{Reference Type}: Case Reports
{Title}: Using large language models for safety-related table summarization in clinical study reports.
{Author}: Landman R;Healey SP;Loprinzo V;Kochendoerfer U;Winnier AR;Henstock PV;Lin W;Chen A;Rajendran A;Penshanwar S;Khan S;Madhavan S;
{Journal}: JAMIA Open
{Volume}: 7
{Issue}: 2
{Year}: 2024 Jul
暂无{DOI}: 10.1093/jamiaopen/ooae043
{Abstract}: <B>UNASSIGNED: </B>The generation of structured documents for clinical trials is a promising application of large language models (LLMs). We share opportunities, insights, and challenges from a competitive challenge that used LLMs for automating clinical trial documentation.<BR><B>UNASSIGNED: </B>As part of a challenge initiated by Pfizer (organizer), several teams (participant) created a pilot for generating summaries of safety tables for clinical study reports (CSRs). Our evaluation framework used automated metrics and expert reviews to assess the quality of AI-generated documents.<BR><B>UNASSIGNED: </B>The comparative analysis revealed differences in performance across solutions, particularly in factual accuracy and lean writing. Most participants employed prompt engineering with generative pre-trained transformer (GPT) models.<BR><B>UNASSIGNED: </B>We discuss areas for improvement, including better ingestion of tables, addition of context and fine-tuning.<BR><B>UNASSIGNED: </B>The challenge results demonstrate the potential of LLMs in automating table summarization in CSRs while also revealing the importance of human involvement and continued research to optimize this technology.