Preoperative urine nitrite versus urine culture for predicting postoperative fever following flexible ureteroscopic lithotripsy: a propensity score matching analysis.

PURPOSE: Multiple studies have reported that preoperative positive urine culture is an independent risk factor for postoperative fever (POF) after ureteroscopy (URS). Urine nitrite is associated urinary tract infections (UTIs). However, none of studies has explored the role of urine nitrite in the prediction of POF after flexible URS (fURS).
METHODS: Patients who underwent fURS by the same surgeon between 2009 and 2019 were screened. Sensitivity and specificity of urine culture and urine nitrite were calculated. Propensity score (PS) matching was performed to get a baseline-balanced retrospective cohort to avoid potential bias. Receiver operating characteristic-area under the curve (ROC-AUC) calculated was used to determine the predictive power of models. Decision curve analysis (DCA) was plotted to obtain the clinical benefit of the models.
RESULTS: Poseoperative fever (POF) is defined as the temperature of the patient higher than 38 ℃ within 72 h after operation, with no sign of infection in other systems. 31(2.8%) of 1095 cases had POF after fURL. Urine nitrite had a better specificity than urine culture for POF diagnosis (P < 0.001). After the PS matching, a well-balanced cohort of 24 POF group and 96 no-POF group was produced. The mean AUC from the bootstrap resampling method for urine nitrite model (AUC: 0.8736; 95% CI: 0.8731-0.8743) was significantly increased than that of the urine culture model (AUC: 0.8385; 95% CI: 0.8378-0.8392). The application of two kinds of POF predicting models could bring clinical net benefit when the probability is < 35%. However, urine nitrite model showed a better clinical net benefit acquirement compared to the urine culture model.
CONCLUSION: Preoperative positive urine nitrite may play a pivotal role in the prediction of POF after fURS and needs to be validated by future evidence.