GIÁ TRỊ TIÊN LƯỢNG CỦA THANG ĐIỂM OAKLAND Ở BỆNH NHÂN XUẤT HUYẾT TIÊU HÓA DƯỚI TẠI BỆNH VIỆN ĐA KHOA TRUNG ƯƠNG CẦN THƠ

Abstract

Background: Prognostic scoring systems for the clinical course of lower gastrointestinal bleeding (LGIB) help stratify patients and identify those at high risk, thereby facilitating early intensive monitoring and treatment, ultimately improving patient outcomes. Objective: To determine the prognostic value of Oakland score in LGIB. Materials and methods: A cross-sectional descriptive study with analytical components was conducted on 38 patients with LGIB admitted to Can Tho Central General Hospital. Results: The female-to-male ratio was 1.38, and the mean age was 58.68 ± 21.60 years, with the majority (79%) being over 40 years old. The prevalence of overweight and obesity was 15.8%. Regarding comorbidities, 65.7% had joint diseases, 39.5% had ischemic heart disease, and 36.8% had diabetes mellitus. Notably, 44.7% had a prior history of LGIB. The use of nonsteroidal anti-inflammatory drugs was high, accounting for 47.3%, while 42.1% were taking aspirin or clopidogrel. Regarding clinical outcomes, 63.2% required blood transfusions, 15.8% underwent endoscopic intervention alone, and 18.4% required both blood transfusion and endoscopic intervention. Severe clinical outcomes accounted for 31.6%, including 21.1% with rebleeding and two in-hospital mortality cases (5.3%). The mean Oakland score was 17.45 ± 5.3. At a cutoff of 26, the Oakland score demonstrated a mortality prediction sensitivity of 67%, specificity of 94%, and an area under the curve (AUC) of 98%. At a cutoff of 21, it showed a rebleeding prediction sensitivity of 80%, specificity of 79%, and an AUC of 79%. Conclusion: The Oakland score has a high discriminative value in predicting mortality at a cutoff of 26 and rebleeding at a cutoff of 21 in patients with LGIB.