IMPROVE DISCHARGE INSTRUCTION THROUGH AUTOMATED PICTOGRAPH ENHANCEMENT

描述（由申请人提供）：出院说明为患者管理自己的护理提供了关键信息。这些指示是满足医疗保险和医疗补助服务中心以及医疗保健组织认证联合委员会的认证和公开报告要求的核心措施。然而，研究表明，许多患者并不完全理解他们收到的指示，这种缺乏理解降低了患者的满意度和依从性。其他因素使患者理解和记住指示具有挑战性：身体和情绪困扰，健康素养水平低，缺乏动力和环境干扰。由于这些其他因素不能在出院时消除，研究人员和临床医生探索各种策略，使指示更容易理解。在实践中，护士可能会用通俗易懂的语言进行解释，并绘制图表。研究表明，文本简化和象形文字增强是有帮助的，象形文字对低识字率的患者特别有益。在这个项目中，我们将重点放在象形文字增强。象形文字的好处已经在许多研究中得到证实。然而，有几个问题阻碍了象形文字在出院指示中的使用：（1）用象形文字手动增强患者特定指示非常耗时;（2）不存在用于患者交流的标准象形文字语言;以及（3）关于如何系统地开发和评估用于患者交流的象形文字的研究有限。因此，患者特定的出院指示很少包括象形文字，患者仍然难以理解。为了解决这个问题，我们将创建一种面向患者的象形语言，并开发一个系统来自动增强带有象形文字的自由文本说明。我们将集中在妇科肿瘤学领域。我们的总体目标是提高患者的理解力、满意度和依从性。具体而言，我们将：（1）使用系统的文本分析、象形图设计和验证方法，为出院说明创建象形词汇和语法;（2）通过信息提取和象形图生成开发自动象形图增强系统;（3）评估自动象形图增强的安全性和益处。我们将在健康志愿者身上测试该系统的安全性。对患者理解、患者满意度和护士满意度的影响将在Brigham and Women's Hospital患者的RCT中进行评估。

DESCRIPTION (provided by applicant): Discharge instructions provide critical information for patients to manage their own care. Such instructions are required as core measures to meet accreditation and public reporting requirements by the Centers for Medicare and Medicaid Services and the Joint Commission on Accreditation of Healthcare Organizations. However, studies show that many patients do not fully understand the instructions they receive and that this lack of comprehension reduces patient satisfaction and compliance. Other factors make understanding and remembering instructions challenging for patients: physical and emotional distress, low health literacy level, lack of motivation, and environmental distractions. Since these other factors cannot be eliminated at the time of discharge, researchers and clinicians explore various strategies to make the instructions more understandable. In practice, nurses may provide explanations in lay-friendly terms and draw figures and tables. Studies have shown that text simplification and pictograph enhancement are helpful and pictographs were especially beneficial for low-literacy patients. In this project we will focus on pictograph enhancement. The benefit of pictographs has been demonstrated in numerous studies. However, several problems inhibit the use of pictographs in discharge instructions: (1) manually enhancing patient-specific instructions with pictographs is prohibitively time consuming; (2) there exists no standard pictographic language for patient communication, and (3) there has been limited research on how to systematically develop and evaluate pictographs for patient communication. As a result, patient-specific discharge instructions rarely include pictographs and remain difficult for patients to understand. To solve this problem, we will create a patient-oriented pictographic language and develop a system to automatically enhance free-text instructions with pictographs. We will focus on the domain of gynecologic oncology. Our overall goal is to improve patient comprehension, satisfaction and adherence. Specifically, we will: (1) create a pictographic vocabulary and grammar for discharge instructions, using systematic text analysis, pictograph design, and validation methods; (2) develop an automated pictograph-enhancement system through information extraction and pictograph generation; and (3) evaluate the safety and benefits of automated pictograph enhancement. We will test the system's safety on healthy volunteers. The impact on patient understanding, patient satisfaction, and nurse satisfaction will be assessed in a RCT with patients at Brigham and Women's Hospital.