Prospective Study Of Early Risk Factors For High-risk Non-variceal Upper Gastrointestinal Bleeding

Background:Acute non-variceal upper gastrointestinal bleeding is one of the most common clinical critical diseases.Although there are already many scoring systems about the integrated clinical indicators and the results of endoscopy,studies on the scoring systems still have a great value because of the complexity or inaccuracy in the currently existing scoring systems.Objectives:To get some simple and accurate pre-endoscopic risk factors or high-risk scoring system for early detection of high-risk non-variceal upper gastrointestinal bleeding found to help clinicians accurately and quickly screen out the high risk of non-variceal upper gastrointestinal bleeding patient.Methods:A total of 622 NVUGIB patients between January 2015 and December 2018 were prospectively studied.The clinical characteristics and laboratory data were statistically analyzed.The severity of NVUGIB was based on high-risk NVUGIB(Forrest?a-?b),and low-risk NVUGIB(Forrest ?c and ?).We used to obtain some simpler and more accurate risk factors for early non-variceal upper gastrointestinal bleeding by univariate analysis(t-test,chi-square test),logistic regression,and receiver operating characteristic curve(ROC curve)and other statistical analysis methods.According to the specificity and sensitivity of each risk factor,different score systems of conbination were composed,and the best score systems was selected for clinical use.Results:By single factor analysis and Binary Logistic regression analysis*We found that hemoglobin,red haematemesis,red stool can predict high-risk NVUGIB patients.In order to determine the cutoff point of hemoglobin,using ROC curve analysis,the results showed that the Hb?83g/L can predict high-risk NVUGIB.Thus,Hb?83g/L,the red haematemesis,red stool was choose to be the risk factors for our new risk scoring system.the specificities and sensitivities of ?Hb?83g/L,?red haematemesis,?red stool were 66.3%?86.4%?96.4%and 61.7%?41.1%?18.3%.we define the color of haematemesis and melena was scored 1 when it was red hematemesis,red stool(dark red,bright red or containing red clot),and scored 0 when presented as other presentations.The??????groups were combined to form two new risk scoring systems,which were defined as score1 and score2.The above two risk scoring systems and the Blatchford risk system(BRS)were compared with each other by ROC curves.The specificity,sensitivity and AUROC of high-risk patients of score1 and score2 predict were higher than BRS.Therefore,these two scoring systems can well distinguish between high and low risk NVUGIB patients.Among the two systems,score2 has the best accuracy.The area under the ROC curve was 0.720,with sensitivity 0.775,specificity 0.518 in score2 when the cutoff point was 0.5.The new scoring system need only three factors,a hematological indicator and two clinical manifestations to obtain results,which are high accuracy,short time-consuming,easy to remember,and easy to promote.Conclusion:In our study,we used a prospective study to screen for "Hb?83g/L,?red haematemesis,?red stool" as three early factors for high-risk NVUGIB.The combination of the three factors as new scoring systems can better predict early high-risk NVUGIB patients.