Rivera Rivera Jessica Nathalie, AuBuchon Katarina E, Smith Marjanna, Starling Claire, Ganacias Karen G, Danielson Aimee, Patchen Loral, Rethy Janine A, Blumenthal H Joseph, Thomas Angela D, Arem Hannah
Healthcare Delivery Research Network, MedStar Health Research Institute, Washington, DC, United States.
Lombardi Comprehensive Cancer Center, Georgetown University, Washington, DC, United States.
JMIR Pediatr Parent. 2024 Nov 14;7:e56807. doi: 10.2196/56807.
The 42 days after delivery ("fourth trimester") are a high-risk period for birthing individuals and newborns, especially those who are racially and ethnically marginalized due to structural racism.
To fill a gap in the critical "fourth trimester," we developed 2 ruled-based chatbots-one for birthing individuals and one for newborn caregivers-that provided trusted information about postbirth warning signs and newborn care and connected patients with health care providers.
A total of 4370 individuals received the newborn chatbot outreach between September 1, 2022, and December 31, 2023, and 3497 individuals received the postpartum chatbot outreach between November 16, 2022, and December 31, 2023. We conducted surveys and interviews in English and Spanish to understand the acceptability and usability of the chatbot and identify areas for improvement. We sampled from hospital discharge lists that distributed the chatbot, stratified by prenatal care location, age, type of insurance, and racial and ethnic group. We analyzed quantitative results using descriptive analyses in SPSS (IBM Corp) and qualitative results using deductive coding in Dedoose (SocioCultural Research Consultants).
Overall, 2748 (63%) individuals opened the newborn chatbot messaging, and 2244 (64%) individuals opened the postpartum chatbot messaging. A total of 100 patients engaged with the chatbot and provided survey feedback; of those, 40% (n=40) identified as Black, 27% (n=27) identified as Hispanic/Latina, and 18% (n=18) completed the survey in Spanish. Payer distribution was 55% (n=55) for individuals with public insurance, 39% (n=39) for those with commercial insurance, and 2% (n=2) for uninsured individuals. The majority of surveyed participants indicated that chatbot messaging was timely and easy to use (n=80, 80%) and found the reminders to schedule the newborn visit (n=59, 59%) and postpartum visit (n=66, 66%) useful. Across 23 interviews (n=14, 61% Black; n=4, 17% Hispanic/Latina; n=2, 9% in Spanish; n=11, 48% public insurance), 78% (n=18) of interviewees engaged with the chatbot. Interviewees provided positive feedback on usability and content and recommendations for improving the outreach messages.
Chatbots are a promising strategy to reach birthing individuals and newborn caregivers with information about postpartum recovery and newborn care, but intentional outreach and engagement strategies are needed to optimize interaction. Future work should measure the chatbot's impact on health outcomes and reduce disparities.
分娩后的42天（“产后期”）是产妇和新生儿的高危时期，尤其是那些因结构性种族主义而在种族和族裔上处于边缘地位的人群。
为填补关键“产后期”的空白，我们开发了两个基于规则的聊天机器人——一个针对产妇，一个针对新生儿护理人员——提供有关产后警示信号和新生儿护理的可靠信息，并将患者与医疗服务提供者联系起来。
2022年9月1日至2023年12月31日期间，共有4370人收到了新生儿聊天机器人的推广信息，2022年11月16日至2023年12月31日期间，有3497人收到了产后聊天机器人的推广信息。我们用英语和西班牙语进行了调查和访谈，以了解聊天机器人的可接受性和可用性，并确定改进的领域。我们从分发聊天机器人的医院出院名单中抽样，按产前护理地点、年龄、保险类型和种族群体分层。我们在SPSS（IBM公司）中使用描述性分析来分析定量结果，在Dedoose（社会文化研究顾问公司）中使用演绎编码来分析定性结果。
总体而言，2748人（63%）打开了新生儿聊天机器人的消息，2244人（64%）打开了产后聊天机器人的消息。共有100名患者与聊天机器人互动并提供了调查反馈；其中，40%（n = 40）为黑人，27%（n = 27）为西班牙裔/拉丁裔，18%（n = 18）用西班牙语完成了调查。有公共保险的个人的支付者分布为55%（n = 55），有商业保险的为39%（n = 39），无保险的个人为2%（n = 2）。大多数接受调查的参与者表示，聊天机器人的消息及时且易于使用（n = 80，80%），并发现安排新生儿就诊提醒（n = 59，59%）和产后就诊提醒（n = 66，66%）很有用。在23次访谈中（n = 14，61%为黑人；n = 4，17%为西班牙裔/拉丁裔；n = 2，9%用西班牙语；n = 11，48%有公共保险），78%（n = 18）的受访者与聊天机器人互动。受访者对可用性和内容给予了积极反馈，并提出了改进推广信息的建议。
聊天机器人是一种很有前景的策略，可以向产妇和新生儿护理人员提供有关产后恢复和新生儿护理的信息，但需要有针对性的推广和互动策略来优化互动。未来的工作应该衡量聊天机器人对健康结果的影响并减少差异。
