AI-driven speech-to-text (STT) documentation systems are increasingly adopted in clinical settings to reduce documentation burden and improve workflow efficiency. However, their rapid deployment has outpaced understanding of the associated socio-technical risks, including transparency, reliability, patient autonomy, workflow alignment, and organizational governance. A clearer analysis of these risks is needed to support safe and equitable integration into healthcare practice. This study synthesizes interdisciplinary evidence from technical performance research, regulatory and ethical standards, clinical workflow analyses, and organizational policy guidance. The synthesis was used to develop a multi-layered socio-technical conceptual framework for evaluating and governing STT systems. Findings show that STT systems operate within tightly coupled socio-technical environments in which model performance, clinician oversight, patient rights, workflow design, and institutional governance are interdependent. The study offers a structured socio-technical governance framework and an implementation roadmap that outlines readiness assessment, vendor evaluation, pilot deployment, clinician training, ongoing monitoring, and iterative improvement. The framework emphasizes safeguards that protect patient autonomy, documentation integrity, and institutional trust while enabling the efficient and beneficial use of STT technologies. This work provides actionable guidance for healthcare organizations seeking to adopt STT systems responsibly and equitably.

翻译：人工智能驱动的语音转文本（STT）文档记录系统在临床环境中日益普及，旨在减轻文档记录负担并提升工作流程效率。然而，其快速部署已超越了对相关社会技术风险的理解，包括透明度、可靠性、患者自主权、工作流协调与组织治理等方面。需要更清晰地分析这些风险，以支持其在医疗实践中的安全与公平整合。本研究综合了来自技术性能研究、监管与伦理标准、临床工作流分析以及组织政策指导的跨学科证据，并以此构建了一个用于评估和治理STT系统的多层次社会技术概念框架。研究发现表明，STT系统在紧密耦合的社会技术环境中运行，其中模型性能、临床医生监督、患者权利、工作流设计与机构治理相互依存。本研究提出了一个结构化的社会技术治理框架及实施路线图，涵盖就绪度评估、供应商评价、试点部署、临床医生培训、持续监测与迭代改进等环节。该框架强调建立保护患者自主权、文档完整性与机构信任的保障措施，同时促进STT技术的高效与有益应用。本工作为寻求负责任且公平采用STT系统的医疗机构提供了可操作的指导。