Mechanisms of Change for an Effective Alcohol Text Message Intervention

 描述(由申请人提供)：18-25岁的年轻人有很高的危险饮酒率和与酒精相关的后果，但很少接触酒精干预。急诊科(ED)的设置提供了一个重要的机会来识别年轻的成年危险饮酒者，他们可以从酒精干预中受益。在ED中看到的年轻成年危险饮酒者中，通过9个月的跟踪调查，短信(TM)干预已经证明在减少危险饮酒(酗酒)方面是有效的。然而，TM对酗酒干预的“有效成分”(TM策略，有助于减少酗酒)和“改变机制”(个体内发生的导致酗酒减少的过程)仍然不清楚，并阻碍了构建最大影响和持久性的移动干预措施的进展。为了解决这一差距，需要一种拆解设计来确定TM干预的组件效应，并确定组件的运行机制，这将对移动干预的设计和行为变化的通知理论产生广泛的影响。该项目将确定TM干预“有效成分”是否包括“目标提示”(即TM提示致力于低饮酒量目标)、“反馈”(例如，关于是否愿意设定饮酒目标的实时TM反馈)和“提示自我监控”的关键组合，或者三个组成部分中的每一个单独进行。TM干预是行为改变机制研究中的一种方法论创新，因为它通过计算机算法实现干预传递的高保真，并收集特定干预成分的实时响应数据。对危险饮酒/AUD筛查呈阳性的年轻成人(18-25岁)ED患者(N=1,875)将被招募参加为期12周的TM干预的随机试验，以减少酗酒。患者将被随机分为五种情况之一：(1)“提示自我监测”(TM控制条件)，(2)“目标提示”，(3)“反馈”，(4)所有3个TM成分组合，以及(5)仅评估(控制条件)。所有参与者在基线(ED访问)、随机化后3个月和6个月完成评估。四个TM组的参与者在干预阶段每周完成TM询问，以跟踪饮酒行为的变化，并提出干预效果(意图、动机、内隐认知)的中介变量，以确定特定成分的变化机制。我们还将检查与干预效果相关的个体差异(例如，基线酒精严重程度)，以确定“对谁”治疗有效。确定TM干预“如何”和“对谁”产生影响将被用来修订TM干预内容，以最大限度地影响已确定的机制，并在确定更普遍地包括在移动干预中的关键组成部分方面具有更广泛的影响。此外，实时收集关于干预反应的细粒度数据将使基于更动态的行为变化模型的范式转变成为可能。

 DESCRIPTION (provided by applicant): Young adults ages 18-25 have high rates of hazardous alcohol use and alcohol-related consequences, but are rarely exposed to alcohol interventions. The Emergency Department (ED) setting provides an important opportunity to identify young adult hazardous drinkers who could benefit from an alcohol intervention. Among young adult hazardous drinkers seen in the ED, a text message (TM) intervention has demonstrated effectiveness in reducing hazardous alcohol use (binge drinking) through 9-month follow-up. However, the "active ingredients" (TM strategies that facilitate reductions in binge drinking) and "mechanisms of change" (processes occurring in the individual that lead to reduced binge drinking) of TM intervention for binge drinking remain unclear, and impede progress in constructing mobile interventions with maximum impact and durability. To address this gap, a dismantling design is needed to determine the TM intervention's component effects, and to identify the mechanisms through which components operate, which will have broad implications for the design of mobile interventions and informing theories of behavior change. This project will determine whether TM intervention "active ingredients" involve the critical combination of "goal prompt" (i.e., TM prompt to commit to a low drinking quantity goal), "feedback" (e.g., real-time TM feedback on willingness to set a drinking goal), and "cued self-monitoring" or each of the three components separately. TM intervention is a methodological innovation in research on mechanisms of behavior change because of high fidelity of intervention delivery via computerized algorithm, and collection of real-time data on response to specific intervention components. Young adult (ages 18-25) ED patients who screen positive for hazardous drinking/AUD (N=1,875) will be recruited to participate in a randomized trial of a 12-week TM intervention to reduce binge drinking. Patients will be randomized to one of five conditions: (1) "Cued Self-Monitoring" (TM control condition), (2) "Goal Prompt", (3) "Feedback", (4) all 3 TM components combined, and (5) Assessment Only (control condition). All participants complete assessments at baseline (ED visit), 3- and 6-months post-randomization. Participants in the four TM arms complete weekly TM queries during the intervention phase to track changes in drinking behavior and proposed mediators of intervention effects (intention, motivation, implicit cognitions) in order to determine component-specific change mechanisms. We also will examine individual differences (e.g., baseline alcohol severity) associated with intervention effects, to identify "for whom" treatment has effects. Determining "how" and "for whom" TM intervention has effects will be used to revise TM intervention content to maximally impact identified mechanisms, and has broader implications in terms of determining key components to include in mobile interventions more generally. Further, fine-grained data on response to intervention collected in real time will enable a paradigm shift based on the development of a more dynamic model of behavior change.