Embracing Innovation: Integrating AI (ChatGPT) in Clinical Trials Programming Workflow

The usage of ChatGPT in clinical trials and other generative AI models are on the rise, and in the constantly evolving pharmaceutical industry, efficiency and productivity remain paramount. During the PHUSE US Connect 2024 conference, Senior Statistical Programmer Mykyta Vysotskyi brought forth a few practical ideas of integration of artificial intelligence (AI), specifically ChatGPT, into the clinical programming world.

The presentation, titled “Navigating the Safe Integration of AI (ChatGPT) into Clinical Trials Programming Workflow,” challenges the current thoughts for a wholesale replacement of current statistical programming practices with AI.

Instead, Vysotskyi advocates for a more pragmatic approach to marry the immediate benefits of generative AI with existing tools and expertise within the pharmaceutical industry.

Routine tasks can be made easier with AI

At the conference, Vysotskyi shared practical strategies for effortlessly incorporating AI, particularly ChatGPT, into routine tasks performed by clinical trial programmers, like creating common reports.

Starting with simple everyday tasks is seen as an ideal first step to enhance efficiency without sacrificing the depth of analysis or quality of results.

Central to Vysotskyi’s discourse is the notion of leveraging AI as a complementary tool rather than a substitute. ChatGPT’s potential within the framework of established methodologies should be embraced by statistical programmers with both enthusiasm and caution as the industry looks for new ways to speed up drug development.

Enhance without replacement

Statistical programmers are under a magnifying glass to find ways to incorporate AI into their current workflow, just as so many in the pharmaceutical industry are.

There are significant staffing challenges for an industry that is expected to grow by seven percent through 2028, according to an article in Pharmaceutical Executive Magazine. Embracing ChatGPT in clinical trials has the potential to allow today’s statistical programmers to do more with less.