Dynamic Risk Prediction of Life-Threatening Mass Effect After Ischemic Stroke

项目概要/摘要 Ong博士是一位神经学家和年轻的研究者，致力于以患者为导向的临床研究。对于这架K-23 提交，她将开发一个新的框架，用于威胁生命的质量效应的动态风险预测， 使用基线和纵向变量分析入院后120小时内的缺血性卒中。一 K23奖将为她提供获得关键职业发展技能的途径，使她能够 执行和建立她的项目，包括：1）动态风险预测和轨迹分析; 2）多模式 神经统计学方法; 3）临床试验设计; 4）专业发展。这些目标 将帮助Ong博士实现她的长期职业目标，成为一名独立的临床数据研究者- 驱动的工具，支持临床决策和优化急性脑损伤后的结果。王医生 招募了一个多学科的导师团队，以帮助她执行她的项目，并实现科学 独立她将由大卫格里尔博士共同指导，格里尔博士是R 01资助的临床科学家， 神经重症监护、神经系统诊断和临床研究设计，以及R 01资助的Emelia Benjamin博士 领先的脑血管流行病学家和风险预测专家，具有丰富的指导经验。博士 波士顿大学公共卫生学院生物统计学主席JoséeDupuis将作为一种方法论， 导师监督王博士的动态建模策略的进展。王博士的首要假设是 动态风险模型用最新的纵向数据更新预测， 预测危及生命的质量效应，更好地支持实时临床决策。目标1将 使用3000例大卒中患者的回顾性医疗记录数据集来识别变量轨迹 预测放射学LTME，并使用此信息来开发更新的多变量动态风险 LTME模型包括入院前120小时的基线和纵向变量。在目标2中， 她将通过前瞻性招募研究每小时定量瞳孔测量和LTME的关系 的60例大卒中患者，并开发了一个探索性的动态多变量模型的LTME使用 博士Ong提出的研究是 这是非常重要的，因为改善LTME预测可以促进更及时的生命和功能保留干预。 她的研究是创新的，因为她将开发和应用一种新的动态风险建模框架来预测 缺血性卒中后继发性损伤，并研究新的有前途的纵向变量，定量 瞳孔测量法 目标1中确定的预测变量以及每小时瞳孔测量数据。 她的目标，培训计划，和跨学科的导师团队将准备王博士成为一个 数据驱动工具的独立研究者，支持临床决策并优化 急性脑损伤预期的结果将是R 01提案的有力的初步数据，以测试以下方面的影响： 在临床实践中对干预时间和结局进行动态LTME评估。

PROJECT SUMMARY/ABSTRACT Dr. Ong is a neurologist and young investigator pursuing patient-oriented clinical research. For this K-23 submission, she will develop a novel framework for dynamic risk prediction of life-threatening mass effect after ischemic stroke using both baseline and longitudinal variables through the first 120 hours after admission. A K23 award will provide her with the means to acquire critical career development skills that will enable her to execute and build upon her project including: 1) dynamic risk prediction and trajectory analysis; 2) multi-modal methods of neuroprognostication; 3) clinical trial design; and 4) professional development. These objectives will help Dr. Ong to achieve her long-term career goal of becoming an independent clinical investigator of data- driven tools that support clinical decision making and optimize outcomes after acute brain injury. Dr. Ong has recruited a multi-disciplinary mentorship team to assist her in executing her project and achieving scientific independence. She will be co-mentored by Dr. David Greer, an R01 funded clinician scientist with expertise in neurocritical care, neuroprognostication, and clinical study design, and Dr. Emelia Benjamin, an R01 funded leading cerebrovascular epidemiologist and risk prediction specialist with extensive mentorship experience. Dr. Josée Dupuis, Chair of Biostatistics at Boston University School of Public Health, will serve as a methodologic mentor overseeing Dr. Ong’s progress in dynamic modeling strategies. Dr. Ong’s overarching hypothesis is that dynamic risk models that update their predictions with newly available longitudinal data will improve prediction of Life-Threatening Mass Effect and better support clinical decision making in real-time. Aim 1 will use a retrospective medical record dataset of 3000 large stroke patients to identify variables trajectories predictive of radiographic LTME, and use this information to develop updating multivariable dynamic risk models of LTME comprised of baseline and longitudinal variables for the first 120 hours of admission. In Aim 2, she will study the relationship of hourly quantitative pupillometry and LTME through the prospective recruitment of 60 patients with large stroke, and develop an exploratory dynamic multivariable model of LTME using Dr. Ong’s proposed research is significant because improving LTME prediction can facilitate more timely life- and function-sparing interventions. Her research is innovative because she will develop and apply a novel dynamic risk modeling framework to predict secondary injury following ischemic stroke, and study the new promising longitudinal variable, quantitative pupillometry. predictive variables identified in Aim 1 as well as hourly pupillometry data. Her aims, training plan, and interdisciplinary mentorship team will prepare Dr. Ong to become an independent investigator of data-driven tools that support clinical decision making and optimize outcomes after acute brain injury. The anticipated results will be strong preliminary data for a R01 proposal testing the effect of dynamic LTME assessments on time to intervention and outcome in clinical practice.