AI, Open Source, and the Future of Biostatistics
Driven by the integration of artificial intelligence (AI), the rise of open-source tools, and evolving data standards, statistical programming is experiencing some exciting innovations.
At the PharmaSUG 2025 conference, industry leaders and professionals shared insights on how these innovations are shifting the approach to statistical programming, data analysis, and regulatory submissions. In addition to the several attendees sent for continuing education, Catalyst employees had three presentations accepted:
How companies are implementing AI
AI is actively being integrated into clinical workflows. Presentations such as Igor Goldfarb’s “Application of Advanced GenAI Tools in Sample Size Estimation” and Kevin Lee’s “AI is Coming for You: New Biometric Leadership in the Era of Gen AI” highlighted both the promise and the pitfalls of AI in biostatistics.
One speaker likened ChatGPT to a “smart and capable intern”—useful for initiating tasks but requiring human oversight for validation and accountability. For example, while GenAI can assist in sample size estimation, its outputs often lack reproducibility, a critical requirement in scientific research.
Benefits and risks of AI adoption
AI offers significant advantages, including automation of repetitive tasks, enhanced data processing, and improved efficiency. However, unsupervised use in clinical settings can pose risks, especially when statistical accuracy and regulatory compliance are at stake. The consensus: AI should augment, not replace, human expertise.
Python and R gaining ground
There’s a noticeable industry move toward open-source programming languages like Python and R. These tools offer flexibility, cost-effectiveness, and a vibrant community of contributors. Presentations at PharmaSUG emphasized how open-source tools are being used for statistical programming tasks, data visualization, and automation.
One standout example involved using SAS to automate the annotation of case report forms (CRFs), reducing a 40-hour task to just a few hours. This kind of innovation demonstrates the power of combining traditional tools with modern programming techniques.
Despite the rise of open source, SAS remains a cornerstone in clinical research due to its validated environment and regulatory acceptance. Conversations with SAS representatives revealed how the platform is evolving to integrate Python and R, ensuring it remains competitive while maintaining its strengths in compliance and reliability.
CDISC, SDTM, and ADaM in focus
Data standards continue to be a critical area of focus. Presentations on topics like “Cytokine Release Syndrome (CRS): Data Collection, Clinical Database Integration, and Analyses” and “ADaM Fundamental Principles vs. Rules” provided practical guidance on navigating complex data scenarios.
These sessions emphasized the importance of aligning with CDISC standards to ensure data integrity and regulatory readiness. Real-world examples illustrated how to balance flexibility with standardization in ADaM dataset development.
Innovation through imagination
One of the most inspiring takeaways from PharmaSUG was the creativity of programmers in solving complex problems. As one attendee noted, “There’s one end goal, but countless ways to get there. Your only limitation is your imagination.”
This spirit of innovation is driving the development of new tools and processes, pushing the boundaries of what’s possible in clinical programming.
To remain relevant, professionals must stay attuned to industry trends and continuously expand their skill sets. Cross-platform collaboration, ongoing training, and strategic adoption of AI are essential for fostering innovation and maintaining high standards in clinical research.
The convergence of AI, open-source tools, and evolving data standards marks a new era in clinical programming. While challenges remain, the opportunities for innovation, efficiency, and improved outcomes are greater than ever. With a strong foundation in programming and a willingness to embrace change, clinical programmers are well-positioned to lead the industry into the future.
Gabrielle Legaspi, Rachna Patel, and Jake Gallagher contributed to this post.
