摘要
We read with great interest the study by Rao et al.[1]Rao and colleagues showed that the Rockall score,Glasgow-Blatchford score(GBS)and AIMS65 are poorly predictive of 30-day mortality,or the need for endoscopic intervention in intensive care unit(ICU)patients with an upper GI bleed(UGIB).The Rockall[2]score and AIMS65[3]were developed for predicting mortality,whereas the GBS[4]was developed for predicting composite outcomes(The need for a blood transfusion or intervention to control bleeding,rebleeding,or death)We were not surprised to see that these scores were poor predictors of endoscopic intervention,as they were not designed to predict high-risk endoscopic stigmata(HRS),which is arguably a very meaningful endpoint to determine the need for endoscopic intervention.
