The extraction of structured clinical information from unstructured EHR notes is a persistent bottleneck in healthcare informatics. While large language models (LLMs) offer high performance, their deployment in clinical settings is hindered by privacy risks, inference costs, and the tendency to hallucinate beyond textual evidence. We address these challenges for the CL4Health 2026 Case Report Form (CRF) filling task by proposing a fully local, domain-adapted pipeline using the MedGemma-27B model. Our two-stage architecture, which separates binary presence classification from value extraction, enforces strict adherence to textual evidence and ensures deterministic outputs for negated, uncertain, or unknown states. By leveraging item-specific, few-shot in-context learning without external API calls or fine-tuning, our approach achieves a macro-F1 score of 0.55 on the official English test track. This result secures second place among all locally-hosted, open-source submissions. Our work demonstrates that privacy-preserving, on-premise LLM pipelines can achieve near-competitive performance with proprietary frontier models, providing a practical, data-sovereign framework for clinical NLP.

翻译：从非结构化电子健康记录（EHR）文本中提取结构化临床信息一直是医疗信息学的持续瓶颈。尽管大型语言模型（LLM）表现出高性能，但其在临床环境中的部署受到隐私风险、推理成本以及超出文本证据的幻觉倾向的阻碍。针对CL4Health 2026病例报告表（CRF）填写任务，我们提出了一种完全本地化、领域适配的流水线，采用MedGemma-27B模型来解决这些挑战。我们的两阶段架构将二元存在性分类与值提取分离，强制严格遵循文本证据，并确保对否定、不确定或未知状态生成确定性输出。通过利用特定项目的少量样本上下文学习（无需外部API调用或微调），我们的方法在官方英文测试集上实现了0.55的宏F1分数。该结果在所有本地部署的开源提交中排名第二。我们的工作表明，保护隐私的本地LLM流水线能够达到与专有前沿模型相近的竞争性能，为临床自然语言处理提供了实用且数据自主的框架。