Can a Chatbot-delivered Alcohol Intervention Engage Users and Enhance Outcomes Over a Smartphone App? Development and Feasibility Testing of a StepAway 'Bot'

项目摘要/摘要 需要酒精治疗的人数与实际接受酒精治疗的人数之间存在巨大差距。 造成这一差距的因素很多，包括耻辱、治疗的可获得性和可负担性，以及个人 想要独立解决他们的问题。技术提供的干预可能会极大地 增加接受所需酒精干预服务的人数，因为他们有能力 绕过治疗障碍。智能手机是提供经验性干预的理想设备 随时随地需要帮助。我们的研究团队之前开发并测试了一种 基于智能手机的独立、自我管理的酒精使用障碍干预系统 基于经验支持的面对面治疗。这种基于位置的监控和干预 酒精使用障碍(LBMI-A)系统通过心理教育模块提供进展 以及即时干预，以降低饮酒的可能性。在为期6周的初步试验研究中，我们 将LBMI-A与基于网络的干预进行对比，补充了饮酒者体检(DCU) 用阅读疗法。虽然在这两种干预措施中饮酒量都减少了50%，但只有LBMI-A 显著增加了禁欲天数的百分比。关联了更大的LBMI-A使用率 随着饮酒量的大幅减少。基于iPhone的修订和改进版本Step Away已经发布 开发并已被从iTunes应用商店下载的个人使用。我们一步之遥 使用数据表明用户对未被推送给用户的模块的参与度相对较低， 这是许多健康应用程序都会遇到的问题。一种新的方法，通过 在这项研究中，人工智能聊天机器人将被开发出来，它具有提供增强用户的潜力 参与度和有效性，因为它可以通过文本界面接触到引入新的干预措施 通过Step Away即时帮助用户获得渴望、体验 痛苦的或需要社会支持的。它可以充当个人助理，引导用户通过 改变他们饮酒的过程。这个项目将集中在开发一个步骤远离聊天机器人在第一个 研究的阶段。我们还将增强现有Step Away应用程序的用户界面，以实现最佳可用性。 第二阶段将需要进行一项试点研究，以确定Step Away聊天机器人是否已增强 将用户参与度、干预保真度和结果功效与Step Away应用程序进行比较 一群酗酒者。控制条件将包括最低限度的干预；仅评估 应用程序。参与者将接受访谈，以确定他们对这两种干预措施的看法，以期 了解用户参与的障碍。最后，这个项目将为进一步的、大规模的 对Step Away进行评估，其内容将由试点研究的结果决定。

PROJECT SUMMARY / ABSTRACT A wide gap exists between the number of individuals needing alcohol treatment and those actually receiving it. Many factors account for this gap, including stigma, treatment availability and affordability, and individuals wanting to solve their problems independently. Technologically-delivered interventions may dramatically increase the number of individuals who receive needed alcohol intervention services due to their ability to circumvent treatment barriers. Smartphones are the ideal devices to provide empirically-supported intervention assistance whenever and wherever it is needed. Our research team previously developed and tested a stand-alone, self-administered smartphone-based intervention system for alcohol use disorders that was based on empirically supported face-to-face treatments. This Location-Based Monitoring and Intervention for Alcohol Use Disorders (LBMI-A) system provided a progression through psychoeducational modules and in-the-moment interventions to reduce the likelihood of drinking. In a preliminary 6-week trial study, we contrasted the LBMI-A with a web-based intervention, the Drinker's Checkup (DCU) that was supplemented with bibliotherapy. While alcohol consumption was reduced by >50% in both interventions, only the LBMI-A produced a significant increase in percentage of days abstinent. Greater LBMI-A usage was associated with greater reductions in drinking. A revised and improved iPhone-based version, Step Away, has been developed and has been used by individuals downloading it from the iTunes App Store. Our Step Away usage data indicate that user engagement with modules that are not “pushed” to the user is relatively low, which is a problem that many health apps experience. A new method of delivering Step Away through an artificially intelligent chatbot will be developed in this study that holds potential for providing enhanced user engagement and effectiveness as it can reach out through a text interface to introduce new intervention steps and respond to the user with Step Away's in-the-moment help with having a craving, experiencing distress or needing social support. It could serve as a personal assistant that guides a user through a process of changing their drinking. This project will focus on developing a Step Away chatbot during the first phase of the study. We will also enhance the existing Step Away app's user interface for optimal usability. The second phase will entail performing a pilot study to determine if the Step Away chatbot has enhanced user engagement, intervention fidelity and outcome efficacy in comparison to the Step Away app among a group of problem drinkers. A control condition will consist of a minimal intervention; an assessment-only app. Participants will be interviewed to determine their perceptions of both interventions with a view towards understanding barriers to user engagement. Finally, this project will set the stage for a further, large-scale evaluation of Step Away, the content of which will be determined by the outcome of the pilot study.