Digital support for data-driven recruitment planning of clinical trials

Clinical studies are essential for the research and development of new therapies and medical engineering devices. Efficient implementation benefits employees, companies and patients. However, almost half of all clinical studies fail due to recruitment difficulties. Complex inclusion and exclusion criteria are another common reason for failure. This results in delays in approvals and high financial losses for the sponsors.

Currently, clinical studies are often organized using Excel lists and phone calls between the organization conducting the study (often a clinical research organization, or CRO) and the study centers (often hospitals). During this process, employees manually review the inclusion and exclusion criteria and compare them with the patients' data. This process is time-consuming and prone to error.

 

AI-based decision support for more efficient planning and execution of clinical studies

The aim of the project is to develop an AI-based decision support system for the implementation of clinical studies. During the development phase, research centers, research-based pharmaceutical companies and infrastructures for conducting clinical studies that are interested in improving the recruitment process should be addressed at an early stage. This is because more efficient planning and execution of clinical studies saves money and time and prevents study discontinuations due to insufficient numbers of participants.

To this end, we at Fraunhofer CIMD are combining our expertise in the section of artificial intelligence, in particular generative AI, natural language processing (NLP) and large language models (LLMs), and supplementing this with extensive experience in conducting and analyzing clinical studies.

 

Outlook

In hospitals, medical and administrative data is collected, processed and shared in the hospital information system (HIS). Integrating the developed system for AI-based decision making could further simplify its use in everyday clinical practice. In addition, it could be extended to other indications.