Is the Promising AIMS-65 Score Really Useful? An European Center Experience

Abstract

Introduction: Current guidelines recommend risk assessment as an initial step in the management of patients with upper gastrointestinal bleeding. AIMS65 is a five factor score recently proposed to be used in the UGIB setting. The aim of this study was to compare the accuracy of AIMS65 score in predicting clinically relevant outcomes between variceal and non-variceal bleeding patients. Methods: Retrospective single-center study including 315 UGIB episodes between January 2011 and December 2014. For each patient clinical, demographic, laboratorial and endoscopic data were collected and AIMS65 score (albumin, INR, altered mental status, systolic blood pressure, and age older than 65 years) was calculated. The accuracy of AIMS65 score to predict clinical outcomes (30-day mortality, 30-day rebleeding, blood transfusion, endoscopic therapy, and surgery) was evaluated using area under the ROC curve (AUROC). Results: From the analyzed patients, 195 (61.9%) were men, and the mean age was 68 years. Variceal and non-variceal bleeding was confirmed in 75 (23.8%) and 186 (59.0%) patients, respectively. In 54 patients (17.1%) no bleeding origin was found during upper endoscopy. There were no significant differences in the mean AIMS65 score between the variceal and non-variceal bleeding group (1.81 vs 1.70; p=0.442). When applied to variceal bleeding patients, AIMS65 score could fairly predict 30-day mortality (AUROC=0.723), but not 30-day rebleeding (AUROC=0.529) and need for blood transfusion (AUROC=0.612). As all patients in this group received endoscopic therapy and any of them required a surgical intervention, AIMS65 score could not be assessed for these outcomes. Contrarily, when applied to non-variceal bleeding patients, AIMS65 score was not able to predict any of the studied outcomes, namely 30-day mortality (AUROC=0.604), 30-day rebleeding (AUROC=0.521), endoscopic therapy (AUROC=0.535), and need for blood transfusion (AUROC=0.657) or surgery (AUROC=0.586). Conclusion: Although AIMS65 score may be useful to predict mortality in patients with variceal bleeding, it is not a reliable score to predict other clinically relevant outcomes. Furthermore, this score seems to be useless in non-variceal bleeding cases, which may preclude its routine use for initial assessment of patients in the emergency department.