严重颅脑损伤患者应激性高血糖列线图预测模型构建与验证OA

Objective To constract and validate a nomogram prediction model of stress hyperglycemia(SHG)in patients with severe traumatic brain injury.Methods A total of 262 patients with severe traumatic brain injury admitted to Affiliated People's Hospital of Jiangsu University from January 2020 to August 2024 were selected as study subjects.Based on whether SHG occurred,the subjects were divided into SHG group(n=111)and non-SHG group(n=151).Univariate and multivariate Logistic regression analysis were performed to identify characteristic variables,and a regression model was constructed.The model's efficacy was validated using the Bootstrap resampling method.Results Among 262 patients with severe traumatic brain injury,111 patients developed SHG,yielding an incidence rate of 42.37%.Hemoglobin levels,glucose administration,mechanical ventilation use were the risk factors influencing the occurrence of SHG.The constructed nomogram model was found to have an area under the receiver operating characteristic curve of 0.724(95%CI:0.662-0.786).The calibration curve analysis demonstrated excellent consistency between predicted probabilities and actual observed probabilities,while the decision curve analysis revealed that the model exhibits significant clinical net benefit.Conclusion This study constructed a risk prediction model for SHG in patients with severe traumatic brain injury comprising three factors,with the nomogram model demonstrating high accuracy,discrimination power,and clinical applicability.