摘要
目的 分析医院获得性急性肾衰竭(HA-ARF)的临床特征并构建其预警模型。方法回顾性分析65例HA-ARF患者及同期94例社区获得性急性肾衰竭(CA-ARF)患者的临床资料,分析HA-ARF发生的危险因素并构建预警模型,绘制ROC曲线评估模型的预测效能。结果HA-ARF病因主要为感染、手术相关及肾脏低灌注。HA-ARF组年龄以及肾前性因素、入住外科、多器官功能衰竭比例高于CA-ARF组,肾后性因素比例和死亡率低于CA-ARF组(P<0.05)。多因素logistic回归分析结果显示,高龄、入住外科、多器官功能衰竭为HA-ARF发生的独立危险因素(P<0.05)。建立预警模型:Logit(P)=-7.219+0.105×年龄+2.190×入住外科+2.679×多器官功能衰竭。该模型ROC曲线下面积为0.832[95%CI(0.766~0.898)],取最佳诊断界值为0.476时,该模型预测HA-ARF发生的灵敏度为70.82%,特异度为84.03%。结论 HA-ARF病因复杂,高龄、入住外科、多器官功能衰竭是其独立危险因素,据此构建的模型对HA-ARF发生具有较高的预测效能。
Objective To analyze the clinical characteristics of hospital-acquired acute renal failure(HA-ARF) and construct an early warning model.Methods The clinical data of 65 patients with HA-ARF(group A) and 94 patients with community-acquired acute renal failure(group C) were retrospectively analyzed.The risk factors for the occurrence of HA-ARF were analyzed and an early warning model of HA-ARF was constructed.The ROC curve was drawn to estimate the predictive performance of the model.Results HA-ARF was mainly caused by infection, surgery and renal hypoperfusion.The age and the proportions of the patients with prerenal factor, admission to surgical ward and multiple organ failure were higher, while the proportions of those with postrenal factor and mortality rate were lower in group A than those in group C(P<0.05).Multivariate logistic regression analysis showed that advanced age, admission to surgical ward and multiple organ failure were the independent risk factors for the occurrence of HA-ARF(P<0.05).The early warning model constructed was as Logit(P)=-7.219+0.105×age+2.190×admission to surgical ward+2.679×multiple organ failure.ROC curve analysis showed that the AUC of the model to predict HA-ARF was 0.832 [95%CI(0.766-0.898)].Taking the best cut-off value of 0.476,the sensitivity and specificity for the model to predict HA-ARF were 70.82% and 84.03%,respectively.Conclusion The cause of HA-ARF is complex.Advanced age, admission to surgical ward and multiple organ failure are the independent risk factors for the occurrence of HA-ARF.The model constructed based on these factors has higher predictive performance for that.
作者
尹帅
舒适
姚明明
尹春彪
YIN Shuai;SHU Shi;YAO Mingming(Department of Critical Care Medicine,Hangzhou Cancer Hospital,Hangzhou 310002,CHINA)
出处
《江苏医药》
CAS
2022年第8期768-772,共5页
Jiangsu Medical Journal
基金
杭州市医药卫生科技项目(A20220169)。
关键词
急性肾衰竭
预警模型
Acute renal failure
Early warning model
