Analysis Of Risk Factors For Obstructive Pyonephrosis In Upper Urinary Tract Caculi

Objective:To investigate retrospectively the risk factors for causing pyonephrosis in patients with upper urinary tract caculi,and to develop a regression predictive model for probability of pyonephrosis.Finally,some clinical evidences is provided for the time of preventing and treating pyonephrosis in patients with urinary tract caculi.Methods:175 cases patients with upper urinary tract caculi admitted to the Second Hospital of Tianjin Medical University from January 2018 to October 2019 were collected.Among them,24 cases were with pyonephrossis and included into the case group,and 151 cases were without pyonephrossis and selected as the control group.The observational data are compared in the group,including the sex,age,body mass index(BMI),side,diabetes mellitus,hypertension,past history of ipsilateral urological surgery,grade of hydronephrosis,stones position,stones number,stones size,staghorn caculi etc.The factors that were statistically significant between the two groups were selected by Single variable analysis,and factors selected above were further analyzed in logistic multivariable regression analysis.And then,we established a logistic regression predictive model for assessing probability of pyonephrosis pyonephrosis in patients with upper urinary tract caculi.Results:In single-factor analysis,age(x~2=21.713,P<0.01),diabetes mellitus(x~2=45.277,P<0.01),past history of ipsilateral urological surgery(x~2=22.518,P<0.01),grade of hydronephrosis(x~2=11.344,P<0.01),stones position(x~2=8.217,P=0.016),staghorn caculi(x~2=28.407,P<0.01),stones size(Wilcoxon W=12288.50,P<0.01)were statistically significant,and then in logistic multivariable regression analysis,combined diabetes mellitus(OR 24.845,95%CI 4.213-146.51),having past history of ipsilateral urological surgery(OR 6.579,95%CI 1.321-32.769),moderate or severe hydronephrosis(OR 7.155,95%CI 1.693-30.234),staghorn caculi(OR207.403,95%CI 2.485-17309.133)is statistically significant(P<0.05).Based on the results of logistic multivariable regression analysis,a risk predictive model for pyonephrossis patients with upper urinary tract caculi was established.The predictive model equation was logit(P)=3.213*diabetes mellitus+1.884*past history of ipsilateral urological surgery+1.968*grade of hydronephrosis+5.335*staghorn caculi-5.323,and the area under ROC curve(AUC)is 0.95,95%confidence interval(CI)(0.918-0.982),Sensitivity 95.8%,Speciality 83.4%.Conclusion:These observational data,incuding diabetes mellitus,past history of ipsilateral urological surgery,grade of hydronephrosis,staghorn caculi,were independent risk factors that affect whether pyonephrossis occurred in patients with upper urinary tract caculi.Age,stones position,stones sizes were also risk factors.The predictive model we established have a good predictive capability and plays a significant role in preventing pyonephrossis and early stage treatment.Therefore,positively controlling blood glucose level,avoiding or degreasing the damage to the renal pelvis or ureter asfar as possibally in urological surgery,early decompression of the renal pelvis and removing obstruction in the discovery of hydronephrosis of moderate and above,early treatment of staghorn caculi were very significant to prevent the occurrence of pyonephrossis in patients with upper urinary tract caculi.