Xu Changhai, Wang Xueying, Wu Haibo, Li Wei, Lin Fei, Lin Na, Shen Shiyin, Pan Shubin, Chen Tong, Zhang Donghui, He Long, Cui Yan
Intensive Care Medicine Department, General Hospital of Northern Theater Command of Chinese PLA, 110000 Shenyang, Liaoning, China.
Urinary Surgery, General Hospital of Northern Theater Command of Chinese PLA, 110000 Shenyang, Liaoning, China.
Arch Esp Urol. 2025 Jul;78(6):758-765. doi: 10.56434/j.arch.esp.urol.20257806.101.
This study explored infection risk factors within one month post-kidney transplantation (KT) and developed a clinical prediction model.
We retrospectively analyzed clinical data from KT patients treated at our hospital (January 2015-December 2024). Patients were categorized into infection or control groups based on 1-month postoperative infection status. Infection incidence and risk factors were analyzed, and a multivariate logistic regression model was developed. The receiver operating characteristic (ROC) curves and a nomogram were generated. Patients were randomly split 7:3 into training and validation sets to assess model performance.
A total of 410 patients were included in this study, of whom 131 had postoperative infection, with an incidence rate of 31.95%. Multivariate logistic regression analysis showed that history of smoking (odds ratio (OR) = 2.96, 95% confidence interval (CI) (1.20-7.27)), drainage tube indwelling time (OR = 1.41, 95% CI (1.17-1.71)), catheter indwelling time (OR = 1.66, 95% CI (1.36-2.03)) and albumin (ALB) (OR = 0.78, 95% CI (0.71-0.86)) and haemoglobin (HGB) (OR = 0.70, 95% CI (0.59-0.83)) levels were independent risk factors for early infection after KT ( < 0.05). The area under the ROC curve of the training set was 0.954 (95% CI (0.925-0.982)), the specificity was 0.855 and the sensitivity was 0.896. In the validation set, the area under the ROC curve was 0.914 (95% CI (0.861-0.967)), the specificity was 0.832 and the sensitivity was 0.903. The Hosmer-Lemeshow goodness-of-fit test of the model showed that the training set χ = 6.962 ( = 1.000) and the validation set χ = 8.813 ( = 0.450). Multivariate risk factors were used to construct a nomogram model, and the calibration curve was consistent with the ideal curve, suggesting that the model had good stability. The clinical decision curve showed that it had good clinical value.
History of smoking, drainage tube indwelling time, catheter indwelling time and ALB and HGB levels are the risk factors of infection after KT. The model based on these factors can effectively predict the occurrence of infection after KT.
本研究探讨肾移植（KT）术后1个月内的感染危险因素，并建立临床预测模型。
我们回顾性分析了我院2015年1月至2024年12月接受KT治疗的患者的临床资料。根据术后1个月的感染状况将患者分为感染组和对照组。分析感染发生率和危险因素，并建立多因素逻辑回归模型。生成受试者工作特征（ROC）曲线和列线图。将患者按7:3随机分为训练集和验证集，以评估模型性能。
本研究共纳入410例患者，其中131例术后发生感染，发生率为31.95%。多因素逻辑回归分析显示，吸烟史（比值比（OR）=2.96，95%置信区间（CI）（1.20 - 7.27））、引流管留置时间（OR = 1.41，95% CI（1.17 - 1.71））、导尿管留置时间（OR = 1.66，95% CI（1.36 - 2.03））以及白蛋白（ALB）（OR = 0.78，95% CI（0.71 - 0.86））和血红蛋白（HGB）（OR = 0.70，95% CI（0.59 - 0.83））水平是KT术后早期感染的独立危险因素（<0.05）。训练集的ROC曲线下面积为0.954（95% CI（0.925 - 0.982）），特异性为0.855，敏感性为0.896。在验证集中，ROC曲线下面积为0.914（95% CI（0.861 - 0.967）），特异性为0.832，敏感性为0.903。模型的Hosmer-Lemeshow拟合优度检验显示，训练集χ = 6.962（ = 1.000），验证集χ = 8.813（ = 0.450）。使用多因素危险因素构建列线图模型，校准曲线与理想曲线一致，表明该模型具有良好的稳定性。临床决策曲线显示其具有良好的临床价值。
吸烟史、引流管留置时间、导尿管留置时间以及ALB和HGB水平是KT术后感染的危险因素。基于这些因素的模型可以有效预测KT术后感染的发生。
