Analysis Of Risk Factors Of Postoperative Infection Of RIRS And Establishment Of Nomogram Prediction Model

BackgroundUrinary calculus is a common disease in the clinical work of urology.The lifetime prevalence of Urinary calculus is 5%to 12%,and the incidence rate varies by age,sex,region and race,and climate characteristics.The overall prevalence of kidney stones in China is 5.8%.The incidence of stones in children and the elderly is low,while the incidence of adults in working age is high,which also brings more direct and indirect economic losses.The causes of Urinary calculus and the mechanism of stone formation are more complicated.In general,most people have one or more supersaturated substances in their urine,such as calcium,uric acid,oxalic acid,etc.The crystals are easily precipitated to form stones,which are precipitates of mineral crystals.The current commonly used stone treatment methods include:drug row stone（MET）,extracorporeal shock wave lithotripsy（MET）,ureteroscopy lithotripsy（URL）,ureteroscopy lithotripsy lithotripsy（RIRS）,percutaneous nephrolithotomy Stone taking（PCNL）,etc.Due to the high success rate of the first treatment of RIRS and less surgical trauma,the number of patients receiving RIRS treatment has increased in recent years.The 2016 American Urological Association（AUA）guidelines and the 2020 European Urinary Association（EAU）guidelines recommend that for stones of<2 cm in the middle,upper,and renal pelvis,both ESWL and RIRS are first-line options;and for the next stone,if stone<1cm is preferred ESWL or RIRS,if the stone size is 1-2cm,you can choose RIRS or PCNL;ureteral stones>1cm are recommended RIRS as the first-line choice.Although RIRS has fewer complications,9 to 25%of patients still have postoperative complications.Most of these complications are not serious.Postoperative infection is the most common complication.It occurs in 7.7%of patients.It can be manifested as postoperative fever,Urosepsis or systemic inflammatory response（SIRS）.In severe cases,it needs to be transferred to ICU monitoring treatment,which may even threaten the patient’s life.The current study believes that there are urinary tract infections（positive urine leukocytes,urine nitrite,urine culture,etc.）,poor body condition（low albumin,many basic diseases,high ASA score,etc.）,previous surgery history,preoperative DJ ureteral stenting,insufficient anti-infection before operation,high intra-pelvic pressure during operation（without ureteral sheath or low mirror sheath ratio,high perfusion volume,fast perfusion rate,etc.）,large stone load,long operation time,infectious stones is a risk factor for infection after RIRS.However,the conclusions of different studies are not the same,and there is still a lack of studies with a large number of cases and a comprehensive analysis of these factors.Therefore,it is of great significance to comprehensively analyze the risk factors of postoperative infection of RIRS and make Nomogram prediction model to provide reference for clinical decision making for the prevention and treatment of postoperative infection of RIRS.Objective1.Analyze the risk factors of infection after RIRS;2.Establish Nomogram prediction model according to the relevant risk factors.Patients and MethodsThis study included a total of 1678 patients from September 1,2015 to December 31,2018.The inclusion conditions were as follows:1.Admission due to urinary calculi;2.RIRS were performed for the purpose of stone removal during hospitalization;3.The patient did not undergo other operations in this hospitalization.The exclusion conditions are as follows:1.Patients are hospitalized in other departments or transferred to our department for surgery;2.Patients with incomplete research data.Patients were divided into two groups according to whether their body temperature exceeded 38.5℃within 7 days after surgery:postoperative fever group and postoperative feverless group.The patient’s preoperative information includes the patient’s medical history,physical examination,laboratory examination,imaging examination,anti-infective treatment,intraoperative conditions and postoperative body temperature and other information.Mean±SD is used for data that conforms to the normal distribution,and Student’s t test（equal variance）or Brown-forsythe test（uneven variance）is used for the differences between groups;the median（P25,P75）is used for the data that does not conform to the normal distribution.It indicates that the difference between groups is tested by Mann-Whitney U test;all qualitative data is expressed by percentage,of which count data is tested by Pearsonχ² test or Fisher’s exact test,and rank data is tested by Mann-Whitney U test.Incorporate statistical variables,previous literature reports or clinical variables that are considered meaningful in the Logistic regression equation for univariate and multivariate analysis（stepwise method）.P value was selected by two-sided test.When P<0.05,the difference was considered statistically significant.According to multi-factor Logistic regression,the Nomogram prediction model is established and internal verification is completed.ResultsThe study included 1678 patients,including 1002 males（59.71%）and 676females（40.29%）,with an average age of 50.28±13.18 years.The maximum diameter of the stone was 20（13.5,30）mm,and the stone area was111.5（63,193.75）mm².There were 339（20.20%）cases with positive urine culture.There were 119 patients in the fever group and 1559 patients in the no-fever group,with a fever rate of 7.09%.According to the logistic univariate regression analysis,the differences between the postoperative fever group and the postoperative no-fever group were statistically significant（P<0.05）,with Female patients（57.98%vs 38.94%）,BMI（23.35±3.39vs 24.08±3.54）kg/m²,no smoking history（2.52%vs 8.34%）,blood pressure at admission（127.27±19.74 vs 130.62±16.87,76.07±13.28 vs 78.47±11.66）mm Hg,with a history of febrile urinary tract infection（31.93%vs 7.44%）,positive physical examination at admission(38.66%vs History of urologic surgery（76.47%vs62.35%）,calculus area（130（78,251）vs 110（62,191））,infectious calculus（31.25%vs14.09%）,blood leukocyte>10.0×10⁹/L（15.13%vs 6.41%）,urine RBC<4.0×10¹²/L（11.76%vs 5.00%）,hemoglobin<110g/L（21.85%vs）11.10%),platelet<100×10⁹/L（3.36%vs 0.77%）,platelet>400×10⁹/L（11.76%vs 2.89%）,potassium<3.5mmol/L（10.92%vs 6.62%）,calcium<2.03mmol/L（4.20%vs 1.16%）,carbon dioxide<21.0mmol/L（20.17%vs 11.31%）,direct bilirubin>6umol/L（3.36%vs0.45%）,albumin<35g/L（23.53%vs 6.22%）,uric acid<220umol/L（5.04%vs0.91%）,urine p H≥6.5（48.73%vs 35.99%）,positive urinary occult blood（96.64%vs91.65%）,positive nitrite（35.29%vs 11.38%）,pyuria（47.06%vs 16.90%）,positive urine culture（47.90%vs 47.90%）,the completion of preoperative anti-infection time≥9 days（10.92%vs 1.15%）,replacement of≥3 antibiotics（8.40%vs 0.90%）,the ratio of endoscopic sheath<0.7（39.50%vs 29.70%）,and the annual operation quantity of the surgeons<20（19.33%vs 9.49%）.According to the logistic multivariate regression analysis,with a history of febrile urinary tract infection（OR 2.636,95%CI:1.554-4.4,P<0.001）,positive physical examination at admission（OR 1.580,95%CI:1.018-2.428,P=0.039）,blood uric acid<220umol/L（OR 4.271,95%CI:1.167-13.73,P=0.020）,and blood leukocyte>10.0×10⁹/L（OR 2.307,95%CI:1.185-4.277,P=0.010）,albumin<35g/L（OR 1.952,95%CI:1.069-3.455,P=0.025）,pyuria（OR 2.39,95%CI:1.457-3.908,P=0.001）,positive urine culture（OR 1.984,95%CI:1.177-3.308,P=0.009）and the completion of preoperative anti-infection time≥9 days（OR 3.667,95%CI:1.42-9.179,P=0.006）were independent risk factors for postoperative fever.The annual number of operations≥20（OR 0.542,95%CI:0.311-0.982,P=0.006）was the protective factor.According to ROC curve,AUC=0.824,and fitting curve,c-index=0.812.This indicates that the Nomogram prediction model has good reliability and differentiation.The DCA curve shows that the use of this model below the 40%threshold can benefit clinical decision making.Conclusion1.The risk factors for fever after RIRS operation were the presence of cold and fever before surgery,blood leukocyte>10.0×10⁹/L,albumin<35g/L,pyuria,serum uric acid<220umol/L,preoperative anti-infection time≥9 days,and the annual operating quantity of the operator<20 sets.2.The Nomogram prediction model in this study has a good reliability and differentiation in predicting the risk of postoperative fever in patients undergoing RIRS surgery,which can benefit clinical decision making.