生成式人工智能时代的临床决策

Clinical Decision-Making inthe Era of Generative AI Emerging behavioral signals from European IQVIA Data Strategy and Architecture (DS&A) Table of contents Executive summary: The GenAI frontier in clinical cognition1Introduction2Survey findings: Adoption patterns and emerging behavioral signals3How healthcare professionals are using generative AI5Trust, challenges, and institutional readiness9Implications for commercial and medical strategy in a GenAI-enabled environment13Closing perspective14Survey design, scope, and considerations for interpretation14About IQVIA Data Strategy and Architecture (DS&A)16References17 Executive summary: The GenAI frontier in clinical cognition Generative AI (GenAI) is no longer a peripheralexperiment. It is fundamentally reconfiguring clinicaldecision-making by decentralizing how evidence issynthesized and applied. Rather than awaiting formalinstitutional deployment, healthcare professionals Key findings Strategic implications Medical Affairs:Medical Affairs must transition from a“sole source of truth” to a strategic partner in an AI-mediated environment. Scientific communication must Widespread adoption:85% ofsurveyed HCPs utilize GenAI forwork, peaking at 92% among Commercial and sales operations:GenAI accelerates theestablishment of the “pull” dynamic where HCPs activelylook for information and veryify claims in real-time via High-frequency use:Integrationis deeply habitual; 41% of HCPs useGenAI daily, and 73% engage at Moving from observation to integration The shift toward distributed individual usage requires are-engineering of engagement models. Organizationsmust now codify governance structures to reflect a Geographic variation:Franceleads adoption at 91%, followed by About the research As a leader in healthcare data and analytics, IQVIAprovides the foundational insights and frameworksnecessary to navigate this transition. This study is part Clinical utility:Usage is driven byimmediate needs, including clinical on previous IQVIA research examining how healthcareprofessionals use artificial intelligence tools to accessand interpret scientific information, highlightingthe growing role of generative AI within clinical Introduction Healthcare systems across Europe are undergoingstructural transformation in how clinical informationis accessed, interpreted and applied.¹ While formal The findings suggest that GenAI use among healthcareprofessionals is neither speculative nor peripheral.Instead, it represents a meaningful evolution in clinical Generative artificial intelligence tools have movedrapidly from novelty to routine utility.2What beganas exploratory experimentation has, in many cases,become embedded in the cognitive processes At the same time, institutional enablement has notkept pace with individual behavior. Most usageremains self-initiated and frequently accessed For pharmaceutical and life sciences organizations,this development raises important questions. Howwidespread is GenAI adoption among healthcareprofessionals? What specific tasks are these tools This white paper explores the implications of thesefindings. It examines how GenAI is currently beingused in clinical contexts, where trust and validation To better understand this emerging dynamic, IQVIAconducted a multi-country survey of oncologists andgeneral physicians across Germany, France and the Survey findings: Adoption patterns and emerging Adoption patterns and access models usage together represent the dominant pattern,indicating that HCPs are returning to these tools GenAI adoption has reached widespread Survey findings indicate that generativeAItools arealready integrated into the professional workflows ofa large majority of healthcare professionals. Overall, This pattern holds across both HCP specialties includedin the survey. While intensity may vary slightlybetween markets, no country reflects low structural Frequency of use further reinforces the maturity ofadoption. More than four in ten respondents report dailyuse, with a substantial share engagement on a weekly This reinforces the view that generative AI has crossedthe threshold from curiosity-driven experimentation to Adoption patterns are broadly consistent acrossspecialties, including general practitioners andoncologists. While there are variations across General-purpose GenAI tools dominate Survey responses show that healthcare professionalsoverwhelmingly rely on general-purpose generativeAI platforms rather than healthcare-specific solutions.ChatGPT clearly dominates usage across countries Usage frequency reflects operational, not The distribution of usage frequency suggests thatGenAI is functioning as a regular support tool rather as secondary tools, while dedicated healthcare-focused AI solutions are used only by a small minorityof respondents. This pattern indicates that adoptionis currently driven by accessibility, familiarity This imbalance between individual usage andinstitutional enablement