NGHIÊN CỨU GIÁ TRỊ MỘT SỐ THANG ĐIỂM TRONG TIÊN LƯỢNG CHẢY MÁU TIÊU HÓA TRÊN KHÔNG DO TĂNG ÁP LỰC TĨNH MẠCH CỬA Ở BỆNH NHÂN CAO TUỔI

Abstract

Objectives: To evaluate the value of the Glasgow Blatchford (GB), clinical Rockall, and AIMS65 scoring systems in predicting the risk of rebleeding and in-hospital mortality in elderly patients with non-variceal upper gastrointestinal bleeding (UGIB). Methods: A prospective, cross-sectional descriptive study was conducted on 192 patients aged ≥ 60 with non-variceal UGIB at the Department of Gastroenterology, Military Hospital 103 and Military Hospital 354. Scores for GB, clinical Rockall, and AIMS65 were calculated and compared with the rates of rebleeding and in-hospital mortality. Results: All three scoring systems had predictive value of rebleeding, with AIMS65 showing the best predictive value (AUC = 0.925) at a cut-off point of 2, a sensitivity of 100% and specificity of 80.4%. The GB and clinical Rockall scores had lower AUC values of 0.837 and 0.815, respectively. For predicting mortality, both AIMS65 and GB scores demonstrated very good predictive values (AUC value of 0.945 and 0.901, respectively); however, AIMS65 at a cut-off point of 2 with a sensitivity of 93.8% and specificity of 84.1% was higher than that of the GB score. Conclusion: AIMS65 is a highly valuable scoring system for predicting the risk of rebleeding and in-hospital mortality in elderly patients with non-variceal UGIB and therefore, should be recommended for use.