Mp35-17 Comparison of a Novel Risk Calculator With the Aua 2020 Guidelines for Stratifying Urothelial Malignancy Risk in Patients With Microscopic Hematuria

You have accessJournal of UrologyBladder Cancer: Epidemiology & Evaluation II (MP35)1 May 2024MP35-17 COMPARISON OF A NOVEL RISK CALCULATOR WITH THE AUA 2020 GUIDELINES FOR STRATIFYING UROTHELIAL MALIGNANCY RISK IN PATIENTS WITH MICROSCOPIC HEMATURIA Emeka Udedibia, Kerli Orellana, Kenneth Solosky, Richard Amdur, and Simon Hall Emeka UdedibiaEmeka Udedibia , Kerli OrellanaKerli Orellana , Kenneth SoloskyKenneth Solosky , Richard AmdurRichard Amdur , and Simon HallSimon Hall View All Author Informationhttps://doi.org/10.1097/01.JU.0001009372.61513.54.17AboutPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookLinked InTwitterEmail Abstract INTRODUCTION AND OBJECTIVE: To enhance the ability of the American Urological Association (AUA) 2020 guidelines for stratifying risk of urothelial neoplasms we developed two novel risk models. Accurate risk prediction is critical for clinical decisions, and the new tools, a simple model and a regression model, may offer enhanced stratification. METHODS: A retrospective analysis of 2552 patients with microscopic hematuria (MH) from a single center was conducted to develop two risk models. After preliminary statistical analysis, we identified four key predictors: age, microscopic red blood cell count (logRBC), history of genitourinary malignancy (hxGUmal), and tobacco use. The first model, derived using multivariable logistic regression, incorporates the four predictors. The second, a simple model, assigns points to each of these predictors to quantify risk. For the regression model, we calculated mean predicted probabilities for the presence of urothelial neoplasm and established a probability threshold to categorize risk into low, medium, or high. We compared both models' performance with AUA guidelines using standard metrics. RESULTS: With the regression model, mean predicted probabilities of urothelial neoplasm for patients without and with tumors were 0.03 and 0.21, respectively (p<.0001), a significant enhancement compared to the AUA's risk categorization. Of our cohort, the simple model classified more patients (64%) as low risk compared to the AUA (4%) and identified a high-risk group with a 26.8% incidence of urothelial neoplasm, surpassing the AUA's 3% incidence (Figure 1). The regression model classified 49% as low risk and showed a 13.7% incidence in the high-risk category. Notably, the regression model achieved a sensitivity of 0.91 and a negative predictive value of 0.997, both higher than those of the AUA guidelines. CONCLUSIONS: Both the simple model and the regression model demonstrated superior predictive accuracy for stratifying patients into urothelial neoplasm risk categories when compared to the AUA 2020 guidelines. These models showed an enhanced ability to classify low risk patients potentially leading to a reduction in unnecessary evaluations (Figure 1). Further validation studies are recommended. Download PPT Source of Funding: None © 2024 by American Urological Association Education and Research, Inc.FiguresReferencesRelatedDetails Volume 211Issue 5SMay 2024Page: e588 Advertisement Copyright & Permissions© 2024 by American Urological Association Education and Research, Inc.Metrics Author Information Emeka Udedibia More articles by this author Kerli Orellana More articles by this author Kenneth Solosky More articles by this author Richard Amdur More articles by this author Simon Hall More articles by this author Expand All Advertisement PDF downloadLoading ...