Objective  To establish a short-term mortality risk scoring standard for sepsis-associated acute respiratory distress syndrome (sARDS) and provide a reference tool for clinicians to evaluate the severity of sARDS patients. Methods  A retrospective cohort study was conducted on sARDS patients admitted to the adult intensive care unit (ICU) of the First Affiliated Hospital, Hengyang Medical School, University of South China from January 1, 2013 to August 31, 2020. They were divided into a death group and a survival group according to whether they died within 28 days after admission to ICU. Clinical data of the patients was collected within 24 hours admitted to ICU. Related risk factors for mortality within 28 days after admission to ICU were screened out through univariate logistic regression analysis. A risk prediction model for mortality within 28 days after admission to ICU was established by multivariate logistic regression analysis. The Hosmer-Lemeshow χ² test and the area under the receiver operating characteristic (ROC) curve were used to evaluate the model’s goodness-fit and accuracy in predicting 28-day mortality of the sARDS patients, respectively. Finally, the clinical prognosis scoring criteria 28-day mortality of the sARDS patients were established according to the weight coefficients of each independent risk factor in the model. Results  A total of 150 patients were recruited in this study. There were 67 patients in the survival group and 83 patients in the death group with a 28-day mortality rate of 55.3%. Four independent risk factors for 28-day mortality of the sARDS patients, including invasive mechanical ventilation, the number of dysfunctional organs≥3, serum lactic acid≥4.3 mmol/L and the severity of ARDS. A risk prediction model for mortality within 28 days of the sARDS patients was established. The area under the ROC curve and 95% confidence interval (CI), sensitivity and specificity of the risk prediction model for 28-day mortality for the sARDS patients were 0.896 (95%CI 0.846 - 0.945), 80.7% and 82.1%, respectively, while that for acute physiology and chronic health evaluation Ⅱ (APACHEⅡ) score were 0.865 (95%CI 0.805 - 0.925), 71.1% and 89.6%; for sequential organ failure assessment (SOFA) score were 0.841 (95%CI 0.7799 - 0.904), 68.7%, and 82.1%; for the prediction scores of lung injury were 0.855 (95%CI 0.789 - 0.921), 81.9% and 82.1%, respectively. It was indicated that the prediction accuracy of this risk prediction model of 28-day mortality maybe was better than that of APACHE-Ⅱ score, SOFA score and prediction score of lung injury. In addition, four risk factors were assigned as invasive mechanical ventilation (12 points), serum lactic acid≥4.3mmol /L (1 point), number of organs involved≥3 (3 points), and severity of ARDS (mild for 13 points, moderate for 26 points, severe for 39 points). Further more, the score of each patient was 13 - 55 points according to the scoring criteria, and the score grade was made according to the percentile method: 13 - 23 points for the low-risk group for 28-day mortality, 24 - 34 points for the medium-risk group for 28-day mortality, 35 - 45 points for the high-risk group for 28-day mortality, and over 45 points for the extremely high-risk group for 28-day mortality. According to the scoring criteria, the prognosis of the patients in this study was analyzed. The mortality probability of each group was 0.0% in the low-risk group, 13.8% in the medium-risk group, 51.9% in the high-risk group, and 89.7% in the extremely high-risk group, respectively. Conclusions  The invasive mechanical ventilation, the number of involved organs≥3, serum lactic acid≥4.3 mmol /L and the severity of sARDS are independent risk factors for 28-day mortality of the sARDS patients. The scoring criteria may predict the risk of 28-day mortality for the sARDS patients.

Citation： WANG Qiaosheng, WANG Meiqiu, FU Hui. A short-term mortality risk scoring standard for sepsis-associated acute respiratory distress syndrome. Chinese Journal of Respiratory and Critical Care Medicine, 2022, 21(3): 162-169. doi: 10.7507/1671-6205.202201007 Copy