RESUMO
The pharmaceutical industry is undergoing a sweeping transformation, driven by technological innovations, demographic shifts, regulatory changes and consumer expectations. For adaptive players in pharma to excel in this rapidly changing landscape, which will be markedly different from today by 2030 and beyond, they will require a different set of skills, capabilities and mindsets, as well as a willingness to collaborate and co-create value with multiple stakeholders. The industry needs to rewrite the textbook for pharma by embracing and implementing four key dimensions of change: digitalization, personalization, collaboration and innovation. In this article, we will examine how these dimensions of change are reshaping the industry, and provide practical and strategic guidance based on best practices and examples. Specifically, adaptive pharma companies should embrace the use of advanced digital technologies, such as artificial intelligence and machine learning, to streamline processes and solve challenges rapidly. Personalization, both in medicine and patient engagement, will also be key to success in the 'digital revolution', and a collaborative approach involving partnerships with tech start-ups, health-care providers and regulatory bodies will also be essential to create an integrated and responsive health-care ecosystem. Using these ideas for a rewritten textbook for pharma, adaptive players in pharma will evolve to be personalized and digitized health-focused organizations that provide comprehensive solutions which go beyond drugs and devices.
RESUMO
Objectives: 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. Materials and Methods: 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. Results: 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. Discussion: We discuss areas for improvement, including better ingestion of tables, addition of context and fine-tuning. Conclusion: 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.