Research Ethics and Safety Promoted by Embodied Conversational Technology

描述（由申请人提供）：背景：尽管数十年来一直倡导改善人类受试者的知情同意，但多项研究表明，研究受试者通常不了解参与所需的基本概念。在过去的十年里，我们已经开发并测试了具体化的对话代理（ECA）——用英语和西班牙语用声音、手势、凝视线索和其他非语言行为模拟面对面对话的计算机角色。我们还试点测试了ECA平台的使用，作为支持潜在研究对象知情同意程序的方法。我们现在建议采用这一创新的尖端技术来创建和测试“体现对话技术促进的研究伦理与安全”（RESPECT）系统，该系统将：(1)改善潜在受试者了解研究的方式（ECA-Study Finder）；(2)帮助使潜在研究对象的知情同意更有意义（ECA-Consent Advocate）；(3)提高研究对象的体验（ECA-Study Buddy）。设计：随机对照试验假设：RESPECT系统参与者将：(1)有更好的研究发现者体验；(2)对知情同意有更好的理解；(3)更加遵守研究方案；(3)报告更多不良事件。研究人群：研究发现系统的用户和被纳入一组肿瘤学研究和波士顿医学中心的人。方法：ECA-Study Finder将通过一项随机对照试验对使用基于文本或基于eca的研究Finder系统的人进行评估。ECA-Consent Advocate和ECA-Study Buddy组件将通过一项随机对照试验进行评估，该试验纳入102名18岁以上的患者，这些患者正在接受波士顿大学医学中心的一项肿瘤学试验。结果测量：主要结果是对知情同意的理解程度。研究发现者的次要结果包括基于文本的与基于eca的研究发现过程的准确性。同意倡导者的次要结果包括入组可能性和满意度。研究伙伴系统的次要结果包括遵守研究方案、报告更多不良事件和满意度。预期结果：在本项目中，我们将评估一种可扩展的以患者为中心的干预措施，以使人们更容易搜索研究，提高知情同意的质量，提高对研究方案的依从性，并促进更好地报告不良事件。

DESCRIPTION (provided by applicant): Background: Despite decades of advocacy to improve informed consent for human subjects, multiple studies have shown that research subjects often do not understand fundamental concepts required for participation. Over the past decade we have developed and tested embodied conversational agents (ECA) - computer characters that simulate face-to-face conversation using voice, hand gesture, gaze cues and other nonverbal behavior in both English and Spanish. We have also pilot tested the use of the ECA platform as a method of supporting the informed consent process for potential research subjects. We now propose to adapt this innovative cutting edge technology to create and test the Research Ethics and Safety Promoted by Embodied Conversational Technology (RESPECT) system which will: (1) improve the way potential subjects learn about research studies (ECA-Study Finder); (2) help make informed consent for potential research subjects more meaningful (ECA-Consent Advocate); and (3) improve the experience of research subjects (ECA-Study Buddy). Design: Randomized Controlled Trial Hypotheses: That RESPECT system participants will: (1) have a better study finder experience; (2) have better comprehension of informed consent; (3) be more adherent to the study protocol; and (3) report more adverse events. Population Studied: Users of a study finder system and people being approached for enrollment into a set of oncology studies and Boston Medical Center. Methods: The ECA-Study Finder will be assessed with a randomized controlled trial of people who use either a text-based or ECA-based study finder system. The ECA-Consent Advocate and ECA-Study Buddy components will be assessed with a randomized controlled trial of 102 patients over 18 years old, who are being approached to participate in an oncology trial at Boston University Medical Center. Outcome Measures: The primary outcome is level of comprehension of informed consent. Secondary outcomes for the study finder include accuracy of the text-based versus ECA-based study finder process. Secondary outcomes for the consent advocate include likelihood to enroll and satisfaction. Secondary outcomes for the study-buddy system include adherence to the study protocol, reporting more adverse events, and satisfaction. Expected Results: In this project we will evaluate a scalable patient-centered intervention to make it easier for people to search for research studies, improve the quality of informed consent, improve adherence to research protocols, and promote better reporting of adverse events. PUBLIC HEALTH RELEVANCE: Despite decades of advocacy to improve informed consent for human subjects, multiple studies have shown that research subjects often do not understand fundamental concepts required for participation. In this project we will advance our research toward the development of a scalable, cost-effective, and sustainable patient-centered Embodied Conversational Technology intervention to make it easier for people to search for research studies, improve the quality of informed consent, improve adherence to research protocols, and promote better reporting of adverse events.