RÉSUMÉ
Objective:To study the factors influencing repeatedly hospitalization in patients with acute pancreatitis (AP), and to analyse the predictive value of triglyceride for repeated hospitalization.Methods:The clinical data of 1 958 patients with AP treated at the First Affiliated Hospital of Anhui University of Science and Technology from January 2012 to April 2022 were analyzed. Of 1 733 AP patients who were enrolled, there were 1 000 males and 733 females, with mean ± s. d age being (49.4±16.4) years. Patients were grouped based on their ID numbers to determine their number(s) of hospitalization. Those who were admitted only once were included in the initial hospitalization group ( n=1 030), and those who were admitted twice or more were included in the repeated hospitalization group ( n=703). The factors influencing repeated hospitalization were analyzed by univariate analysis and multivariate logistic regression analysis. The predictive value of triglyceride for repeated hospitalization was evaluated by receiver operating characteristic (ROC) curve. Results:Multivariate logistic regression analysis showed that hypertriglyceridemia ( OR=1.445, 95% CI: 1.144-1.825, P=0.002) and biliary causes ( OR=3.184, 95% CI: 1.978-5.125, P<0.001) were independent risk factors for repeated hospitalization. When triglyceride <10.9 mmol/L, the prediction of AP patients without repeated hospitalization was 90.6%. The area under the ROC curve was 0.589, and the Yoden index was 0.170. Conclusion:Hypertriglyceridemia was risk factor for repeat hospitalization in AP patients and the efficacy of triglyceride in predicting repeat hospitalization in AP patients was good.
RÉSUMÉ
Cascading overload failures are widely found in large-scale parallel systems and remain a major threat to system reliability; therefore, they are of great concern to maintainers and managers of different systems. Accurate cascading failure prediction can provide useful information to help control networks. However, for a large, gradually growing network with increasing complexity, it is often impractical to explore the behavior of a single node from the perspective of failure propagation. Fortunately, overload failures that propagate through a network exhibit certain spatial-temporal correlations, which allows the study of a group of nodes that share common spatial and temporal characteristics. Therefore, in this study, we seek to predict the failure rates of nodes in a given group using machine-learning methods. We simulated overload failure propagations in a weighted lattice network that start with a center attack and predicted the failure percentages of different groups of nodes that are separated by a given distance. The experimental results of a feedforward neural network (FNN), a recurrent neural network (RNN) and support vector regression (SVR) all show that these different models can accurately predict the similar behavior of nodes in a given group during cascading overload propagation.
