Dynamic markers of intraoperative instability

描述（由申请人提供）：本研究的广泛和长期目标是建立实时动态心血管风险指标，以帮助预测结果并指导高危患者接受复杂手术的管理。随着心血管高危患者出现在复杂的心血管手术中，主要不良围手术期事件（MAE）如中风、心力衰竭和心肌梗死的频率增加，并与住院时间延长、死亡率增加和医疗费用增加有关。传统的“静态”风险分层指标，如年龄或合并症，不能预测哪些患者有患MAE的风险，也不能为个性化治疗策略提供见解。这些方法未能考虑到个体专利中生理控制的波动，并使用简化的线性模型，这些模型无法捕捉作为“现实世界”生理信号标志的复杂、时变特征。该项目将应用最先进的非线性方法来实时术中血压信号，并创建一套新的动态指标，以促进早期发现术中细微的血流动力学紊乱。通过将这些复杂的术中信号与MAE（由经过验证的胸外科学会（STS）国家结局数据库确定）联系起来，将确定具有MAE预测价值的血流动力学“特征”。为了实现这些目标，拟议计划的三个具体目标是：1)确定a)特定个体在手术不同阶段的BPV是否固定，b) BPV对心脏手术后MAE的预测能力2)测试BPV从基线到cpb后期间的变化，比基线或cpb后BPV更能预测MAE，并验证BPV对结果的预测能力3)创建一个独特的开放获取数据库(术前，术中血流动力学信号记录和术后结果数据)可通过美国国立卫生研究院赞助的复杂生理信号PhysioNet研究资源（www.physionet.org）公开获取。术中每搏的血流动力学数据将直接从手术室监视器收集，并将与自动化麻醉信息系统和STS结果数据库集成。利用多尺度熵对数据进行去识别和分析。将测试熵数据对MAE的预测能力，并将熵数据单独或作为熵数据的一部分与传统STS风险指标进行比较。术后最佳结果的熵值范围将被确定并用于指导未来的介入研究。提出的动态方法为患者水平区分提供了一个有希望的解决方案；改善患者咨询，术中血流动力学管理和术后预后。

DESCRIPTION (provided by applicant): The broad, long-term objective of this research is to develop real-time dynamic cardiovascular risk indices to help predict outcomes and guide management of high-risk patients undergoing complex procedures. As high-cardiovascular risk patients present for complex cardiovascular surgery, the frequency of major adverse perioperative events (MAE) such as stroke, heart failure, and myocardial infarction has increased and is associated with longer hospitalization, increased mortality, and increased health care costs. Traditional "static" metrics of risk stratification, such as age or co morbid conditions, are not able to predict which patients are at risk for MAE or offer insights into individualized treatment strategies. These approaches fail to take into account fluctuations in physiologic control in individual patents and use simplified linear models that are unable to capture the complex, time varying features that are hallmarks of 'real-world' physiological signals. This project will apply state-of-the-art nonlinear methods to real-time intraoperative blood pressure signals and create a novel dynamical set of indices that facilitate early detection of subtle intraoperative hemodynamic disturbances. By relating these complex intraoperative signals to MAE, determined from the validated Society of Thoracic Surgeons (STS) National outcomes database, hemodynamic "signatures" with predictive value for MAE will be determined. To achieve these goals, the three specific aims of the proposed program are: 1) To determine a) if BPV is fixed for a given individual at various stages of surgery, b) BPV's predictive ability for postoperative MAE following cardiac surgery 2) To test the change in BPV from baseline to post-CPB periods as more predictive of MAE than either baseline or post-CPB BPV and to validate BPV's predictive ability of outcome 3) To create a unique open access database (preoperative, intraoperative hemodynamic signal recordings plus postoperative outcome data) publicly available via the NIH-sponsored PhysioNet Research Resource for Complex Physiologic Signals (www.physionet.org). The intraoperative beat-by-beat hemodynamic data will be collected directly from the operating room monitors and will be integrated with the automated anesthesia information systems and STS outcome database. The data will be deidentified and analyzed with multi scale entropy. The entropy data will be tested for its MAE predictive ability and compared to the traditional STS risk indices by itself or as a part of it. The entropy range at which the postoperative outcome is optimal will be determined and used as guidance for future interventional studies. The proposed dynamic approach offers a promising solution for patient level discrimination; improve patient counseling, intraoperative hemodynamic management and postoperative outcome.