Making ChatGPT better for clinicians

Making ChatGPT better for clinicians

Summary (EN)

OpenAI announced ChatGPT for Clinicians, a specialized version of ChatGPT intended to support documentation, evidence review, and other common clinical tasks, and said it will be offered free to verified U.S. physicians, nurse practitioners, physician assistants, and pharmacists. The company frames the launch as a response to growing clinician adoption of AI tools and mounting pressure on healthcare workers facing heavier administrative workloads and an expanding medical literature base. According to the announcement, clinician use of ChatGPT has more than doubled over the past year, and the product is designed around repeatable workflows such as referral letters, prior authorization, patient instructions, and medical research. OpenAI says the offering includes frontier models for clinical questions, trusted search across peer reviewed medical sources, deep research capabilities for literature reviews, reusable workflow skills, and optional HIPAA support through eligible accounts. The company also emphasizes privacy and account security, including the statement that conversations are not used to train models. OpenAI says physician advisors have reviewed more than 700,000 model responses related to health use cases, and that 6,924 conversations were tested before release. In those internal evaluations, physicians reportedly rated 99.6% of responses as safe and accurate. OpenAI positions the product as decision support rather than a substitute for clinical judgment, while also using the launch to advance a broader healthcare strategy that includes enterprise deployment, compliance controls, and clinician specific evaluation benchmarks.

Summary (ZH)

OpenAI 发布了面向临床场景的 ChatGPT for Clinicians，定位为帮助医护专业人士处理文书、检索证据和完成常见临床辅助任务的专用版本，并宣布将向美国经过认证的医生、执业护士、医师助理和药师免费开放。官方将此次发布定义为对医疗系统现实压力的回应，即临床人员一方面需要应对不断增加的行政事务，另一方面还要跟进快速扩张的医学研究信息。OpenAI 表示，过去一年中临床人员对 ChatGPT 的使用量已增长超过一倍，新产品围绕转诊信、预授权、患者说明和医学研究等可重复流程进行设计。功能上，ChatGPT for Clinicians 包含用于复杂临床问题的前沿模型、基于同行评审医学资料的可信搜索、用于文献综述的深度研究能力、可复用的工作流技能，以及针对合规需求提供的可选 HIPAA 支持。公司同时强调隐私与账号安全，明确表示相关对话不会用于模型训练。OpenAI 还称，其医生顾问团队已累计审阅超过 70 万条与健康应用相关的模型回复，并在发布前测试了 6,924 段真实工作场景对话。在其内部评估中，医生将 99.6% 的回答评为安全且准确。整体来看，这一发布既是面向一线临床工作的产品落地，也是 OpenAI 将医疗行业作为重点应用领域持续推进的一部分。

Source

https://openai.com/index/making-chatgpt-better-for-clinicians/