Expanding Access to Care for Marginalized Caregivers through Innovative Methods for Multicultural and Multilingual Adaptation of AI-Based Health Technologies

项目摘要和相关性 对患有慢性病的儿童的护理影响到1000万家庭照顾者，他们在 美国每年的无偿护理服务。这对照顾者来说，无论是身体上、精神上还是经济上都是巨大的负担。 护理负担在拉美裔社区中尤其普遍，拉美裔社区是规模最大、增长最快的少数族裔 美国人口。尽管需求压倒性和不成比例，支持性医疗服务的使用是 在拉丁裔照顾者中较低，因为障碍包括缺乏可接近的以及在文化和语言上合适的 程序。如果有的话，可以减轻多方面的护理负担，并改善老年人的整体健康状况 照顾者和孩子。语言翻译和文化适应一直是资源密集型的，最近的技术 机器学习和人工智能的进步有可能使劳动密集型工作自动化。然而，机器 如果不小心使用翻译，也可能加剧对边缘化群体的偏见和现有的不平等。 这项提议将在文化和语言上适应、开发和试点一种创新的、可访问的和具有成本效益的 目前基于人工智能(AI)的干预措施适用于说英语的慢性病儿童照顾者 健康状况--“在线照顾照顾者”(COCO)--西班牙语中的拉丁裔家庭照顾者。可可是一个 具有嵌入式聊天机器人的创新移动应用程序，利用尖端技术提供基于证据的 由于资源密集，以前在接触照顾者方面的能力有限的干预措施 照顾者和提供者两方面的要求。我们的初步工作，结合来自 关于COCO提供的特定干预组件(例如，日常健康检查、症状)的研究文献 自我监控和问题解决疗法)，指出了COCO作为一种高影响力和可扩展的工具的潜力 扩大获得护理的机会。 发展一种可获得的、在文化和语言上适当的干预措施具有改善的潜力 照顾患有慢性病儿童的拉丁裔家庭照顾者的健康结果，以及建立长期的 自我管理和照顾能力。这项研究的结果将指导未来的干预改进，通知 评估和招募战略，并为设计一项全面的临床试验提供初步疗效。拉丁裔 COCO是一种可访问和可扩展的干预措施，可解决在支持家庭照顾者方面的关键临床缺口 被边缘化的社区。通过适应COCO，我们将开发新的多元文化和多语言方法 这是可转让的，并有可能应用于改编其他基于对话的卫生干预措施以供使用 在文化和语言上被边缘化的其他人群中，以可扩展和具有成本效益的方式。

Project Summary and Relevance Care for children with chronic conditions impacts 10 million family caregivers, who provide an estimated $100 billion in unpaid care services in the U.S. annually. This is a significant burden for caregivers physically, mentally, and financially. The burden of care is particularly prevalent among Latino communities, the largest and fastest-growing minority population in the U.S. Despite the overwhelming and disproportionate need, the use of supportive healthcare services is lower among Latino caregivers due to barriers including the lack of accessible and culturally and linguistically appropriate programs. Those, if available, could reduce the multifaceted caregiving burden, and improve the overall health of the caregiver and child. Language translations and cultural adaptations have been resource-intensive, and recent technological advancements in machine learning and AI have the potential to automate labor-intensive work. However, machine translations, if used without careful intention, also may exacerbate biases and existing inequities for marginalized groups. This proposal will culturally and linguistically adapt, develop, and pilot test an innovative, accessible, and cost-effective artificial intelligence (AI)-based intervention currently available for English-speaking caregivers of children with chronic health conditions - “Caring of Caregivers Online” (COCO) - for Latino family caregivers in Spanish. COCO is an innovative mobile application with an embedded chatbot leveraging cutting-edge technologies to deliver evidence-based interventions that were previously limited in terms of their ability to reach caregivers due to resource-intensive requirements both on the caregiver side and provider side. Our preliminary work, in tandem with evidence from the research literature about specific intervention components COCO delivers (e.g., daily wellness check-in, symptom self-monitoring, and problem-solving therapy), point to COCO's potential as a high-impact and scalable tool for expanding access to care. The development of an accessible, and culturally and linguistically appropriate intervention has the potential to improve the health outcomes of Latino family caregivers caring for children with chronic conditions, as well as establish long-term self-management and caregiving skills. Results from this study will guide future intervention refinement, inform assessment and recruitment strategies, and provide preliminary efficacy for the design of a full-scale clinical trial. Latino COCO is an accessible and scalable intervention that addresses a critical clinical gap in supporting family caregivers in marginalized communities. Through adapting COCO, we will develop novel multicultural and multilingual methods that can be transferable and have the potential to be applied in adapting other dialog-based health interventions for use among other culturally and linguistically marginalized populations in a scalable and cost-efficient manner.