Endotypes of thrombocytopenia in the critically ill

血小板减少症在重症患者中极为常见。然而，急性血小板反应在 重症患者的血小板减少症尚未得到很好的研究，ICU中血小板减少症的多因素病因使其难以 了解，或了解是否要治疗它。在一些情况下，如创伤性损伤或败血症， 血小板计数与出血、血栓形成和终末器官损伤有关。血小板被广泛研究， 血小板是止血的重要组成部分，但一个迅速出现的概念是，血小板也是全身性炎症的关键效应细胞。 炎症过程既是局部和全身炎症反应的煽动者，也是 导致组织损伤的炎症。血小板和炎症之间的联系是复杂和双向的， 由于炎性配体已显示调节血小板功能， 其他细胞类型的反应。该提案的首要主题是研究危重病患者的血小板动力学， 患者，构建该人群中血小板减少症的临床内型，并将这些内型与 通过计算机模拟的中尺度机制。我们将使用基于大型电子健康记录的 数据库和三态创伤数据库作为源数据来构建这些内型。我们将内型定义为临床型 沿着沿着四个维度定义的模式：（1）基线信息（人口统计学、慢性疾病负担、 疾病和入院诊断），（2）血小板计数时间序列的特征（降低率、最低点等），（三） 并发干预;（4）结果。计算方法将试图将临床内型植根于 机械解释（或替代解释的集合），有助于聚焦基础科学 调查，并关闭关键的知识差距，防止设计和使用有针对性的抗血小板炎症 重症患者的治疗。将在不同的复杂程度上开发计算模型， 注意将潜在机制与在血小板减少症患者中常规进行的功能测定联系起来，例如 凝血酶原时间、活化凝血时间和血栓弹力图。

Thrombocytopenia is extremely frequent in critically ill patients. However, the role of acute platelet responses in critically ill patients is not well studied, and the multifactorial etiology of thrombocytopenia in the ICU makes it difficult to understand, or understand whether or not to treat it. In several situations such as traumatic injury or sepsis, very low platelet counts have been to bleeding, thrombosis and end-organ injury. Platelets have been extensively studied as a key component of hemostasis, but a rapidly emerging concept is that platelets are also key effector cells in systemic inflammatory processes as both instigators of local and systemic inflammatory reactions and also participants in the inflammation that contributes to tissue injury. The link between platelets and inflammation is complex and bidirectional, as inflammatory ligands have been shown to regulate platelet function and activated platelets induce inflammatory responses in other cell types. The overarching theme of this proposal is to study platelet dynamics in critically ill patients, construct clinical endotypes of thrombocytopenia in this population, and to relate these endotypes to underlying mesoscale mechanisms through computational modeling. We will use a large electronic health record-based database and a tri-state trauma database as source data to construct these endotypes. We define endotype as clinical patterns defined along four dimensions: (1) baseline information (demographic, chronic disease burden, severity of illness and admitting diagnosis), (2) features of the platelet count time series (rate of decrease, nadir, etc.), (3) concurrent interventions, and (4) outcome. The computational approach will attempt to root clinical endotypes in mechanistic interpretations (or collections of alternative interpretations), contributing to focus basic science investiagtions, and to close key knowledge gaps preventing the design and use of targeted anti-platelet-inflammatory therapies in the critically ill. Computational models will be developed at different levels of complexity, with a specific attention to tie underlying mechanisms to functional assays routinely performed in thrombocytopenic patients, such as prothrombin time, activated coagulation time, and thromboelastogram.