Introduction: Hemorrhagic and infectious events represent the most relevant complications after percutaneous surgery for renal stones, leading to severe risks for patients and potential need of additional treatments and longer hospitalization. Existing nephrolithometric nomograms have been designed to predict SFR but their reliability in predicting complications is debated. Herein we present a newly designed nomogram with intention to predict hemorrhagic/infectious events after PCNLs
Methods: We conducted a multicentric prospective study involving 20 different high-volume Chinese Centers on adult patients undergoing standard (24 Fr) PCNL or mini (18 Fr) PCNL. Dataset was derived from previous RCT (clinicaltrials.gov, NCT02635048), where patients have been randomly assigned to either mini-PCNL or standard-PCNL to treat renal stones with maximum diameter up to 40 mm. Aim of the study was to identify pre-operative risk factors for early post-operative infectious/hemorrhagic complications. They included fever (defined as one or more episodes of temperature >38°C), septic shock (defined according to Q-sofa criteria), transfusion (in case of hemoglobin <70 g/dL) or angioembolization (in case of significant bleeding with progressive hemoglobin drop and hemodynamic instability)
Results: A total of 1999 patients fulfilled inclusion\exclusion criteria and have been enrolled in the study. 1001 patients (50.07%) received mini-PCNL and 998 received standard PCNL (49.93%). The overall SFR was 86% with a mean maximum stone diameter of 29 mm (SD 25.0-35.0). Mean operative time was 36 minutes (SD 27.0-51.0). 170 patients (8.5%) had fever post-operatively and 14 (0.7%) developed urosepsis. 24 patients (1.2%) required transfusion and 19 (0.9%) needed embolization. The overall risk of post-operative infectious/hemorrhagic complications was (11.35%).
After multivariable logistic model, the elements included in the nomogram were age (p=0.041), BMI (p=0.018), maximum stone diameter (p < 0.001), pre-operative hemoglobin (p=0.005), type 1 or 2 diabetes (p=0.05), eGFR <30 (p=0.0032), hypertension defined as blood pressure >135/85 mmHg (p=0.001), previous PCNL or pyelo/nephrolithotomy on same kidney (p=0.0018), presence of severe hydronephrosis (p=0.002). After internal validation, the AUC of the model was 0.73. The model can be used through the link: https://caterinagregorio.shinyapps.io/app_nomogramma/
Conclusions: This is the first nomogram intended to predict infections and bleedings after PCNLs, the model shows a good accuracy, and it can support clinicians in balancing the risks of post-operative adverse events and optimize the peri-operative patient’s workout and management. Further external validation is necessary to support this model
.jpg)
.jpg)