摘要
目的 探讨输尿管上段结石碎石术后出血的影响因素及预测模型建立.方法 选取2021年2月至2022年2月本院收治的310例输尿管上段结石患者作为研究对象,所有患者均接受微创经皮肾镜碎石术治疗,根据患者术后出血情况分为出血组(33例)和未出血组(277例).采用多因素logistic回归模型筛选出影响患者术后出血的独立危险因素,基于独立危险因素创建列线图预测模型,同时对该列线图的预测性和准确度进行验证.结果 与未出血组比较,出血组患者的手术时间≥60 min、体质量指数(BMI)＞24 kg/m2、糖尿病病史和存在结石残留占比显著更高,差异均有统计学意义(均P＜0.05).多因素logistic回归分析结果显示,手术时间≥60min、BMI＞24kg/m2、糖尿病病史和存在结石残留均是术后出血的独立危险因素(均P＜0.05).列线图结果显示,手术时间≥60 min的总分为98分、BMI＞24 kg/m2为61分、糖尿病病史为45分、存在结石残留为75分.列线图模型的校准曲线预测值均与实际值吻合.Hosmer-Lemeshow拟合优度检验显示模型拟合较好(x2=8.199,P=0.150).受试者工作特征(ROC)曲线的曲线下面积为0.816.结论 手术时间、BMI、糖尿病病史和结石残留均是输尿管上段结石碎石术后出血的危险影响因素,据此构建的列线图预测模型发现有较高的预测效能.
Abstract
Objective To study the risk factors of bleeding after lithotripsy for upper ureteral calculi,and to establish a prediction model.Methods Three hundred and ten patients with upper ure-teral calculi underwent lithotripsy in our hospital from February 2021 to February 2022 were enrolled,and classified into two groups according to the postoperative bleeding status,bleeding group(33 ca-ses)and non-bleeding group(277 cases).Multivariate logistics regression model was used to screen out the independent risk factors of postoperative bleeding.The nomogram prediction model was estab-lished based on independent risk factors,and the discriminative ability and predictive accuracy of the nomogram was verified.Results Compared with the non bleeding group,the operation time ≥60 min,body mass index(BMI)＞24 kg/m2,history of diabetes and the proportion of residual stones in the bleeding group were significantly higher,and the differences were statistically significant(all P＜0.05).Multivariate logistic analysis showed that operation time ≥60 min,BMI＞24 kg/m2,history of diabetes mellitus and presence of residual stones were independent risk factors for postoperative bleed-ing(all P＜0.05).The nomogram showed that the total score of operation time ≥ 60 min was 98 points,BMI＞24 kg/m2 was 61 points,diabetes history was 45 points,and the presence of residual stones was 75 points.The predicted values of the calibration curves of the nomogram model are consist-ent with the actual values.Hosmer lemeshow goodness of fit test showed that the model fitted well(x2=8.199,P=0.150).The area under the receiver operating characteristic curve(ROC)was 0.816.Con-clusions The operation time,BMI,history of diabetes,and presence of residual stones are all risk factors for postoperative bleeding after lithotripsy for upper ureteral calculi,and the nomogram predic-tion model is found to have high predictive performance.
基金项目
广西壮族自治区卫生健康委员会项目(Z20210883)
维普
万方数据