Reprodutibilidade e concordância interobservador do escore nefrométrico R.E.N.A.L.: ênfase nos achados de imagem

Abstract

Objective: To investigate the reproducibility and interobserver agreement for R.E.N.A.L. nephrometry scoring system. Materials and Methods: Two independent radiologists retrospectively analyzed 46 consecutive patients with renal masses, between 2008 and 2012, using the R.E.N.A.L. nephrometry score (RENAL-NS), which is based on the evaluation of five anatomical features of the tumor, as evaluated with computed tomography or magnetic resonance imaging: Radius, Exophytic/endophytic properties, Nearness to the collecting system, Anterior or posterior descriptor, and Location relative to the polar line. Tumor complexity was graded as low, intermediate, or high. The interobserver agreement was calculated for the total score and for the score for each parameter. Surgical excision of the tumors was used as the standard of reference. Results: The interobserver agreement for each of the RENAL-NS parameters, respectively, a hilar location, and the total score was 98%, 80%, 100%, 89%, 85%, 89%, and 93% of patients, corresponding to kappa values of 0.96, 0.65, 1.00, 0.75, 0.72, 0.78, and 0.88, respectively. The Nearness, Radius, and total score showed the best agreement. For the cases that were discordant in terms of the final score, no major implications in surgical planning were observed. Conclusion: The RENAL-NS is a structured, useful system to assess the anatomical features of renal tumors. It is easily applicable and reproducible, even for less experienced radiologists.