Acute upper gastrointestinal bleeding (UGIB) is a common cause for attendance to the Emergency Department with a wide range of clinical severity, ranging from insignificant to life-threatening. While there is robust data to support the benefit of upper endoscopy within 24 hours of admission, the implementation of early upper endoscopy while patients are still in the emergency room has not been widely accepted due to lack of added benefit in terms of patient outcome such as mortality and re-bleeding rate. However, the use of upper endoscopy in the emergency room with the purpose of facilitating early discharge of low risk patients with upper gastrointestinal bleeding has not been studied.

The Glasgow Blatchford Score (GBS) was developed in 2000 to identify very low risk patients who would not need any intervention and were of low risk of rebleeding and death. The GBS used objective clinical parameters that could be easily obtained in the emergency department and did not use any endoscopy findings as parameters. Unlike the Rockall score which was designed to identify high risk patients, the GBS was used to identify low risk patients who could be safely discharged from the emergency department without endoscopy and studies have shown it to be superior in this regard. The main limitation of GBS is its low specificity with only 4-8% of all patients presenting with upper gastrointestinal bleeding stratified as low risk. Furthermore, the GBS was derived from a Scottish gastrointestinal bleeding registry and may not be applicable to the local population.

Suitable patients attending the Accident & Emergency Department for symptoms of upper gastrointestinal bleeding will be identified and recruited by the Accident & Emergency physicians. Patients presenting to Endoscopy Centre after admission will also be recruited to collect clinical data.

Univariate analysis was carried out on the development set using Pearson's chi-square method to examine the association among the factors on the outcome. Variables significantly associated with 30 day re-attendance rate in univariate analyses (P<=0.1) will be entered in multivariate logistic regression models. Risk factors which retained significance in multivariate analyses will be selected for incorporation into the risk score. A weighting will be assigned to each independent variable in the risk score, applying the corresponding adjusted odds ratio (AOR). The risk score for each subject is the sum of all the risk factors. To evaluate the predictive ability of the scoring system, a receiver operating characteristic (ROC) curve was constructed and the area under the curve (AUC) was delineated. A concordance (c)-statistics was used to reflect the discriminative ability of the prediction tool.

Cost-effective economical model will be used in cost effective analysis of the use of early discharge strategy.

Inclusion Criteria:

• All patients attending either the Prince of Wales Hospital or Alice Ho Miu Ling Nethersole hospital with symptoms of upper gastrointestinal bleeding will be initially resuscitated and stabilized in the Accident & Emergency Department (AED) as per local practice

Exclusion Criteria:

1. Incapable of providing informed consent
2. Age <18 years old
3. Pregnancy
4. Medically unfit to receive an upper endoscopy procedure (require O2 supplement >3L/min, persistent haemodynamic instability despite initial resuscitation, mental confusion)
5. Require admission to intensive care unit
6. Require emergency endoscopy for uncontrolled gastrointestinal bleeding