Journal of Medical Economics, Journal Year: 2025, Volume and Issue: unknown, P. 1 - 19
Published: April 1, 2025
The rapid evolution of large language models (LLMs) and machine learning (ML) presents both significant opportunities challenges for market access processes. These sophisticated AI systems, built on transformer architectures extensive datasets, offer potential to forecast claims decisions health technology assessment (HTA) agencies streamline processes such as systematic literature reviews HTA submissions. Furthermore, the analysis real-world data - also deriving causal relationships is being discussed intensively. Despite notable advancements, their adoption in key PRMA still limited at present, with only a small fraction submissions bodies incorporating AI. Key barriers include stringent transparency requirements, necessity explainability human oversight analyses, highly sensitive nature text drafting especially cases where reimbursement or pricing negotiations balance knife's edge. requirements are often not met due immaturity many applications, which lack necessary precision, reliability, contextual understanding. Moreover, AI-generated evidence has yet prove its validity before it can supplement replace traditional study designs, randomized controlled trials (RCTs), critical decisions. Additionally, environmental financial costs training LLMs require careful assessment. This paper explores various current limitations, future prospects from German perspective while considering broader implications EU Health Technology Assessment Regulation (HTAR). It concludes that hold transformative potential, integration into workflows must be approached cautiously, incremental adoption, close collaboration between industry, agencies, academia. Demonstrating robust, unbiased comparative evidence-showcasing superior performance cost savings over methods-could accelerate process.
Language: Английский