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

点击翻译按钮获取中文摘要

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.