ľ¹Ï¸£ÀûÓ°ÊÓ

Tabea Kaul’s research explores how we can better combine and evaluate medical research to support evidence-based medicine. Systematic reviews are important tools in medicine. They help doctors and researchers understand what all the existing studies say about a certain topic. Systematic reviews summarize existing studies, but they often face challenges like missing or biased data. This thesis focuses on improving how we find, assess, and report medical studies.

The thesis shows that searching clinical trial registries can uncover more relevant studies than traditional databases, but both sources are still needed. It also highlights that many clinical trial records do not report their results properly in clinical trials registries, making it harder to use them in reviews. We offer practical recommendations to improve their transparency and usability. Further, this thesis identifies tools used to judge the quality of medical prediction studies and proposes guidance on how to best choose such a tool. Moreover, this thesis reviews a tool to assess prediction model studies and proposes improvements. A new tool, PROBAST+AI, was developed to assess studies using modern techniques like artificial intelligence (AI). Additionally, a reporting adherence tool was updated to help monitor how studies adhere to the reporting checklist TRIPOD+AI, a newly developed checklist for prediction model studies using modern techniques like AI. In short, this thesis provides practical tools and guidance to improve how medical research is collected, assessed, and reported—making healthcare decisions more reliable and transparent.