急诊创伤患者亚谵妄发生预测模型的构建及验证研究

Construction and validation of prediction model of subsyndromal delirium in emergency trauma patients

目的探讨急诊创伤患者亚谵妄发生的影响因素,并对其发生风险进行预测。方法收集2020年1月—12月山东省某三级甲等综合性医院急诊ICU收治的创伤患者资料,采用单因素分析和多因素Logistic回归分析亚谵妄发生的风险因素,构建列线图预测模型,采用C指数验证模型区分度。结果共纳入306例患者,98例发生亚谵妄,发生率为32.0%。Logistic回归分析显示,年龄(OR=1.151)、急性生理与慢性健康评分(OR=1.314)、入住ICU时间(OR=1.881)、手术时间(OR=1.976)、术后疼痛评分(OR=4.314)、人文环境维度得分(OR=1.349)、自身感受维度得分(OR=1.413)是创伤患者发生亚谵妄的独立危险因素。将上述因素建立个体化列线图预测模型,C指数为0.898[95%CI(0.858,0.932)],校正曲线拟合良好。结论本研究建立的列线图预测模型有助于临床工作人员筛选发生亚谵妄的高危创伤患者,为优化急诊创伤患者谵妄的管理提供参考依据。

Objective To investigate the influencing factors of subsyndromal delirium in emergency trauma patients and to predict its risk.Methods The data of patients admitted to the emergency trauma ward of a tertiary first-class general hospital in Shandong province from January to December 2020 was collected.The risk factors of subsyndromal delirium were analyzed by univariate analysis and multivariate logistic regression.A nomogram prediction model was constructed,and the C-index was used to verify the discrimination of the model.Results A total of 306 patients were included,including 98 cases of subsyndromal delirium,the incidence was 32.0%.Logistic regression analysis showed that age(OR=1.151),APACHEⅡscore(OR=1.314),length of stay in ICU(OR=1.881),operation length(OR=1.976),postoperative NRS score(OR=4.314),the total score of humanistic environment dimension(OR=1.349)and total score of self-perception dimension(OR=1.413)were independent risk factors for subsyndromal delirium in trauma patients.The above independent risk factors were included and the individual nomogram prediction model was established.The C-index was 0.898[95%CI(0.858,0.932)],and the calibration curve fitted well.Conclusion The nomogram prediction model established in this study is helpful for early clinical identification of high-risk patients with subsyndromal delirium,and it provides basis for optimizing the management of delirium in patients with severe trauma.