Pseudo-observation regression for sequentially truncated data.

Upper gastrointestinal bleeding (UGIB) is a significant cause of hospital admissions. It is important to safely triage these patients in the emergency department and determine which patients are at risk of adverse outcomes. Several risk classifications have been developed to predict clinically important outcomes, such as mortality, the need for hospitalization, the risk of rebleeding, and the need for endoscopic intervention. In emergency departments, these scores are both useful and easy to calculate for the purpose of early and accurate risk assessment [, ]. Scores that can be calculated in emergency departments, such as the Glasgow-Blatchford score (GBS), Rockall score, and AIMS-65 score, are frequently used in clinics. These scores are useful not only for identifying patients who can be treated on an outpatient basis but also for identifying critically ill patients who require intensive care admission or endoscopic intervention. However, due to the difficulty of using these scores in routine clinical practice, simpler scores have been developed [–]. The 5-step Manchester Triage System (MTS) is a widely used triage tool for patients presenting to the emergency department. It has been reported that the MTS can assist clinicians in determining the urgency of endoscopic evaluation in patients with suspected upper gastrointestinal bleeding (UGIB) [].