Establishment Warning Systems Of Acute Respiratory Distress Syndrome

Background Acute respiratory distress syndrome(ARDS) is a syndrome consisting of acute hypoxemic respiratory failure with bilateral pulmonary infiltrates that is associated with both pulmonary and no pulmonary risk factors and that is not primarily due to left atrial hypertension. About 7-10% ICU patients and 5-8% mechanical ventilation patients occurring ARDS. Although in recent years for ARDS pathogenesis and treatment of have a great progress, but the mortality rate remains high and the length of stay in ICU patients with ARDS mortality rate as high as 25%-75%. It is found that early diagnosis and active intervention can improve the prognosis of ARDS and reduce the mortality of ARDS patients. Early intervention is dependent on early recognition and diagnosis of ARDS and the risk factors of acute respiratory distress syndrome (ARDS) to predict the possibility of the occurrence of ARDS patients, and take measures to prevent and early treatment and may help to improve the prognosis of the patients. Critical Illness and Injury Trials Group has built a system to establishment occurring of ALI/ARDS. The lack of a validated risk model that confirms and consolidates these risk modifiers prevents the systematic determination of a Chinese population at high risk for developing ARDS and is a major limitation to studies aimed at prevention or early intervention in ARDS.Purpose To determine the frequency and outcome of ARDS development in patients at risk and validate a lung injury prediction score base on the Chinese population.Methods This is a multicenter cohort study, patients of the 13 National University three affiliated hospitals were investigated. This research collected cases from January 1,2012 into the modeling group. Search for the possible risk factors for ARDS. Risk factors were screened by univariate analysis diagnosis of ARDS and non ARDS patients, assignment the risk factors by OR. The discrimination of the model was assessed with area under receiver operating curve (AUC). In the prospective multicenter observational cohort study, predisposing conditions and risk modifiers predictive of ARDS development were identified from routine clinical data available during initial evaluation.Results Thirteen hospitals enrolled 698 patients at risk including 500 cases of the model group; the evaluation group of 198 cases. Single factor analysis showed that the risk factors of BMI, immune disorders, cardiovascular disease, acute pancreatitis, acute peritonitis, severe infection, pulmonary contusion, hypoproteinemia, acid is toxic for the development of ARDS. By multivariate analysis, final model into the acute pancreatitis, pneumonia, aspiration, pulmonary contusion 4 risk primary disease and hypoproteinemia, acidosis, and PaO2/FiO2 is equal to or less than 300 the three risk index as a predictor of ARDS of risk factors. Through the risk factors in the model or value assignment to each factor:with a clinical diagnosis of ARDS as the standard, the establishment of the receiver operating curves, by its sensitivity and specificity and Youden index, and ultimately determine the threshold model for 7.5, ACU=0.779, model predictive sensitivity of ARDS was 43.0%, specificity was 92.5%; PPV=0.72, NPV=0.84; and non ARDS group compared with that in patients with acute respiratory distress syndrome (ARDS), mortality rate was significantly higher (28% vs.15%, P=0.001). Through the use of the model,198 patients were tested, and the area of AUC was 0.828, the sensitivity was 77.9%, specificity was 68.3%, and NPV=0.76, PPV=0.62.Conclusion Through the acute pancreatitis, pneumonia, aspiration, pulmonary contusion, hypoproteinemia, acidosis and PaO2/FiO2 is less than or equal to 300 high-risk factors established ARDS and early warning system, can be more effective in predicting ARDS occurrence.