术前尿培养阴性患者输尿管软镜碎石术后发热的危险因素分析及预测发热列线图模型的构建

Analysis of risk factors for postoperative fever in patients with negative preoperative urine culture after flexible ureteroscopy and construction of a nomogram model

目的按探讨术前尿培养阴性患者输尿管软镜碎石术(fURS)后发热的危险因素,并构建列线图模型预测术后发热的风险。方法回顾性分析2019年1月至2023年3月于郑州大学第二附属医院接受fURS且术前尿培养阴性的308例患者的临床资料。男235例,女73例;年龄(46.4±12.1)岁;合并高血压病86例,糖尿病41例,冠心病12例;57例有泌尿系结石手术史;91例有重度肾积水。结石位于左侧164例,右侧144例;肾结石88例,输尿管结石124例,肾输尿管结石96例。结石数量≤2枚者243例,>2枚者65例;结石最大径12.0(9.0,15.0)mm。尿白细胞阳性109例,阴性199例。尿亚硝酸盐阳性2例,阴性308例。记录患者术后48h内有无发热。比较发热组与未发热组一般资料的差异。采用单因素和多因素logistic回归分析fURS术后发热的危险因素。基于独立危险因素构建列线图预测模型,采用1000次Bootstrap重抽样进行内部验证,采用受试者工作特征曲线(ROC)和曲线下面积(AUC)评估其预测性能,采用校准曲线评估模型稳定性。结果308例手术顺利完成,手术时间60.0(40.0,75.0)min。完全清除结石221例。其中14例(4.5%)术后出现发热,294例术后未出现发热。发热组较未发热组的女性比例更高[57.1%(8/14)与22.1%(65/294),P=0.007],合并糖尿病比例更高[50.0%(7/14)与11.6%(34/294),P<0.001],肾结石比例更高[64.3%(9/14)与26.9%(79/294),P=0.022],术中完全清除结石比例更低[42.9%(6/14)与73.1%(215/294),P=0.031],结石最大径更大[15.5(12.519.3mm与11.0(9.0,15.0)mm,P=0.004],手术时间更长[87.5(58.8,106.3)min与55.0(40.0,75.0)min,P<0.001],血小板计数更高[267.0(225.8,354.0)×10^(9)/L与233.0(197.8,272.0)×10^(9)/L,P=0.026],肌酐更低[67.5(52.5,72.3 umol/L与73.0(62.0,84.0)μmol/L,P=0.026],血小板淋巴细胞比值更高[148.8(118.3,189.3)与119.5(93.2,156.0),P=0.030]。单因素logistic回归分析结果显示,女性、合并糖尿病、结石位置、不能完全清除结石、结石最大径、手术时间、血小板计数、肌酐、血小板淋巴细胞比值和尿亚硝酸盐阳性是术后发热的危险因素(均P<0.05)。多因素logistic回归分析结果显示,女性(0OR=11.073,95%CI 1.623~75.521,P=0.014)、合并糖尿病(0R=5.995,95%CI1.441~24.952,P=0.014)和手术时间长(0R=1.024,95%CI1.003~1.046,P=0.024)是术后发热的独立危险因素。根据危险因素构建的列线图模型表现出良好的预测性能(AUC=0.866,95%CI0.781~0.935)。校准曲线展示出较好的一致性。结论女性、手术时间长和合并糖尿病是术前尿培养阴性患者fURS术后发热的危险因素,由此构建的列线图具有良好的预测性能.

Objective To investigate the risk factors for postoperative fever in patients with negative preoperative urine culture undergoing flexible ureteroscopy(fURS),and construct a nomogram prediction model to predict the risk of postoperative fever.Methods The clinical data of 308 patients who underwent flexible ureteroscopy(fURS)at the Second Affiliated Hospital of Zhengzhou University from January 2019 to March2023,were retrospectively analyzed.Among these patients,there were 235 males and 73 females,with an average age of(46.4±12.1)years old.Additionally,86 cases had concomitant hypertension,41 cases had diabetes,and 12 cases had coronary heart disease.A history of urinary stone surgery was present in 57 cases,and 91 cases exhibited severe hydronephrosis.The distribution of stones included 164 cases on the left side and 144 cases on the right side,with 88 cases of renal stones,124 cases of ureteral stones,and 96 cases of renal-ureteral stones.Among them,243 cases had≤2 stones,while 65 cases had>2 stones,with a maximum stone diameter of 12.0(9.0,15.0)mm.Urine leukocyte-positive cases were 109,and urine leukocyte-negative cases were 199.Two cases were positive for nitrite,and 308 cases were negative.The occurrence of postoperative fever within 48 hours was recorded,and differences between the fever and non-fever groups were compared.Logistic regression analysis was employed to identify risk factors for post-fURS fever.A nomogram prediction model based on independent risk factors was constructed,and internal validation was conducted using 10o0 bootstrap resamples.The predictive performance of the model was evaluated using receiver operating characteristic(ROC)curves and the area under the curve(AUC).Model stability was assessed using calibration curves.Results The surgeries for all 308 cases were successfully completed with a median operative time of 60.0(40.0,75.0)minutes.Complete stone clearance was achieved in 221 cases.Among them,14 cases(4.5%)experienced postoperative fever,while 294 cases did not.The fever group had a higher proportion of females[57.1%(8/14)vs.22.1%(65/294),P=0.007],more cases with comorbid diabetes[50.0%(7/14)vs.11.6%(34/294),P<0.001],a higher proportion of renal stones[64.3%(9/14)vs.26.9%(79/294),P=0.022],a lower intraoperative stone clearance rate[42.9%(6/14)vs.73.1%(215/294),P=0.031],larger stone diameter[15.5(12.5,19.3)mm vs.11.0(9.0,15.0)mm,P=0.004],longer operative time[87.5(58.8,106.3)min vs.55.0(40.0,75.0)min,P<0.001],higher platelet count[267.0(225.8,354.0)×10^(9)/L vs.233.0(197.8,272.0)×10^(9)/L,P=0.026],lower creatinine levels[67.5(52.5,72.3)umol/L vs.73.0(62.0,84.0)umol/L,P=0.026],and a higher platelet lymphocyte ratio[148.8(118.3,189.3)vs.119.5(93.2,156.0),P=0.030].Results of univariate analysis showed that female gender,diabetes,stone location,incomplete stone clearance,maximum stone diameter,operative time,platelet count,creatinine,platelet lymphocyte ratio,and positive nitrite in urine(all P<0.05)were risk factors for postoperative fever.Multivariate regression analysis revealed that female gender(0R=11.073,95%CI 1.623-75.521,P=0.014),diabetes(0R=5.995,95%CI 1.441-24.952,P=0.014),and operative time(0R=1.024,95%CI 1.003-1.046,P=0.024)were independent risk factors for post-fURS fever.The nomogram exhibited excellent predictive performance(AUC=0.866,95%CI0.781-0.935),and the calibration curve demonstrated good consistency.Conclusions Female gender,longer operative time,and diabetes are risk factors for post-fURS fever in patients with preoperative negative urine culture.The nomogram demonstrates excellent predictive performance.