VA-DoD Long-Term Impact of Military-Relevant Brain Injury Consortium (LIMBIC): Phenotypes of Persistent Comorbidity in Post‐9/11 Era Veterans with mTBI

VA-DoD 军事相关脑损伤联盟 (LIMBIC) 的长期影响：后 9/11 时代患有 mTBI 的退伍军人持续合并症的表型

• 批准号: 10001099
• 负责人: Mary Jo Pugh
• 金额: --
• 依托单位: VA SALT LAKE CITY HEALTHCARE SYSTEM
• 依托单位国家: 美国
• 财政年份: 2019
• 资助国家: 美国
• 起止时间: 2019-10-01 至 2024-09-30
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10001099
• 关键词: Acute; Algorithms; Area; Behavior; Brain Injuries; Caring; Characteristics; Chronic; Chronic Care; Data; Data Sources; Epidemiology; Functional disorder; Health system; Individual; Injury; Inpatients; Learning; Mental Health; Methods; Military Personnel; Modeling; Nerve Degeneration; Nervous System Trauma; Neurodegenerative Disorders; Neurosecretory Systems; Outcome; Outpatients; Pain; Pathway interactions; Pharmacy facility; Phenotype; Prevalence; Process; Psychological reinforcement; Registries; Risk; Science; Self-Injurious Behavior; Severities; Sleep; Source; Substance Use Disorder; Support System; Text; Trauma; Veterans; Work; cohort; comorbidity; deep learning; economic outcome; health data; improved; interest; longitudinal analysis; mild traumatic brain injury; neurosensory; psychologic; screening

The Chronic Effects of Neurotrauma Warfighter Epidemiology Cohort was developed to identify phenotypes of comorbidity among deployed Post-9/11 Veterans in order to compare emergence of neurosensory, neurodegenerative, pain, and mental health comorbidity in Veterans TBI. The LIMBIC extension of the Warfighter Epidemiology Cohort will extend the work begun by CENC in which we identified a cohort of Post-9/11 Veterans and identified comorbidity phenotypes. We also obtained DoD trauma registry (DODTR) data, where available, and Military Health System (MHS) inpatient, outpatient, and pharmacy data that was included in the DoD Mental Health Data Cube. We now propose to expand upon this important data source for over 600,000 deployed SM’s to include a broader cohort of Post-9/11 era (deployed and nondeployed Veterans and additional data sources that provide unique opportunities to examine long-term comorbidity phenotypes and develop risk models for comorbidities of interest such as neurodegenerative disease, SUD, psychological comorbidities, and self-harm behaviors. These data will allow us to accomplish the following specific aims: Aim 1: Using “all sources” TBI severity algorithm and NLP/text embedding methods, identify phenotypes of mTBI in DoD and DoD+VA data that incorporate acute injury, mechanism of injury, and blast exposure. Aim 2: Identify prevalence of key comorbidities and outcomes at baseline, before and after mTBI exposure, and in VA (where relevant) and compare those rates by TBI severity and study group. Aim 3: Use deep learning models that incorporate mTBI phenotype, acute and chronic treatment approaches, and emergence of diverse comorbidities to develop risk scores for poor military outcomes and developing key comorbidities. Aim 4: Use deep learning models to identify optimal processes of care for mTBI. We will use data in DaVINCI to identify a cohort of Veterans who receive longitudinal VA care (at least once a year for three or more years between FY2002 and FY19 (at least one of which is after 2007 when TBI screening was mandated. We will also identify individuals who did not receive VA care. We will then categorize those with and without VA care as deployed and not deployed, creating four study groups: a) deployed with VA care; b) deployed without VA care; c) not deployed with VA care; d) not deployed without VA care. We will compile VA and DoD data sources and identify key comorbidities (Neuroendocrine dysfunction, substance use disorder, mental health conditions, pain conditions, sleep conditions, self-harm behaviors) and TBI characteristics. Those data will be used for machine/deep learning models that will develop TBI phenotypes, comorbidity phenotypes, and model risk scores for developing key comorbidities, and optimal processes of care for mTBI. Conducting these analyses for these four study groups will inform TBI pathways of care and illuminate specific target areas to improve acute TBI care and subsequent support systems for chronic care following TBI.

神经创伤战士流行病学队列的慢性影响旨在确定 9/11 后部署的退伍军人的共病表型，以便比较出现情况 退伍军人 TBI 中神经感觉、神经退行性、疼痛和心理健康合并症的研究。 战士流行病学队列的 LIMBIC 扩展将扩展 CENC 开始的工作 其中我们确定了一组 9/11 后退伍军人并确定了共病表型。 还获得了国防部创伤登记 (DODTR) 数据（如果有）和军事卫生系统 (MHS) 住院患者、门诊患者和药房数据，包含在国防部心理健康中 我们现在建议将这个重要的数据源扩展到超过 600,000 个。 部署了 SM 以涵盖更广泛的 9/11 后时代群体（已部署和未部署的 退伍军人和其他数据源提供了独特的机会来检查长期 合并症表型并为感兴趣的合并症开发风险模型，例如 神经退行性疾病、SUD、心理合并症和自残行为。 这些数据将使我们能够实现以下具体目标： 目标 1：使用“所有来源”TBI 严重性算法和 NLP/文本嵌入方法，识别 DoD 和 DoD+VA 数据中 mTBI 的表型，包括急性损伤、机制 伤害和爆炸暴露。 目标 2：确定基线、之前和之后主要合并症的患病率和结果 mTBI 暴露，以及 VA（如果相关），并按 TBI 严重程度和比例比较这些比率 学习小组。 目标 3：使用包含急性和慢性 mTBI 表型的深度学习模型 治疗方法以及各种合并症的出现，以制定风险评分 军事成果不佳并出现关键合并症。 目标 4：使用深度学习模型来确定 mTBI 的最佳护理流程。 我们将使用 DaVINCI 中的数据来识别接受纵向 VA 护理的退伍军人队列 （在 2002 财年至 2019 财年之间的三年或三年以上，每年至少一次（其中至少一年是 2007 年强制执行 TBI 筛查后，我们还将确定未进行 TBI 筛查的个人。 然后，我们将接受 VA 护理和未接受 VA 护理的人员分类为已部署和未部署。 部署，创建四个研究组：a) 部署有 VA 护理；b) 部署没有 VA 护理； 未在 VA 护理的情况下部署； d) 未在 VA 护理的情况下部署 我们将编制 VA 和国防部数据。 来源并确定关键合并症（神经内分泌功能障碍、物质使用障碍、 心理健康状况、疼痛状况、睡眠状况、自残行为）和 TBI 这些数据将用于开发 TBI 的机器/深度学习模型。 表型、合并症表型和发生关键合并症的模型风险评分， 以及 mTBI 的最佳护理流程。 对这四个研究组进行这些分析将为 TBI 的护理和治疗途径提供信息 阐明特定目标领域，以改善急性 TBI 护理和后续支持系统 TBI 后的长期护理。