Analysis Of Risk Factors And Establishment Of Prediction Model For Urosepsis After Percutaneous Nephrolithotomy

Objective To clarify the potential risk factors of urosepsis(US)after percutaneous nephrolithotomy(PCNL)by meta-analysis and clinical studies,then establish a novel prediction model of PCNL-US which may provide evidences in future clinical practice.Methods(1)By searching Pub Med,The Cochrane Library,and Embase as foreign language databases,the research literature on PCNL-US related risk factors published from January 2010 to June 2022 was collected.Using ’percutaneous nephrolithotomy’;’sepsis’;’urosepsis’;’risk factors’;’kidney stones’ as the subject headings and keywords,literature retrieval and meta-analysis were carried out,and relevant forest maps and funnel plot evaluation results were drawn potency.(2)This retrospective clinical study included 646 inpatients with kidney stones treated by percutaneous nephrolithotomy from January 2019 to August 2022 in the Second Clinical Urology Department of Lanzhou University.Using R Programming language,the patients were divided into modeling group(452 cases)and verification group(194cases)according to the ratio of 7:3.After digitizing the basic information and clinical data of the patients in the modeling group and performing single factor analysis,there were significant differences in the results.(P < 0.05)indexes were screened by LASSO regression and multivariate Logistic regression analysis was performed.(3)The factors with statistical differences in the multi-factor results in the modeling group were used to build a Logistic regression prediction model.The R statistical software was used to draw the risk nomogram model and visualize it,and at the same time draw the ROC curve,calibration curve,and decision curve of the verification group for verification.Results1.The meta-analysis finally screened 16 high-quality case-control studies,with a total of 5986 patient data.After analysis,it was clear that a total of 8 independent risk factors were identified,including urine culture positive,urine nitrite positive,long operation time,preoperative white blood cell elevation,infection stones,staghorn stones,heavy burden stones,and gender.2.In this study,the clinical data of 646 patients were collected in the Urology Medical Center of the Second Hospital of Lanzhou University.The incidence of US in the modeling group was 7.96%(36/452),and the incidence of US in the verification group was 7.73%(15/194).In the modeling group,the single factor results showed that: female(P<0.001),diabetes(P<0.001),urine culture(P<0.001),urine nitrite(P< 0.001),preoperative fever(P < 0.001),staghorn stones(P < 0.001),multiple(P=0.003),mild renal hydrops(P < 0.001),Albumin-to-Globulin Ratio(AGR)(P <0.001),Lymphocyte-to-Monocyte Ratio(LMR)(P<0.001),Neutrophil-to-Lymphocyte Ratio(NLR)(P=0.027),urine leukocytes(P < 0.001),blood leukocytes(P < 0.001),serum albumin(P<0.001),hemoglobin(P<0.001),surface area(P<0.001),serum creatinine(P=0.049),hospitalization Time(P<0.001),residual calculi(P<0.001)are related to PCNL-US.However,age,Body Mass Index(BMI),operation history,hypertension,bilirubin,blood urea nitrogen,degree of hydronephrosis,Platelet-to-Lymphocyte Ratio(PLR),CT value,blood loss,and operation time were not significantly correlated.3.After screening by LASSO regression,multivariate logistic regression analysis indicated that positive urinary nitrite(OR=8.497,95CI%: 3.666-20.554,P<0.001),preoperative fever(OR=6.832,95CI%: 2.784-16.719,P<0.001),white blood cells(OR=1.231,95CI%: 1.068-1.429,P < 0.001),residual stones(OR=4.661,95CI%:1.990-11.128,P<0.001)were independent risk factors for PCNL-US.Based on the above research results,we constructed a nomogram prediction model of PCNL-US risk factors.The area under the ROC curve(AUC value)of the modeling group was0.889,with a sensitivity of 73.2% and a specificity of 90.7%;the area under the ROC curve of the verification group was 0.829,with a sensitivity of 86.7% and a specificity of 65.9%.The P values in the Hosmer-Lemeshow test were all greater than 0.05,and the calibration curves and clinical decisions performed well.ConclusionThis study suggests that positive urinary nitrite,preoperative fever,white blood cells,and residual stones are independent risk factors for PCNL-US.A clinical prediction model based on the above four risk factors may provide a reference for avoiding the occurrence of PCNL-US.