ABSTRACT
Acute nonvariceal upper gastrointestinal bleeding (ANVUGIB)is a commonly seen gastrointestinal emergency.Rockall and Blatchford scoring system are commonly used for risk stratification in ANVUGIB.Aims:To investigate the predictive values of Rockall and Blatchford scoring system for assessing the risk of blood transfusion,surgical intervention and mortality in patients with ANVUGIB.Methods:Five hundred and ninety hospitalized patients with ANVUGIB were scored by Rockall and Blatchford scoring system,respectively.Predictive values of these two scoring systems for assessing the risk of blood transfusion,surgical intervention and mortality were assessed by area under the receiver operating characteristic (ROC)curve (AUC).Results:Rockall and Blatchford scores in patients with blood transfusion,surgical intervention and died were significantly higher than those in patients without blood transfusion,surgical intervention and survived (P <0.01 ).The AUC of Rockall scoring system for predicting blood transfusion,surgical intervention and mortality were 0.785 (95% CI:0.743-0.828,P=0.000),0.765 (95% CI:0.693-0.837,P=0.000),0.835 (95% CI:0.703-0.966,P=0.005),respectively.The AUC of Blatchford scoring system for predicting blood transfusion,surgical intervention and mortality were 0.812 (95%CI:0.775-0.848,P=0.000),0.870 (95%CI:0.811-0.930,P=0.000),0.784 (95% CI:0.614-0.954,P=0.017),respectively.Conclusions:Rockall and Blatchford scoring system have high predictive value for blood transfusion,surgical intervention and mortality in patients with ANVUGIB.Rockall scoring system is better for predicting mortality,while Blatchford scoring system is better for predicting blood transfusion and surgical intervention.
ABSTRACT
Objective To investigate the accuracy of prognosis risk assessment and clinical applicability of Rockall (RS) and Blatchford scoring system (BRS) for acute nonvariceal upper gastrointestinal bleeding (ANVUGIB). Methods From January 2009 to December 2009, the clinical date 195 ANVUGIB patients who met the standards with complete information and treated in The First Affiliated Hospital of Anhui Medical University were recorded. Each patient's scores of RS and BRS were calculated for risk stratification. Patients were followed up for 30 days after discharged.Death or the prognosis of disease in 30 days after discharged was considered as clinical study endpoints.Checked prognostic capacity of these two scoring system. Results In the 195 patients, there were 150 years, mean age was 53.97±18.34 years. 90 patients' age was over sixty (elderly group), 105 less than sixty (non-elderly group). 182 patients survived (93.3%), while 13 dead (6.7%). In survival patients, 11were re-bleeding (5.6 %). Mortality ( 12.2 %, 11/90), the percentage of patients with comorbidities (43.3%, 39/90) and taking aspirin (24. 4%, 22/90) were higher in elderly patients than non-elderly patients (1.9%, 2/105; 16.2%, 17/105; 11.4%, 12/105 respectively)(P＜0.05).The AUC of RS in predicting risk of death was 0.742 (P=0.004) and re-bleeding was 0.469 (P=0.101). For BRS score system, the AUC of predicting risk of death was 0. 493 (P= 0. 067)and rebleeding was 0.341(P=0.092). The RS score was positively correlated with length of hospital stay,however there was no statistically significant between BRS score and length of hospital stay.Conclusion RS score system was good at predicting the risk of death, and the score was positively correlated with length of hospital stay. While it was poor in predicting the risk of re-bleeding. BRS performed poorly in predicting the risk of both death and re-bleeding, so it was not suitable for predicting the risk of hospitalized patients.