ABSTRACT
Objective:To explore the influencing factors of severity of upper gastrointestinal bleeding (UGIB) and to establish the early warning evaluation model in the form of line chart, so as to provide a feasible basis for emergency nurses' triage.Methods:A total of 680 UGIB patients admitted to the Emergency Department of the First Affiliated Hospital of Wenzhou Medical University from January 2019 to January 2020 were retrospectively analyzed. They were divided into a modeling group ( n=510) and a validation group ( n=170) by random number table method, and were divided into a high-risk group and a low-risk group according to the expert Consensus on Emergency Diagnosis and Treatment Procedures for Acute Upper Gastrointestinal Bleeding in 2020. The differences of various indicators between groups were compared, the factors affecting the severity of the disease were analyzed by Logistic regression, and the nomogram was drawn and validated. Results:Multivariate logistic regression analysis showed that hematemesis ( OR=3.875, 95% CI: 2.212-6.79), diabetes ( OR=2.64, 95% CI: 1.184-5.883), syncope ( OR=10.57, 95% CI: 3.675-30.403), heart rate ( OR=3.262, 95% CI: 1.753-6.068), red blood cell distribution width ( OR=3.904, 95% CI: 2.176-7.007), prothrombin time ( OR=3.665, 95% CI: 1.625-8.269), lactic acid ( OR=3.498, 95% CI: 1.926-6.354) and hemoglobin ( OR=4.984, 95% CI: 2.78-8.938) were the influencing factors of the severity of UGIB patients ( P < 0.05). The nomogram model showed good consistency and differentiation (C-index=0.903, 95% CI: 0.875-0.931), and was verified internally (C-index=0.895) and Hosmer-Lemeshow goodness-of-fit test ( P=0.7936). Externally verified C-index was 0.899 (95% CI: 0.846-0.952). The calibration curve prompt warning evaluation model had good stability and the prediction efficiency was better than the modified early warning score ( P < 0.05). Conclusions:The early warning evaluation model has a reliable predictive value, which can provide a reference for emergency medical staff to screen high-risk patients and formulate targeted nursing interventions.