Abstract: Ｏｂｊｅｃｔｉｖｅ：To implore the relationship between dynamic disease severity scores,bioelectrical impedance analysis(BIA) and the outcomes of critically ill patients.
Ｍｅｔｈｏｄｏｌｏｇｙ：Patients admitted to the intensive care unit in Jinling Hospital from June 2018 to June 2019 were screened,which had the record of 2 times of BIA measurement on the first and the third day of ICU stay was enrolled. Data from BIA including electrical impedance value(Z),current frequency (f),and the regression coefficient S value between Z and f,were collected,as well as clinical data including simplified acute physiology score Ⅱ (SAPS Ⅱ),sequential organ failure assessment (SOFA),and acute physiology and chronic health evaluation Ⅱ (APACHE Ⅱ) . The cutoff values of the receiver operating characteristic (ROC) curve for the S value and disease severity scores to predict the 90day death were used to define the 1st day disease severity as mild and severe. The 25th percentile value of the 3rd day disease severity scores deviation from the baseline was used to define the disease at the 3rd day as improved or deteriorated,while for S value,the definition of improved or deteriorated at the 3rd day was still using the cutoff value of ROC at the 3rd day. Basing on the 1st condition(mild or severe)and the 3rd day change(improved or deteriorated ),the patients were divided into 4 groups: mild,nondeteriorated; mild,deteriorated; severe,improved; severe,nonimproved. The endpoint was the 90day followup,then the relationship between the abovementioned parameters and the outcomes was analyzed.
Ｒｅｓｕｌｔｓ：Among 113 patients enrolled,83 were man,and the 90day mortality rate was 354%. Between the dead and survived patients,significant differences were found for S value and all disease severity scores(P<005). The difference of KaplanMeier survival curves between 4 groups divided according to different parameters was significant only for S value and SAPS Ⅱ score (P<005). A new score combining S value and SASP Ⅱ showed a good linear correlation with survival rate(prediction equation: Y=0174X，R2 096,P<0001）. COX regression analysis showed that the SSAPS Ⅱ score are the only independent risk factor for 90day death (HR=200,95%CI 153261,P<0001）.
Ｃｏｎｃｌｕｓｉｏｎ：The dynamic changes of S value,SAPS Ⅱ score are better to discriminate patients with different outcomes,a new score combining both of them shows a good linear correlation with survival rate,which may imply its value in disease severity assessment.

Key words: bioimpedance analysis, disease severity, dynamic changeou, tcome