Diagnostic yield of upper tract imaging performed for hematuria screening: Results from a national, privately-insured cohort

Abstract

ObjectivesTo increase the population-level understanding of upper tract imaging evaluation during risk-based hematuria assessment. Using a national, privately-insured cohort, we evaluated the rates of imaging and the diagnostic yield of each imaging modality. We hypothesized that the rate of upper tract disease would be low and that the benefit of computed tomography over renal ultrasound would be limited. MethodsWe identified patients with an incident diagnosis of microscopic or gross hematuria between 2010 and 2015 within the MarketScan database. Multivariable logistic regression was used to predict the adjusted likelihood of any upper tract finding, with a subset analysis limited to malignant outcomes. ResultsWe identified 466,710 patients; 150,880 (32.3%) received ultrasound, 219,449 (47.0%) received computed tomography, and 96,381 (20.7%) received other imaging. Ultrasound detected 1,469 findings among 150,880 patients (1%); computed tomography detected 3,764 findings among 219,449 patients (1.7%); other imaging detected 2,612 findings among 96,381 patients (2.7%) (P < 0.0001). When adjusted for comorbidities, increased age, male sex, increased Elixhauser index, gross hematuria, and CT or other imaging were associated with upper tract findings, in the overall and subset analysis (all p < 0.0001). ConclusionsThe diagnostic yield of imaging is higher than expected; however, the detection rate remained low. As supported by updated clinical practice guidelines, clinicians should pursue risk-based imaging strategies during hematuria evaluation, although these data may support increased utilization of ultrasound at the initial evaluation.