恒压灌注下微通道经皮肾镜取石术后尿脓毒血症风险评分系统的初步建立

Establishment of a risk assessment scoring system for urosepsis after MPCNL under constant pressure perfusion

目的通过分析术前危险因素建立风险评分系统预测恒压灌注下微通道经皮肾镜取石术(MPCNL)术后发生尿脓毒血症的风险。方法选取遵义市播州区人民医院泌尿外科2018年12月至2020年12月行恒压灌注下MPCNL术的422例患者作为队列人群,以术后发生尿脓毒血症的26例患者为病例组,采用巢式病例对照研究方法,以1∶5的比例匹配同期对照组130例。进行Logistic回归分析确定预测因素,并按照逻辑评分法构建恒压灌注下MPCNL术后尿脓毒血症危险评分系统。从上述模型样本中随机抽取50%的病例(78例)为验证样本行内部验证。利用Hosmer-Lemeshow拟合优度检验验证评分系统的校准度,用受试者工作特征(ROC)曲线验证评分系统的区分度,并根据最大约登指数进行危险度分组。结果多因素Logistic回归分析显示,恒压灌注下MPCNL术后尿脓毒血症危险评分模型有4个预测因子,包括女性(Woman)(β=1.146,P=0.030)、尿白细胞(WBC)≥2+(β=1.088,P=0.047)、尿亚硝酸盐(NIT)阳性(β=2.653,P=0.000)、中性粒细胞与淋巴细胞比值(NLR)≥2.5(β=1.459,P=0.008)。用上述4个因素依据逻辑评分法构建恒压灌注下MPCNL术后尿脓毒血症危险评分系统(简称“WaN-2评分系统”):女性(1分),尿WBC≥2+(1分),尿NIT阳性(2分),NLR≥2.5(1分),总分为5分;分值≥2分为高危人群。WaN-2评分系统ROC曲线下面积为0.831[95%CI(0.740-0.921),P=0.000];Hosmer-Lemshow拟合优度检验显示,模型预测值与实际观测值之间的差异无统计学意义(χ^(2)=1.124,P=0.771)。从上述模型样本中随机抽取50%的病例(78例)为验证样本行内部验证,验证组证实,该评分系统ROC曲线下面积为0.826[95%CI(0.712-0.941),P=0.000];Hosmer-Lemshow拟合优度检验显示,模型预测值与实际观测值之间的差异无统计学意义(χ^(2)=0.884,P=0.643);结果表明该风险评分系统具有良好的区分度和校准度。结论本课题所建立的WaN-2评分系统可初步评估恒压灌注下行MPCNL手术患者术后并发尿脓毒血症的风险程度,对于WaN-2评分≥2分的高危患者,临床医师在围术期应给予足够重视。

Objective To establish a preoperative risk assessment scoring system to predict the risks of urosepsis after minimally tract percutaneous nephrolithotomy(MPCNL)under constant pressure perfusion.Methods A total of 422 patients undergoing MPCNL under constant pressure perfusion in our hospital during Dec.2018 and Dec.2020 were selected as the cohort,26 patients with postoperative urosepsis were selected as the case group,and 130 patients were enrolled in the control group.The predictive factors were determined with Logistic regression analysis,and a risk scoring system of urine sepsis after MPCNL under constant pressure perfusion was established.From the cohort,78 cases were randomly selected for internal verification.The calibration of the scoring system was verified with Hosmer-Lemeshow goodness of fit.The receiver operating characteristic(ROC)curve was drawn to verify the distinction degree of the scoring system and to carry out the risk grouping based on the maximum Youden index.Results The Logistic regression analysis showed 4 predictive factors,including woman(β=1.146,P=0.030),urine WBC≥2+(β=1.088,P=0.047),positive urine nitrite(β=2.653,P=0.000)and neutrophil to lymphocyte ratio≥2.5(β=1.459,P=0.008).Based on the above 4 factors,the risk scoring system was established(hereinafter referred to as“WaN-2 scoring system”),consisting of woman(1 point),urine WBC≥2+(1 point),urine nitrite positive(2 points),NLR≥2.5(1 point),with a total of 5 points.If the score was≥2 points,the patients had high-risk.The area under the ROC curve(AUC)of WaN-2 scoring system was 0.831[95%CI(0.740-0.921),P=0.000].Hosmer-Lemshow goodness of fit showed that there was no statistical difference between the predicted value of the model and the observed value(χ^(2)=1.124,P=0.771).Internal verification proved the AUC of the scoring system was 0.826[95%CI(0.712-0.941),P=0.000].Hosmer-Lemshow goodness of fit showed that there was no statistical difference between the predicted value of the model and the observed value(χ^(2)=0.884,P=0.643).The results indicated that the risk scoring system was of great distinction and calibration.Conclusion The WaN-2 scoring system established can preliminarily evaluate the risk of urosepsis after MPCNL under constant pressure perfusion.For high-risk patients with the WaN-2 score≥2 points,clinicians should pay enough attention before operation.