Recovery from Opioid Use Disorder: subgroups, transition states, and their association with recovery outcomes

总结 该项目的总体目标是提供对恢复的异质性结果的科学理解 阿片类药物使用障碍（OUD）为此，本项目将证明（1）重要的潜在因素， 存在OUD恢复，（2）亚组解释了OUD恢复的个体间异质性，（3） 个体在恢复过程中在这些亚组之间过渡;（4）个体水平因素是影响恢复的前 亚组转换的指示，和（5）潜在恢复因子的变化与恢复结果相关。 为了实现这些目标，该项目将开发新的统计方法，以确定 个人，和潜在的必要恢复因素，实现，从OUD恢复。为了取得成功，COM- 结合了研究团队在统计方法开发、心理社会数据分析和康复方面的专业知识 包括OUD在内的物质使用障碍的影响。所有分析将在缓解期进行 从慢性阿片类药物成瘾-研究恢复的环境和社会经济因素（RECOVER）研究， 对接受皮下丁丙诺啡注射用于OUD治疗的个体进行的纵向研究。的 RECOVER数据集评估了心理社会指标，包括抑郁、生活压力、疼痛、戒断和 四年的恢复期生活质量。在目标1中，提出了一种新的贝叶斯动态聚类因子模型（BD- CFM）将得到发展。这些潜在的因素将被用来识别个体的亚群， 在个人层面的社会心理因素，是这些亚群之间的过渡沉淀。在目标2中， 将对BDCFM进行扩展，以确定潜在恢复因素的变化，作为恢复的预测因素 结果，如重新开始使用阿片类药物（即，复发）和治疗利用率。总的来说，这个项目非常 创新和有影响力，因为它将提供更深入的了解从OUD恢复的异质性， 从而为亚组特定干预提供信息，以促进成功康复。

Summary The overall goal of this project is to provide a scientific understanding of the heterogeneous outcomes of recovery from opioid use disorder (OUD). Toward this end, this project will demonstrate that (1) latent factors important to recovery from OUD exist, (2) subgroups explain the heterogeneity among individuals in recovery from OUD, (3) individuals transition among these subgroups during their recovery process, (4) individual-level factors are pre- dictive of subgroup transitions, and (5) changes in latent recovery factors are associated with recovery outcomes. To accomplish these objectives, this project will develop novel statistical methodologies to identify subgroups of individuals in, and the latent recovery factors necessary to achieve, recovery from OUD. To succeed, the com- bined expertise of the research team in statistical method development, psychosocial data analysis, and recovery from substance use disorders, including OUD, will be leveraged. All analyses will be performed on the Remission from Chronic Opioid Use–Studying Environmental and Socioeconomic Factors on Recovery (RECOVER) Study, a longitudinal study of individuals who received subcutaneous buprenorphine injections for OUD treatment. The RECOVER data set assessed psychosocial measures including depression, life stress, pain, withdrawal, and quality of life during four years of recovery. In Aim 1, a novel Bayesian Dynamic Clustering Factor Model (BD- CFM) will be developed. These latent factors will then be used to identify subgroups of individuals, and changes in individual-level psychosocial factors that are precipitants of transitions among these subgroups. In Aim 2, an extension of the BDCFM will be developed to identify changes in latent recovery factors as predictors of recovery outcomes, such as re-initiation of opioid use (i.e., relapse) and treatment utilization. Overall, this project is highly innovative and impactful as it will provide a deeper understanding of the heterogeneity of recovery from OUD, and thereby informing subgroup specific interventions to facilitate successful recovery.