Prognostic Analysis Of Patients With Non-variceal Upper Gastrointestinal Bleeding When Taking Aspirin

Objective Aspirin,as a kind of nonsteroidal anti-inflammatory drugs,was first to be known to have analgesic and antipyretic properties.The use of aspirin has been a common preventive measure to reduce the risk of cadiovascular events.This is attributed to aspirin's ability to inhibit platelet activition,due to the gradually in-depth studies,we also found that aspirin can exert the function of anti-tumor through various pathways.And in the process of asprin use,upper gastrointestinal bleeding becomes the common and serious side effect of aspirin.Our aim is to discuss the clinical outcomes in patients with non-variceal upper gastrointestinal bleeding with aspirin use and those not.Methods We performed a retrospective study of patients with NVUGIB(defined by hematemesis,coffee-ground vomit or melena),who underwent gastroscopy from Jan.2016 to Apr.2019 in X Hospital.They were divided into two groups based on whether taking aspirin or not.We collected the general information of the patients,labor data,blood tansfusion units,endoscopic findings,etc.Compare the clinical characteristics(sex,age,endoscopic findings),risk scoring systerms(AIMS65 score,Glasgow-Blatchford score and the full Rockall score)and outcomes(in-hospital mortality,rebleeding rate,lenth of hospital stay and severityof bleeding)between two groups.Results Of the 197 patients were included,71 patients were undertaken NVUGIB when taking aspirin(observation gruop),and 126 patients were not(control).Patients treated with aspirin were older,the mean age is 65 years in the observation group and 54 years in control(P<0.05).Compared to the female,the man were more likely to have NVUGIB regardless of taking aspirin or not(observation gruop:male 76.10% vs female23.90% and male 84.90% vs female 15.10% for control group).There were more cardiocerebro-vascular comorbidities in observation gruop than control(P<0.05),there were 62(87.30%)patients in observation gruop and 30(23.80%)in control.The endoscopic findings in observation gruop were more likely to be gastric ulcer releted bleeing(28(39.40%)vs31(24.60%)),(P<0.05),and duodenal ulcer releted bleeding were more likely to happen in control(13(18.30%)vs55(43.70%)),(P<0.05).In predicting the need for blood transfusion,AUC(95%CI)control versus ATD: 0.74(0.63 to 0.86)vs0.83(0.72 to 0.94)for GBS,0.70(0.57 to 0.82)vs 0.80(0.66 to 0.93)for FRS and 0.69(0.55 to 0.82)vs 0.76(0.63 to 0.88)for AIMS65,(all P<0.05),The need for transfusion was best predicted by GBS in both groups.In predicting the rebleeding,AUC(95%CI)control versus ATD: 0.87(0.77 to 0.95)vs 0.79(0.66 to 0.91)for FRS,0.72(0.61 to0.83)vs 0.71(0.54 to 0.86)for GBS,and 0.64(0.52 to 0.79)vs 0.70(0.55 to 0.89)for AIMS65,(all P<0.05),rebleeding was best predicted by the Full Rockall in both groups.In predicting the in-hospital mortality,AUC(95%CI)control versus ATD:0.78(0.68 to 0.88)vs 0.79(0.00 to 1.00)for FRS,0.72(0.56 to 0.87)vs 0.98(0.00 to 1.00)for GBS,and 0.68(0.53 to 0.83)vs 0.95(0.00 to 1.00)for AIMS65,(all P<0.05),FRS was the strongest predictor for mortality in both groups.In predicting the severity of bleeding,AUC(95%CI)control versus ATD:GBS:0.69(0.59 to 0.78)vs0.79(0.70 to0.91);FRS:0.61(0.50 to 0.72)vs0.73(0.53 to 0.92);AIMS65:0.68(0.57 to0.78)vs0.77(0.61 to 0.92).,(all P<0.05),GBS was the best predictor on severity of bleeding.There were less in-hospital mortality in patients taking asprin than those not(P<0.05).There were 2 died in observation group and 14 died in control,and the length of hospital stay were shorter in observation group(P<0.05),the mean length of hospital stay in observation gruop was 7.85 days,and 9.12 days in the control.There were no stastistical significance on the rebleeding and severity of bleeding between two groups.Conclusion Compared to the NVUGIB patients not taking aspirin,patients with aspirin were older,had more cardiocerebro-vascular comorbidities,the predictive ability of three risk scoring system is not bad,less mortalities and shorter hospital stay,there were no significant differences in severity of bleeding and rebleeding between two groups.The clinical outcomes were not bad in patients taking aspirin,this may contribute to increase patient's compliance.