Objective:To investigate the risk factors of acute respiratory distress syndrome（ARDS）in children with sepsis，and to establish a risk prediction model based on least absolute shrinkage and selection operator(LASSO) regression analysis for early identification of sepsis patients who may develop ARDS.Methods:The children with sepsis treated in the Pediatric Center of the First Affiliated Hospital of Xinjiang Medical University from January 2021 to December 2023 were selected，and divided into ARDS group and non-ARDS group according to whether ARDS occurred.The basic data and related laboratory indexes of the two groups were compared,and the characteristic variables were screened by LASSO regression.The independent risk factors of ARDS in septic children were analyzed by multivariate Logistic regression,and a Nomogram prediction model was constructed.The predictive power of the model was verified by using receiver operating characteristic(ROC) curve,calibration curve and clinical decision curve.Results:A total of 197 children with sepsis were included,including 113 males and 84 females,51 cases (25.9%) in ARDS group and 146 cases (74.1%) in non-ARDS group.There were no statistically significant differences in gender,ethnicity,age and body mass index between the two groups( P>0.05).The pediatric critical illness score,pH value,alkaline residue,and platelets in ARDS group were significantly lower than those in non-ARDS group.The ratio of pulmonary infection and septic shock,mortality,lactic acid,interleukin-6,urea,creatinine,uric acid,N-terminal B-type natriuretic peptide precursor and so on were significantly higher than those in non-ARDS group( P<0.05).According to LASSO regression and multivariate Logistic analysis,high lactate ( OR 1.715,95% CI 1.409-2.088, P<0.001),pulmonary infection ( OR 3.679,95% CI 1.653-8.187, P<0.001),and septic shock ( OR 11.474,95% CI 5.396-24.398, P<0.001) were independent risk factors for ARDS in children with sepsis.The prediction model constructed based on these three independent risk factors had an area under the ROC curve of 0.795,a sensitivity of 0.686,and a specificity of 0.870.The calibration curve showed that the predicted probability of the model was basically consistent with the actual probability.The clinical decision curve showed that the prediction model had good clinical effectiveness. Conclusion:High lactate，pulmonary infection，and septic shock are independent risk factors for ARDS in children with sepsis.The clinical prediction model based on these three independent risk factors has good diagnostic efficacy and clinical application value.

Objective:To investigate the risk factors of acute respiratory distress syndrome（ARDS）in children with sepsis，and to establish a risk prediction model based on least absolute shrinkage and selection operator(LASSO) regression analysis for early identification of sepsis patients who may develop ARDS.Methods:The children with sepsis treated in the Pediatric Center of the First Affiliated Hospital of Xinjiang Medical University from January 2021 to December 2023 were selected，and divided into ARDS group and non-ARDS group according to whether ARDS occurred.The basic data and related laboratory indexes of the two groups were compared,and the characteristic variables were screened by LASSO regression.The independent risk factors of ARDS in septic children were analyzed by multivariate Logistic regression,and a Nomogram prediction model was constructed.The predictive power of the model was verified by using receiver operating characteristic(ROC) curve,calibration curve and clinical decision curve.Results:A total of 197 children with sepsis were included,including 113 males and 84 females,51 cases (25.9%) in ARDS group and 146 cases (74.1%) in non-ARDS group.There were no statistically significant differences in gender,ethnicity,age and body mass index between the two groups( P>0.05).The pediatric critical illness score,pH value,alkaline residue,and platelets in ARDS group were significantly lower than those in non-ARDS group.The ratio of pulmonary infection and septic shock,mortality,lactic acid,interleukin-6,urea,creatinine,uric acid,N-terminal B-type natriuretic peptide precursor and so on were significantly higher than those in non-ARDS group( P<0.05).According to LASSO regression and multivariate Logistic analysis,high lactate ( OR 1.715,95% CI 1.409-2.088, P<0.001),pulmonary infection ( OR 3.679,95% CI 1.653-8.187, P<0.001),and septic shock ( OR 11.474,95% CI 5.396-24.398, P<0.001) were independent risk factors for ARDS in children with sepsis.The prediction model constructed based on these three independent risk factors had an area under the ROC curve of 0.795,a sensitivity of 0.686,and a specificity of 0.870.The calibration curve showed that the predicted probability of the model was basically consistent with the actual probability.The clinical decision curve showed that the prediction model had good clinical effectiveness. Conclusion:High lactate，pulmonary infection，and septic shock are independent risk factors for ARDS in children with sepsis.The clinical prediction model based on these three independent risk factors has good diagnostic efficacy and clinical application value.