Moderated Poster
Presentation Authors: Zine-Eddine KHENE*, Benoit Peyronnet, Romain Mathieu, Jean-Christophe BERNHARD, Rennes, France, Nicolas DOUMERC, Jean-Baptiste BEAUVAL, Toulouse, France, Grégory Verhoest, Rennes, France, Alexandre DE LA TAILLE, Créteil, France, Franck Bruyere, Tours, France, Morgan Roupret, Paris, France, Jay Raman, Hershey, PA, Karim Bensalah, Rennes, France
Introduction: Many studies reported risk factors associated with complications after robotic partial nephrectomy (RPN); however, there is no predictive model available. Our objective was to generate a nomogram based on preoperative parameters to predict the occurrence of a major complication within 30-days of RPN.
Methods: The study included 1342 patients with a clinically localized renal tumor who underwent RPN between 2010 and 2017 at seven academic centers. The primary outcome was the major complication rate. A multivariable logistic regression model was fitted to predict the risk of major complications after RPN. Model-derived coefficients were used to calculate the risk of major complications. Local regression smoothing technique was used to plot the observed rate against the predicted risk of major complications.
Results: In multivariate logistic regression, male gender (OR: 3.06; p = 0.001), Charlson comorbidity index (OR: 1.19; p = 0.005), ECOG PS (OR: 1.89; p = 0.02), low hospital volume (OR: 0.38; p = 0.03) and high RENAL score (OR: 3.67; p = 0.001) were significant predictors of major postoperative complications. A preoperative nomogram incorporating these risk factors was constructed with an area under curve of 75%.
Conclusions: Using standard preoperative variables from this multi-institutional RPN experience, we constructed and validated a nomogram to predict post-operative complications after RPN.
