Predictive Value Of Different Scoring Systems For Risk Stratification In Patients With Acute Non-variceal Upper Gastrointestinal Bleeding

Objective: To compare the value of Glasgow-Blatchford bleeding score(GBS),AIMS65 score and Rockall score systems for the prediction of outcome in ANVUGIB patients,and to guide clinical risk stratification.Methods: A retrospective study was conducted on 246 patients with ANVUGIB who were admitted to the Department of Emergency Medicine,Tianjin Medical University General Hospital from January 01,2013 to December 31,2015.The clinical data(name,sex,age,past history,current history,medication history,vital signs,biochemical indexes,gastroscopy,imaging examination,etc.)were collected within24 hours after admission and after hospitalization.Two hunderd and forty-six ANVUGIB patients was score by Blatchford scoring system,AIMS65 scoring system,Rockall scoring system.Then compare the value of Glasgow-Blatchford bleeding score(GBS),AIMS65 score and Rockall score systems for the prediction of outcome in ANVUGIB patients.SPSS23.0 software was used for statistical analysis,The receiver operating characteristic curve(ROC)was used to evaluate the sensitivity and specificity of the Blatchford scoring system,the AIMS65 scoring system,the Rockall scoring system.Predictive values of those scoring systems for assessing the risk of blood transfusion,rebleeing,intervention,and death by area under the receiver operating charactertic curve(AUC).Result: of the 246 patients,95(38.62%)received transfusion,15(6.10%)had rebleeding,and 4(1.62%)died.In terms of predicting the probability of needed transfusion in the ANVUGIB patients,the AUC of Blatchford,AIMS65 and Rockall scoring systems were 0.850[95% CI(0.796,0.904),P<0.001 ],0.781[95% CI(0.714,0.847),P<0.001],0.745[95% CI(0.673,0.817),P<0.001],rspectively;in regard to predicting the probability of rebleeding in the ANVUGIB patients,the AUC of Blatchford,AIMS65 and Rockall scoring systems were 0.900 [ 95%CI(0.843,0.958),P<0.001],0.842[95% CI(0.737,0.946),P<0.001],0.929[95% CI(0.877,0.982),P<0.001],respectively;in the aspect of predicting the probability of intervention in the ANVUGIB patients,the AUC of Blatchford,AIMS65 and Rockall scoring systems were 0.963[95% CI(0.927,0.996),P=0.002 ],0.802[95% CI(0.607,0.997),P=0.039],0.892[95% CI(0.766,0.998),P=0.007],respectively;in the aspect of predicting the probability of death in the ANVUGIB patients,the AUC of Blatchford,AIMS65 and Rockall scoring systems were 0.956[95% CI(0.893,0.999),P=0.002],0.919[95% CI(0.828,0.999),P=0.004],0.992[95% CI(0.766,0.998),P=0.004],respectively.According to the optimal cut-off point of Blatchford score,when the Blatchford score ≧ 7 points,ANVUGIB patients may have higher risk outcome by transfusion,rebleeding,intervention and death;according to AIMS65 score optimal cut-off point,when the AIMS65 score ≧ 2 points,ANVUGIB patients may have higher risk outcome by transfusion,rebleeding,intervention and death;according to Rockall score the best cut-off point,when the Rockall score ≧ 3 points,ANVUGIB patients have a higher risk of transfusion,rebleeding,intervention and death outcomes.Conclusion: To guide the risk stratification of ANVUGIB patients,Blatchford scoring system is more suitable for emergency assessment of ANVUGIB patients,and can better improve the prognosis of patients with ANVUGIB.When the GBS score≦ 2,can be temporarily considered as low-risk patients,these low-risk patients can receive outpatient treatment.