[Retrospective comparison of the reproducibility and the performances of R.E.N.A.L. score, 12 segments score and...

Compare three morphometrics scores of the kidney on the reproductibility and the predictability of complications per- and post-partial nephrectomy.

This comparative study analyzed a monocentric retrospective cohort of 47 patients affected by a unique renal tumor and operated by partial nephrectomy by lumbotomy excluding the genetic diseases from January till December, 2014. The preoperative scanners were reviewed in double-blind by a radiologist and an internal surgeon who classified each tumor according to the three classifications: R.E.N.A.L. score, score in 12 segments and the University Paris Renal Cancer Classification (UPRCC). The judging criteria were the operative and ischemic duration, the blood loss, the preoperative injuries, the length of the hospital stay, the postoperative complications according to the Clavien-Dindo classification. The Cohen's kappa coefficient and the one-way analysis of variance were used respectively to compare the inter-observer reproducibility and the predictability of the complications for each classification.

Forty-four tumors of which 30 (64%) were on the left kidney, with an average size of 55.19mm with 17 (39%) on the upper pole, 12 (27%) on the lower pole and 15 (34%) medio-renal. The UPRCC was the most reproductible classification with 90% inter-observer concordance against 84% for the R.E.N.A.L. score and 72% for the score in 12 segments. The most predictable classification of complications was the score in 12 segments in particular thanks to the subgroup including the number of segments reached which allows to predict significantly the risk of a duration of intervention, prolonged ischemia and the risk of blood loss (P=0.03, 0.06 and <0.001).

The main result of this study is that, as in the majority of the previous studies on morphometrics scores, no score allows to predict the postoperative complications.

4.