
  • Study Start-up Site Contracts streamlines clinical contract drafting by analysing historical clinical contracts and creating comprehensive near-final draft contracts
  • OMR AI Navigation Assistant uses generative AI to provide industry-leading analytics to transform data into business insights
  • Cassandra predicts the needs for FDA and EMA post-marketing studies
  • ICONex enables key opinion leaders and their networks to be more easily identified.

The future of AI in pharma

ICON’s survey confirms a shift from AI hype to practical application, with organisations now integrating AI into digital ecosystems and prioritising interoperability and sustainability.

For example, developing a drug nowadays takes eight to ten years and costs over $2bn, yet less than 10% of clinical candidates gain regulatory approval. AI can make drug discovery and development more cost-effective, rapidly analysing vast data sets to identify promising drug candidates, fine-tune molecular structures and generate novel compounds.

Choosing the right AI partner

Sponsors should conduct thorough due diligence when selecting AI partners. Many AI solutions were developed as ‘solutions without a cause’, meaning their value propositions may not align with pharmaceutical research objectives. These tools need to be tailored for clinical research and seamlessly integrated across the complex clinical trial process, while maintaining data integrity.

Organisations should prioritise partners with expertise in clinical development, strong data security protocols and proven interoperability.

Conclusion

AI and digital technologies in clinical research continue to progress. While full-scale implementation remains limited, growing investment and shifting attitudes indicate strong commitment to digital transformation. As AI applications become more sophisticated and better integrated, they will enhance R&D productivity, improve trial efficiency and reduce costs.


Sheila Kelly is Vice President, IT Strategic Programs at ICON