A Framework to Enhance Decision Support by Invoking NLP: Methods and Applications

项目总结 电子健康记录(EHR)可以通过以下方式提高美国的医疗保健服务质量 向临床医生和患者提供自动最佳实践提醒。然而，这样的功能是 目前仅限于临床实践的狭窄领域，因为现有的决策支持系统只能处理 结构化数据，由于缺乏合适的框架以及对准确性和可移植性的担忧。 PI的初步工作表明，基于规则的方法可以用于开发广泛的领域 除了结构化数据之外还可以利用自由文本的提醒系统。PI已开发出样机 预防宫颈癌和结直肠癌的系统。这些系统由基于规则的复合模型组成 国家指南和基于规则的自然语言处理(NLP)解析器。NLP解析器提取 应用指南所需的患者变量。然而，还需要进一步的研究来扩大 并确保它们在临床部署中的准确性。在指导阶段，PI将进行协作 与临床医生一起扩展、迭代优化和验证系统，并将在 开放源码，以便将其改编用于其他机构的部署(目标1-K99)。在 独立阶段，PI将研究方法，以促进快速开发、部署和交叉 类似系统的机构可移植性。具体地说，PI将为解析器和 研究领域适应和主动学习方法，以减少开发的手动工作 和NLP解析器的改编(目标2-R00)。以使其他研究人员能够重复使用开发的 方法和软件资源，将开发一个工具包，以支持建设和 部署类似系统(AIM 3-R00)。该工具包将由用户友好的工具和模板组成 复制案例研究中设计的流程，并将建立在SHARPn数据标准化的基础上 工具和其他开源工具。独立阶段将与InterMountain合作 医疗保健。PI的职业目标是成为临床信息学的科学领导者，专注于 优化临床决策。该协会有很强的临床医学和医学背景 信息学，并将接受刘红芳博士、克里斯托弗·丘特博士、罗伯特·格林斯博士和 Rajeev Chaudhry，他们拥有免费的专业领域。导师(K99)阶段将为2 在罗切斯特梅奥诊所工作多年，在那里，PI将进行决策支持课程，并将获得 在NLP和卫生信息标准方面进行辅导培训。这将使PI准备好独立 在R00阶段研究可移植性和工装。拟议工作的完成将使PI能够 寻求更多资金用于试行已开发系统的临床部署，测量其临床应用 影响，并将该方法扩展到其他临床领域和机构。职业补助金将使 独立调查员的身份，并为 推进临床决策支持，以改善护理服务。

PROJECT SUMMARY Electronic Health Records (EHRs) can improve the quality of healthcare delivery in the United States, by providing automated best-practice reminders to clinicians and patients. However such functionality is currently limited to narrow areas of clinical practice, as existing decision support systems can process only structured data, due to lack of a suitable framework and concerns about accuracy and portability. Preliminary work by the PI has shown that rule-based approach can be used to develop broad-domain reminder systems that can utilize free-text in addition to the structured data. The PI has developed prototype systems for cervical and colorectal cancer prevention. These systems consist of rule-based composite models of national guidelines, and rule-based Natural Language Processing (NLP) parsers. The NLP parsers extract the patient variables required for applying the guidelines. However further research is needed to extend the systems and to ensure their accuracy for clinical deployment. In the mentored phase, the PI will collaborate with clinicians to extend and iteratively optimize and validate the systems, and will make them available in open-source so that they can be adapted for deployment at other institutions (aim 1 - K99). In the independent phase, the PI will research methods to facilitate rapid development, deployment and cross- institutional portability of similar systems. Specifically, the PI will develop a hybrid design for the parsers and investigate domain adaptation and active learning methods, for reducing the manual effort for development and adaptation of the NLP parsers (aim 2 - R00). To enable other researchers to reuse the developed methodologies and software resources, a toolkit will be developed that will support the construction and deployment of similar systems (aim 3 - R00). The toolkit will consist of user-friendly tools and templates to replicate the processes engineered in the case studies, and will build on the SHARPn data normalization tooling and other open-source tools. The independent phase will be in collaboration with Intermountain Healthcare. The PI's career goal is to become a scientific leader in clinical informatics with a focus on optimizing clinical decision making. The PI has strong background in clinical medicine and medical informatics, and will receive mentoring from Drs. Hongfang Liu, Christopher Chute, Robert Greenes and Rajeev Chaudhry, who have complimentary areas of expertise. The mentored (K99) phase will be for 2 years at Mayo Clinic Rochester, wherein the PI will undertake courses on decision support and will get mentored training in NLP and health information standards. This will prepare the PI for independent research in R00 phase on portability and tooling. Completion of the proposed work will enable the PI to seek further funding for piloting clinical deployment of the developed systems, measuring their clinical impact, and for scaling the approach to other clinical domains and institutions. The career grant will enable the PI to establish himself as an independent investigator and to make significant contributions towards advancing clinical decision support for improving care delivery.