Transforming Bayer's clinical operation with GenAI

RnDAssist - GenAI clinical trial protocol authoring tool

Bayer, in collaboration with Capgemini’s Applied Innovation Exchange (AIE), launched RnDAssist, a GenAI tool, to revolutionize Clinical Study Protocol (CSP) authoring. RnDAssist aims to assist human writers, clinical and medical experts to collaboratively draft the protocols with greater speed, accuracy, and efficiency.

As the backbone of any clinical trial, Clinical Study Protocol is a highly complex, 100+ page document defining every aspect of how a trial will be conducted, from study design to patient eligibility, dosing schedules, and safety monitoring. Traditionally, the creation can take 6 months and is labor-intensive involving numerous stakeholders across various departments, which can slow down clinical operations and even delay the time to market for new treatments.

RnDAssist went live in July 2024, offering intuitive GenAI features designed to significantly reduce the time and effort involved in research and protocol drafting. The tool is already making an impact, being used by Bayer’s medical writing team with live clinical studies, with positive feedback on its usability and quality.

RnDAssist leverages LLM (Large Language Model), RAG (Retrieval Augmented Generation) and AI Agent to enable medical writers and clinical experts to work collaboratively to draft study concept and protocol with efficiency and accuracy. It allows users to conduct literature research with public data, enrich study concept text with historical data and example libraries and conduct consistency check and adhere to regulatory requirements.

Just the beginning

Launched in July 2024, RnDAssist is already used by Bayer’s medical writing team on live clinical studies with positive feedback. The estimated business impact is substantial, with a projected 50% reduction in protocol drafting time and annual cost savings of €1.8 to 4 million through avoidable amendment cost thanks to increased consistency check and validation offered by the tool.

RnDAssist marks the beginning of a broader GenAI-driven transformation at Bayer, with a vision to leverage GenAI to enhance efficiency and innovation across multiple stages of clinical operations. Ultimately, this will contribute to the acceleration of trials and most importantly, the faster delivery of life-saving treatment to patients. Capgemini Applied Innovation Exchange (AIE) is excited to collaborate with Bayer at this pivotal moment.

Highlights

RnDAssist leverages LLMs and RAG to streamline the creation
of Clinical Study Protocol documents.
Internal and external data sources are integrated to ensure accurate, high-quality output.
Advanced prompt builders, consistency checks, and validation against CSP templates all ensure optimum results.
Early user feedback is promising, with usability rating (4.5/5) and quality rating (4.2/5).
It is estimated to reduce CSP drafting times by 50% and avoid amendment cost by €1.8 to €4 million annually.