A Prospective, Multicenter Study of the AIMS65 Score Compared with the Glasgow-Blatchford Score in Predicting Upper Gastrointestinal Hemorrhage Outcomes

Background: The AIMS65 score and the Glasgow-Blatchford risk score (GBRS) are validated preendoscopic risk scores for upper gastrointestinal hemorrhage (UGIH). Goals: To compare the 2 scores' performance in predicting important outcomes in UGIH. Study: A prospective cohort study in 2 tertiary referral centers and 1 community teaching hospital. Adults with UGIH were included. The AIMS65 score and GBRS were calculated for each patient. The primary outcome was inpatient mortality. Secondary outcomes were 30-day mortality, in-hospital rebleeding, 30-day rebleeding, length of stay, and a composite endpoint of in-hospital mortality, transfusions, or need for intervention (endoscopic, radiologic, or surgical treatment). The area under the receiver operating characteristic curve (AUROC) was calculated for each score and outcome. Results: A total of 298 patients were enrolled. The AIMS65 score was superior to the GBRS in predicting in-hospital mortality (AUROC, 0.85 vs. 0.66; P<0.01) and length of stay (Somer's D, 0.21 vs. 0.13; P=0.04). The scores were similar in predicting 30-day mortality (AUROC, 0.74 vs. 0.65; P=0.16), in-hospital rebleeding (AUROC, 0.69 vs. 0.62; P=0.19), 30-day rebleeding (AUROC, 0.63 vs. 0.63; P=0.90), and the composite outcome (AUROC, 0.57 vs. 0.59; P=0.49). The optimal cutoffs for predicting in-hospital mortality were an AIMS65 score of 3 and a GBRS score of 10. For predicting rebleeding, the optimal cutoffs were 2 and 10, respectively. Conclusions: The AIMS65 score is superior to the GBRS for predicting in-hospital mortality and hospital length of stay for patients with UGIH. The AIMS65 score and GBRS are similar in predicting 30-day mortality, rebleeding, and a composite endpoint.