Objective:To construct a predictive model for sepsis associated-acute respiratory distress syndrome(ARDS)by combining multiple indicators.Methods:We performed a retrospective analysis of the medical records of pa-tients with sepsis at the Sun Yat-sen Memorial Hospital from March 2018 to October 2022.Univariate and multivariate Lo-gistic regression were performed to identify the independent risk factors of ARDS,and then a joint prediction model was constructed based on the results of multivariate regression analysis.Results:A total of 303 patients were en-rolled in this study which including 145 patients with ARDS and 158 patients without ARDS.There was statistical signifi-cance between the two groups of patients(P<0.05)in terms of oxygenation index,blood oxygen,urea nitrogen,dia-stolic blood pressure,albumin,SOFA score,septic shock,pulmonary and urinary tract infections.Multivariate logis-tic regression analysis showed that albumin,oxygenation index,diastolic blood pressure,septic shock and pulmonary in-fection were independent predictors for the occurrence of septic ARDS.The area under the curve of the ARDS joint predic-tion model is 0.812(95%CI 0.76-0.86),the sensitivity is 69.7%,and the specificity is 84%.Conclusion:The ARDS prediction model constructed by combining indicators has good predictive performance,which is helpful for the clinical diagnosis and treatment of patients.
维普
万方数据