Adapting the Penny Chatbot for Perinatal OUD Patients: COPILOT

提案摘要/摘要 美国阿片类药物危机对许多人群产生了重大影响，其中受影响最大的是 围产期妇女。在医院分娩的妇女中阿片类药物使用障碍（OUD）的患病率增加 从2010年到2017年增长了131%。怀孕和产后的OUD增加了许多医疗风险。 并发症，并可能导致不良的出生结果，包括早产和宫内胎儿死亡， 以及产妇发病率和死亡率。孤独和社会孤立也与较差的进展有关 急性、慢性、身体和精神状况以及死亡率，并在围产期期间加剧 孕期（指怀孕和产后第一年）。孤独和社会孤立 与阿片类药物使用量增加，复发率和过量有关，即使在控制OUD时也是如此 治疗、就业、关系状况和抑郁症。 为了解决围产期妇女在围产期和OUD护理中的孤独感和参与感，我们计划 调整现有的短信支持聊天机器人Penny，使其适合怀孕女性使用 产后和处理OUD新适应的聊天机器人Penny COPILOT将允许双向 短消息服务（SMS）消息传送，其使用自然语言处理来适当地响应， 精确到用户生成的输入。通过这种方式，Penny COPILOT感觉就像给朋友发短信一样，因为它使用 真实的句子，并尽量减少尴尬的确认信息。我们的团队与宾夕法尼亚混合大学合作 方法研究实验室（MMRL）和Memora健康技术公司，将使用干预映射指导 实施科学综合框架。我们将进行需求评估， 一个咨询委员会，从事预测试，以确保安全和完善的内容，并试点测试所产生的适应 Penny COPILOT在20例OUD围产期女性样本中评价可接受性、可行性和患者 满意我们的目标是开发和完善一个可接受的，可行的，令人满意的支持短信聊天机器人 促进患者参与围产期和OUD护理，减少孤独感。

PROPOSAL SUMMARY/ABSTRACT The US opioid crisis has had a significant effect on many populations, not the least of which effected are perinatal women. The prevalence of opioid use disorder (OUD) among women giving birth in hospitals increased by 131% from 2010-2017. OUD in pregnancy and postpartum increases the risks for a host of medical complications and can lead to adverse birth outcomes including preterm delivery and intrauterine fetal demise, as well as maternal morbidity and mortality. Loneliness and social isolation are also linked to poorer progression of acute, chronic, physical, and mental conditions, as well as mortality, and are compounded during the perinatal period (which is defined as pregnancy and the first postpartum year). Both loneliness and social isolation have been linked to increased opioid use amount, relapse rates, and overdose even when controlling for OUD treatment, employment, relationship status, and depression. To address both loneliness and engagement in perinatal and OUD care among perinatal women, we plan to adapt an existing texting support chatbot, Penny, to make it appropriate for use by women who are pregnant and postpartum and dealing with OUD. The newly adapted chatbot, Penny COPILOT, will allow for two way short message service (SMS) messaging using natural language processing to respond appropriately and accurately to user generated input. In this way, Penny COPILOT feels like texting a friend, as it responds using real sentences and minimizes awkward confirmatory messages. Our team, in collaboration with the Penn Mixed Methods Research Lab (MMRL) and Memora Health Technology Company, will use intervention mapping guided by the Consolidated Framework for Implementation Science. We will conduct a needs assessment, assemble an advisory board, engage in pretesting to ensure safety and refine content, and pilot test the resultant adapted Penny COPILOT in a sample of 20 perinatal women with OUD to evaluate acceptability, feasibility, and patient satisfaction. Our goal is to develop and refine an acceptable, feasible, and satisfactory supportive texting chatbot to promote patient engagement in perinatal and OUD care and decrease perceived loneliness.