Systematic Measurement of Membranous Urethral Length Using a Cross-Reference System: A Study on Interrater Variability.

Abstract

Purpose: To evaluate the effectiveness of a cross-reference system (CRS) in improving the reproducibility of membranous urethral length (MUL) measurements by reducing the interrater variability.

Materials and methods: A total of 100 patients who underwent robot-assisted radical prostatectomy (RARP) and preoperative MRI were enrolled. MUL measurements were independently performed by two urologists with different levels of experience, both before and after CRS implementation. The radiological reference MUL served as the baseline for comparison. Interrater variability and agreement with the radiological reference were assessed.

Results: Mean MUL values measured by both raters and the radiological reference showed no significant difference after CRS implementation. However, interrater measurement error was substantially reduced after CRS implementation, and Bland-Altman analysis demonstrated markedly narrower limits of agreement. At 6 months, preoperative MUL showed no significant association with continence (area under the curve [AUC] and Brier score unaffected by CRS). By 12 months, MUL remained a significant predictor of continence (unadjusted odds ratio [OR]: 0.79, p=0.003; adjusted OR: 0.76, p=0.001), and CRS modestly improved discrimination (ΔAUC +0.02) and calibration (ΔBrier: -0.01).

Conclusions: A systematic CRS-based measurement approach improved MUL reproducibility by reducing interrater variability. These findings emphasize the need for standardized MUL measurements to enhance the anatomical accuracy and continence prediction after RARP.