Beyond the Hype: Transformative Impact of Large Language Models on Clinical Decision-Making and Healthcare Delivery Systems

Abstract

Large language models (LLMs) are rapidly transforming healthcare delivery through unprecedented capabilities in clinical text processing and knowledge synthesis. This research examines the multifaceted impact of LLMs across healthcare settings through a systematic mixed-methods investigation. We analyzed 218 peer-reviewed studies published between 2022-2025 utilizing a framework analysis approach supplemented by expert consensus development via modified Delphi technique. Our findings reveal differentiated adoption patterns across healthcare domains, with administrative applications (documentation, coding) showing 72% greater implementation than direct clinical decision support. Performance metrics demonstrate 31-47% efficiency improvements in documentation workflows, 26% reduction in clinician cognitive load, and 18-29% enhancement in diagnostic accuracy across multiple specialties when used as augmentation tools. However, significant implementation barriers persist, including data privacy concerns, workflow integration challenges, and algorithmic bias. This research establishes an evidence-based framework for healthcare organizations navigating LLM integration while highlighting critical governance mechanisms necessary to balance innovation with patient safety. The findings suggest targeted application areas offering maximum clinical value while minimizing risks in this rapidly evolving technological landscape.

Ähnliche Arbeiten