肺炎克雷伯菌尿路感染167例碳青霉烯耐药风险列线图预测模型的构建

目的 构建肺炎克雷伯菌尿路感染碳青霉烯耐药风险列线图预测模型。方法 回顾性分析2018年1月至2020年12月在秦皇岛市第一医院确诊为肺炎克雷伯菌尿路感染的成人住院病人167例临床资料，将尿液标本中检出耐碳青霉烯类肺炎克雷伯菌（CRKP）的62例病人设为CRKP组，非CRKP的105例病人设为非CRKP组。采用logistic回归分析发生CRKP尿路感染的独立危险因素，并将167例病人按照分层随机抽样以7∶3比例分为训练集（118例）和验证集（49例），然后使用训练集根据独立危险因素建立CRKP尿路感染列线图风险预测模型。采用校准曲线和受试者操作特征曲线（ROC曲线）评估列线图预测模型的准确度和区分度。结果 CRKP组肺炎克雷伯菌对头孢呋辛、哌拉西林/他唑巴坦、美罗培南、阿米卡星耐药性分别为100%(62/62)、100%(62/62)、98.4%(61/62)、51.6%(32/62)，明显高于非CRKP组的42.9%(45/105)、8.6%(9/105)、1.0%(1/105)、3.8%(4/105)。多因素logistic回归分析显示，入住重症监护室、两周内使用碳青霉烯和酶抑制...

Establishment of a nomogram model for carbapenem resistance in 167 cases of Klebsiella pneumoniae urinary tract infection

Objective To construct a predictive model for the risk of carbapenem resistance in Klebsiella pneumoniae urinary tract in. fection using a nomogram.Methods The clinical data of 167 adult patients diagnosed with Klebsiella pneumoniae urinary tract infec.tion in the First Hospital of Qinhuangdao City from January 2018 to December 2020 were retrospectively analyzed. Sixty-two patients with carbapenem-resistant Klebsiella pneumoniae (CRKP) detected in urine specimens were assigned to the CRKP group. One hundredand five patients without CRKP were assigned to the non-CRKP group. Logistic regression was used to analyze the independent risk fac.tors for CRKP urinary tract infection, and 167 patients were randomly divided into a training set (118 cases) and a validation set (49cases) according to stratified 7:3 ratio, and then the training set was used to establish a CRKP urinary tract infection risk predictionmodel based on the independent risk factors. Calibration curves and receiver operating characteristic (ROC) curves were used to evalu.ate the accuracy and discrimination of nomogram prediction models.Results The resistance of Klebsiella pneumoniae to cefuroxime,piperacillin/tazobactam, meropenem and amicacin in CRKP group was 100% (62/62), 100% (62/62), 98.4% (61/62) and 51.6% (32/62),respectively, which were significantly higher than those of the non-CRKP group 42.9% (45/105), 8.6% (9/105), 1.0% (1/105), 3.8% (4/105). Multivariate logistic regression analysis showed that admission to the intensive care unit, usage of carbapenem and enzyme inhibi.tors within two weeks were independent risk factors for urinary tract infections in CRKP (OR=8.95, 5.52, 6.12, P<0.05). The consisten. cy index (C-index) of the training set and the validation set were 0.88 and 0.90, respectively. The area under the ROC curve is 0.88.The calibration curve and ROC curve indicate that the model has good accuracy and good discrimination.Conclusions By analyzingthe risk factors of CRKP urinary tract infection, a risk nomogram prediction model was constructed. The model has good accuracy anddiscrimination, and can effectively predict the risk of urinary tract infection with carbapenem-resistant Klebsiella pneumoniae. Then more targeted protective measures and formulate reasonable drug treatment plans can be formulated.