Abstract :
Background: The need for therapeutic endoscopy in patients with upper GI Bleeding (UGIB) is important
in determining the risk and disposition of these patients. Pre-endoscopic risk scores may be helpful in predicting
this need.
Methods: This is a prospective (cohort) study was conducted at Bhumibol Adulyadej Hospital. 212
patients who were admitted to the hospital with a principal diagnosis of UGIB between January 2012 - December
2012 and who arrived at the endoscopy room were recruited in the study.
Results: We identified 212 patients with the following characteristics: mean age of 58.4 years (SD, 16.1
years), 29.7% women, 17.5% with a history of liver disease, and 22.2% with history of previous gastrointestinal
bleeding. There were 55 patients (25.9%) need endoscopic therapy. Of these, two patients (0.9%) underwent surgery.
Rebleeding was found in 4 patients (1.9%), and 2 patients (0.9%) were rehospitalized for UGIB within 30
days after admission. By ROC curve analysis, the post-endoscopy Rockall score (RS) was superior to the Glasgow
Blatchford Score (GBS) for prediction of intervention (AUC 0.75 (95% CI 0.68-0.82) vs. 0.53 (95% CI (0.44-0.61)
and the AUC for GBS was inferior to pre-endoscopic RS (AUC 0.61 (95% CI 0.52-0.69). The specificity of GBS
and RS was suboptimal at all potential decision thresholds.
Conclusions: Glasgow Blatchford Score of 0 predicts low-risk patients who will not require an immediate
EGD for endoscopic hemostasis with 100% accuracy. The specificity of GBS and RS was insufficient to
recommend use of either score in clinical practice. |