Developing Methods to Estimate Mediation Effects in Adaptive Interventions

项目摘要 拟议的K 01指导研究科学家发展奖将准备Ahnalee Brincks博士。到 领导开发新的定量方法，以了解 以人为本的适应性干预措施影响药物使用和成瘾，包括海洛因和其他阿片类药物 使用障碍（OUD）。Brincks博士，密歇根州立大学学院生物统计学助理教授 人类医学，将受益于NIDA资助的密歇根州立大学药物依赖流行病学的建立 培训计划和国家西班牙裔科学网络指导计划，继续与大学的支持 在经验丰富的NIDA PI，Jim Anthony博士的指导下，他是K 01的主要赞助商。 以人为本的适应性干预（AI）由临床医生使用，根据以下因素调整治疗： 在治疗开始之前和之后观察到的参与者特征。作为有前途的新方法， DUA和OUD的临床和预防干预措施，AI为从业者提供了基于证据的工具， 干预，以适应每个病人的需求，因为这些需求的演变，以提供最有效的干预， 理想的时间和最佳剂量。现有的统计方法可以评估人工智能是否有效， 人工智能是有效的。然而，没有统计工具来了解认可机构如何在这一层面上运作， 人的经验，通过超分子机制和介导途径在临床研究 and prevention预防researchcontext研究.需要新的中介方法来通过更好的理解来改善AI 介入的途径。识别干预途径可以发现新的干预靶点 在随后的OUD和DUA干预的现场驾驶AI试验中受到挑战或确认。 K 01的支持将加强Brincks博士的职业发展。除了学习OUD/DUA自然历史之外， 临床课程，诊断和病人护理服务，她还将学习1）流行病学应用于DUA， 其决定因素和并发症，定义为包括艾滋病毒/艾滋病、丙型肝炎病毒和其他传染病，2） 当前DUA干预措施及其假设的介质，3）用于建立适应性的实验设计 干预（例如，序贯多重分配随机试验），以及4）科学水平的因果关系 NIDA研究中的中介途径。在她的K 01研究中，她的主要目标是开发 新的定量方法，以更好地了解人工智能减少OUD和其他DUA的机制。一 第二个目标是应用这些方法来评估人工智能试验数据中的中介假设。她将发射 一个专注于OUD/DUA中介途径的独立计划，将当前的人工智能方法扩展到 临床干预评估，重点是降低社区、学校和社区的OUD/DUA发病率， 目标人群。因此，这项工作应该对NIDA的战略目标产生深远的影响。

PROJECT SUMMARY The proposed K01 Mentored Research Scientist Development Award will prepare Ahnalee Brincks, Ph.D. to lead the development of novel quantitative methods needed to understand the mechanisms through which person-focused adaptive interventions affect drug use and addiction (DUA), including heroin and other opioid use disorders (OUD). Dr. Brincks, Assistant Professor of Biostatistics at the Michigan State University College of Human Medicine, will benefit from the NIDA-funded establishment of MSU's Drug Dependence Epidemiology Training Program and National Hispanic Science Network Mentoring Program, continued with university support under the direction of an experienced NIDA PI, Dr. Jim Anthony, who serves as here lead K01 sponsor. Person-focused adaptive interventions (AIs) are used by clinicians make adaptations to treatment based on participant characteristics that are observed before and after treatment starts. As promising new approaches for clinical and preventive interventions in DUA and OUD, AIs give practitioners evidence-based tools to adjust an intervention to fit each patient's needs as these needs evolve, to deliver the most effective intervention at the ideal time and in the optimal dose. Current statistical approaches exist to evaluate whether AIs work and for whom AIs are effective. However, there are no statistical tools for understanding how the AIs work at the level of the person's experience, via supra-molecular mechanisms and mediational pathways investigated in clinical and prevention researchcontexts. New mediation methods are needed to improve AIs via greater understanding of intervening pathways. Identification of intervening pathways can lead to discover of new intervention targets to be challenged or confirmed in subsequent field-driving AI trials of OUD and DUA interventions. K01 support will strengthen Dr. Brincks' career development. In addition to studying OUD/DUA natural history and clinical course, diagnosis and patient care services, she also will learn 1) epidemiology as applied to DUA, their determinants and complications, defined to encompass HIV/AIDS, HCV, and other infectious diseases, 2) current DUA interventions and their hypothesized mediators, 3) experimental designs used to build adaptive interventions (e.g., Sequential Multiple Assignment Randomized Trials), and 4) state-of-the-science causal approaches to mediational pathways in NIDA research. In her K01 research, her primary aim is development of new quantitative methods to better understand mechanisms through which AIs reduce OUD and other DUA. A second aim involves applying these methods to evaluate mediation hypotheses in AI trial data. She will launch an independent program focused on OUD/DUA mediational pathways, extending current AI approaches beyond clinical intervention evaluations to focus on reducing OUD/DUA incidence rates in communities, schools, and target populations. As such, this work should have far-reaching implications for NIDA strategic goals.