Predicting perioperative outcomes of robot assisted partial nephrectomy: The role of hospital case volume in the early phase of adoption

INTRODUCTION & OBJECTIVES: Information about the relationship between case volume and perioperative outcome after Roboticassisted Partial Nephrectomy [RAPN] are lacking. Our hypothesis is that perioperative outcomes are affected by hospital case volume in the early phase of RAPN adoption. MATERIAL & METHODS: Using the GQI-RUS project database, 687 patients treated with RAPN at 8 centers between 2006 and 2013 for a single cT1 small renal mass were evaluated. Hospital case volume was defined as the number of procedure performed after the first case at each center and computed as a continuous variable. Primary outcome was achievement of MIC [Margin, Ischemia and complications score], which represents a comprehensive assessment of optimal outcomes after RAPN, namely absence of positive surgical margins, warm ischemia time [WIT] ≤20 minutes and absence of major complications (Clavien ≥3). Secondary outcome was WIT. Multi-variable logistic (MIC) and linear (WIT) regression models were fitted to test the impact of hospital case volume on the outcomes after adjustment for patient (age, ASA score and eGFR) and cancer characteristics (clinical size, clinical stage, side and nephrometry risk class). RESULTS: Median patient age was 60 years (IQR 50-67). Median tumour size was 30 mm (IQR 22-40). Clinical stage was T1a and T1b in 77 and 23% of the population, respectively. Histology was clear cell, papillary, chromophobe and other type of carcinoma in 55, 16, 8 and 2% of the population, respectively. Median hospital case volume was 71 (IQR 28-153). MIC was achieved in 64% of the population. The positive surgical margins, WIT >20 minutes and major complications rate were 3, 33 and 3%, respectively. Median WIT was 17 minutes (IQR 14-22). After adjustment for patient and cancer characteristics, hospital case volume emerged as a significant predictor of MIC achievement (OR 1.06 p<0.001) and shorter WIT (Est. -7 seconds p=0.01). Specifically, a 1.06 fold increase in the probability of MIC achievement and a estimate of -7 seconds of WIT were recorded every 10 procedures performed. Conversely, tumour size and intermediate or high nephrometry class emerged as significant predictors of MIC failure and longer WIT (all p<0.05). Age, ASA score and eGFR were not significant predictors of MIC achievement or longer WIT (all p>0.05). CONCLUSIONS: Perioperative outcomes are significantly affected by hospital case volume. The completion of 10 procedures is associated with a 1.06 fold increase in the probability of MIC achievement and with a WIT reduction of -7 seconds. Conversely, tumour size and intermediate or high nephrometry class increase the risk of sub-optimal perioperative outcome or prolonged WIT.