Program in Child Health Evaluative Sciences, The Hospital for Sick Children, Toronto, ON, Canada.
Genomics Health Services Research Program, Li Ka Shing Knowledge Institute, St. Michael's Hospital, Unity Health Toronto, Toronto, ON, Canada.
Hum Genet. 2023 Mar;142(3):321-330. doi: 10.1007/s00439-022-02512-2. Epub 2023 Jan 11.
Chatbots, web-based artificial intelligence tools that simulate human conversation, are increasingly in use to support many areas of genomic medicine. However, patient preferences towards using chatbots across the range of clinical settings are unknown. We conducted a qualitative study with individuals who underwent genetic testing for themselves or their child. Participants were asked about their preferences for using a chatbot within the genetic testing journey. Thematic analysis employing interpretive description was used. We interviewed 30 participants (67% female, 50% 50 + years). Participants considered chatbots to be inefficient for very simple tasks (e.g., answering FAQs) or very complex tasks (e.g., explaining results). Chatbots were acceptable for moderately complex tasks where participants perceived a favorable return on their investment of time and energy. In addition to achieving this "sweet spot," participants anticipated that their comfort with chatbots would increase if the chatbot was used as a complement to but not a replacement for usual care. Participants wanted a "safety net" (i.e., access to a clinician) for needs not addressed by the chatbot. This study provides timely insights into patients' comfort with and perceived limitations of chatbots for genomic medicine and can inform their implementation in practice.
聊天机器人是一种基于网络的人工智能工具，可模拟人类对话，越来越多地用于支持基因组医学的多个领域。然而，患者对在各种临床环境中使用聊天机器人的偏好尚不清楚。我们对接受过自己或子女基因检测的个人进行了一项定性研究。参与者被问及他们在基因检测过程中使用聊天机器人的偏好。采用解释性描述的主题分析方法。我们采访了 30 名参与者（67%为女性，50%为 50 岁以上）。参与者认为聊天机器人在非常简单的任务（例如，回答常见问题）或非常复杂的任务（例如，解释结果）方面效率低下。聊天机器人在中等复杂程度的任务中是可以接受的，因为参与者认为他们在时间和精力上的投资会有回报。除了达到这个“最佳点”之外，如果聊天机器人被用作常规护理的补充而不是替代，参与者预计他们对聊天机器人的舒适度会增加。参与者希望有一个“安全网”（即，与临床医生联系）来满足聊天机器人无法解决的需求。这项研究及时深入了解了患者对基因组医学中聊天机器人的舒适度和感知限制，可为其在实践中的实施提供信息。
