The Effect Of AIMS65,Glasgow-Blatchford And Rockall Scoring Systems In The Prognosis Assessment Of Acute Upper Gastrointestinal Bleeding

Background and AimsThe Glasgow-Blatchford score(GBS)and the Rockall score are currently the most widely used scoring systems for assessing the prognosis of patients with upper gastrointestinal bleeding(UGIB).However,due to its complicated calculation,the practical application in clinical work is cumbersome.A simple calculation of the AIMS65 scoring system developed in 2011 has received increasing attention in recent years.The aim of this study was to compare these three scores in predicting clinical outcomes in patients with variceal and Nonvariceal UGIB.MethodsPatients with AUGIB who were admitted to hospital from June 2015 to July 2017 were retrospectively enrolled.The patients were divided into the acute non-variceal upper gastrointestinal bleeding(ANVUGIB)group,the acute variceal upper gastrointestinal bleeding(AVUGIB)group and the full-cause group(ANVUGIB and AVUGIB),and were scored by using the AIMS65,GBS,and Rockall scores.The main outcomes are: 1)inhospital death(referring to death due to AUGIB during hospitalization);2)non-drug hemostasis intervention(any or more of the following: endoscopic,radiological or surgical intervention);3)Bleeding(refers to the occurrence of black stool or hematemesis after the first acute hemorrhage stopped during hospitalization);4)Blood transfusion(refers to the clinical infusion of red blood cells or plasma products during hospitalization).The area under the receiver operating characteristic curve(AUROC)and the optimal cut-off value of the three scoring systems were calculated for different outcomes.And the prognostic outcomes of each group were compared according to the risk stratification methods of the three scores.ResultsOf the 284 AUGIB patients enrolled in the study,233(82.0%)were ANVUGIB and 51(18.0%)were AVUGIB.10 patients(3.5%)died,48 patients(16.9%)required endoscopic hemostasis,18 patients(6.3%)required surgery to stop bleeding,9 patients(3.2%)required interventional hemostasis,and 5 patients(1.8%)required three lumens The second cystic tube was pressed to stop bleeding,54 cases(19.0%)rebleeded,and 130 cases(45.8%)required blood transfusion.(1)For the outcome of predicting in-hospital death,the AIMSOC scores of the AIMS65 scores were highest in the full-cause group,ANVUGIB group and AVUGIB group,respectively,0.955,0.969,0.885,and the best cut-off values were 2.5 points(the full-cause group: Sensitivity 100%,specificity 89%).(2)For the clinical outcome of predictive hemostasis intervention,the difference between the three scores of AUROC was not statistically significant in the full-cause group and ANVUGIB group.In the AVUGIB group,the three scores showed no statistical significance for the AUROC in predicting the outcome of the hemostasis intervention(P>0.05).(3)For the clinical outcome of rebleeding,the difference between the three scores of AUROC was not statistically significant in the full-cause group and ANVUGIB group.In the AVUGIB group,only the AIMS65 score was meaningful for predicting the AUROC of the rebleeding outcome(AUROC=0.735,P=0.006).(4)In the full-cause group and the ANVUGIB group,the AIMS65 score and the GBS score were higher in predicting the clinical outcome of blood transfusion.There was no statistical difference between the AUROC and the AUROC of the Rockall score(the fullcause group AUROC: AIMS65 0.810 GBS 0.841 Rockall 0.743,AIMS65 vs Rockall: p<0.05 GBS vs Rockall: p<0.05;ANVUGIB group AUROC: AIMS65 0.807 GBS 0.867 Rockall 0.743,AIMS65 vs Rockall: p<0.05 GBS vs Rockall: p<0.05).In the AVUGIB group,the difference between the three scores of AUROC was not statistically significant.(5)According to the three-strategy risk stratification method,the AIMS65 score and the Rockall score can accurately distinguish the high risk and low risk of in-hospital death,rebleeding,non-drug hemostasis,and four outcomes requiring blood transfusion in all AUGIB patients included in this study.However,GBS can only effectively distinguish between the high risk and low risk of the need for non-drug hemostasis intervention and blood transfusion,but cannot effectively stratify the occurrence of in-hospital death and rebleeding.ConclusionsThe AIMS65 score can better predict the risk of in-hospital death and the need for transfusion in patients with a full-cause group(non-variceal and variceal upper gastrointestinal bleeding).In the subgroup analysis,the AIMS65 score was more suitable for the prognosis evaluation of patients with non-variceal hemorrhage,and the prognosis evaluation of patients with variceal hemorrhage was slightly weaker.