RESUMEN
PURPOSE: This study was done to evaluate by direct comparison the image quality of magnetic resonance urography (MRU) and computed tomography urography (CTU) and to assess the diagnostic confidence of the two techniques in detecting urothelial malignancy in patients with haematuria MATERIALS AND METHODS: Thirty-five patients with haematuria underwent both CTU and MRU. Two different investigators evaluated calyceal, renal pelvis, ureteral and bladder visibility. Their diagnostic confidence in detecting urothelial malignancy with the two procedures was assessed. A Wilcoxon matched-pairs test was performed to compare results. Inter-reader agreement was calculated by weighted kappa (WK) statistic. Patient history (further examinations, cystoscopy and histological specimens) was considered as reference standard to calculate receiver operating characteristic (ROC) curves of diagnostic confidence. RESULTS: CTU provided better visibility of urothelial structures (p<0.01) and allowed for greater diagnostic confidence (ROC area 0.994 vs. 0.938) than MRU, with a good inter-reader agreement (WK=0.62). Nevertheless, in obstructive patients with impaired excretory function, MRU, thanks to the static-fluid technique, offered better visualisation than CTU. CONCLUSIONS: There is a potential role for MRU in urinary tract imaging, but as diagnostic confidence in detecting urothelial malignancy is poorer than in CTU, it might be stareserved for patients at low risk for malignancy and for evaluating obstructed patients.