Edinburgh Research Explorer

Accuracy of ultrasound vs computed tomography scan for upper urinary tract malignancies and development of a risk-based diagnostic algorithm for haematuria in a UK tertiary centre

Research output: Contribution to journalArticle

Original languageEnglish
JournalInternational urology and nephrology
Publication statusPublished - 25 Aug 2020


Purpose There is no consensus across guidelines on a diagnostic algorithm for upper urinary tract (UUT) evaluation following presentation with haematuria. Our aim is to compare the diagnostic accuracy of ultrasound (USS) compared to CT-scan for UUT malignancies and also determine the considerations important for a risk-based diagnostic protocol for haematuria. Methods We reviewed our ‘haematuria clinic’ database to identify patients who had both USS and CT-scan for UUT evaluation between September 2015 and August 2017, and calculated the diagnostic accuracy of these imaging modalities for histologically confirmed UUT cancers. Furthermore, we identified risk factors in our diagnostic algorithm for haematuria and conducted regression analysis to determine their ability to predict UUT malignancies. Results Overall, 575 patient records were studied. Age range was 21–92 years, M:F was 1.4:1, majority (81.2%) had visible haematuria, and 12 (2.1%) UUT cancers were diagnosed [renal cell carcinoma—1.4%; upper tract urothelial cancer—0.7%]. USS and CT-scan had diagnostic accuracy for UUT cancers of 95.8 and 99.1%, respectively (p < 0.001). Haematuria type was a significant consideration only on univariate analysis, while multivariate binary logistic regression showed that male gender, smoking, occupational exposure, and positive urologic history were the main risk factors associated with UUT malignancies. Conclusion USS and CT-scan have comparably high diagnostic accuracy for detecting UUT malignancies. USS may, therefore, be considered as the first-line UUT imaging modality when utilized in a risk-based diagnostic algorithm. Larger, multicentred studies are needed to validate our findings and influence guideline development.

ID: 163129041