Development of an acute lung injury specific severity of illness measure

描述（由申请人提供）：开发旨在降低急性肺损伤死亡率的新疗法需要了解死亡的独立危险因素。大多数报告的急性肺损伤死亡风险因素来自于在学术医疗中心接受治疗的患者，尽管事实上大多数急性肺损伤患者在社区接受治疗。更好地描述死亡预测因素还可以促进对急性肺损伤患者疾病严重程度的量化，从而使研究人员在进行急性肺损伤观察研究和临床试验时改进风险调整和分层。本建议的目的是：1)在基于人群的急性肺损伤患者队列中确定死亡率的预测因素，2)开发和验证基于人群的急性肺损伤疾病严重程度测量，以及3)比较急性肺损伤特异性疾病严重程度测量与目前可用的通用疾病严重程度测量区分死亡的能力。这些目标将通过对金县肺损伤项目（KCLIP）进行二次分析来实现。KCLIP是一项大型的、前瞻性的、基于人群的队列研究，研究对象是来自华盛顿州金县的急性肺损伤患者，纳入了主要在社区医院接受治疗的患者。将评估急性肺损伤特异性变量以及先前显示的预测普通ICU患者死亡的变量与出院时死亡的关系。这些显著将进入一个多变量逻辑回归方程，以产生一个新的疾病严重程度模型。在独立患者队列中验证后，该模型的性能将与现有的疾病严重程度措施进行比较。相关性：尽管在重症监护方面取得了进展，但急性肺损伤仍然是一个重大的公共卫生负担，仅在美国每年就造成大约75,000名患者死亡。拟议的研究将提供关于急性肺损伤死亡率预测因子的新知识，以促进疾病严重程度的测量和确定急性肺损伤治疗的潜在治疗靶点。

DESCRIPTION (provided by applicant): Development of novel therapies aimed at reducing the mortality in acute lung injury requires an understanding of the independent risk factors for death. The majority of reported risk factors for death in acute lung injury derive from patients cared for in academic medical centers despite the fact that most acute lung injury patients are cared for in the community. Better characterization of the predictors of death can also facilitate quantification of an acute lung injury patient's severity of illness allowing investigators to improve risk-adjustment and stratification when performing observation studies and clinical trials in acute lung injury. The aims of this proposal are: 1) to identify predictors of mortality in a population-based cohort of patients with acute lung injury, 2) to develop and validate a population-based severity-of-illness measure for acute lung injury, and 3) to compare the acute lung injury specific severity-of-illness measure's ability to discriminate death to that of currently available generic severity-of-illness measures. These aims will be achieved by conducting a secondary analysis of the King County Lung Injury Project (KCLIP). KCLIP is a large, prospective population-based cohort of patients with acute lung injury from King County, Washington incorporating patients cared for predominantly in community hospitals. Acute lung injury specific variables as well as variables previously shown to predict death in general ICU patients, will be evaluated for their association with death at hospital discharge. Those significant will be entered into a multivariable logistic regression equation to generate a novel severity-of-illness model. After validation in an independent patient cohort the performance of this model will be compared to that of existing severity-of-illness measures. Relevance: Despite advances in critical care, acute lung injury remains a significant public health burden killing approximately 75,000 patients each year in the United States alone. The proposed study will provide new knowledge about the predictors of mortality in acute lung injury cared for in the community with the long term goal of facilitating severity-of-illness measurement and identifying potential therapeutic targets for acute lung injury treatment.