Diagnostic utility of axial imaging in the evaluation of hematuria: A systematic review and critical appraisal of the literature

Abstract

Increasing severity of hematuria is instinctively associated with higher likelihood of urological malignancy. However, the robustness of the evidentiary base for this assertion is unclear, particularly as it relates to the likelihood of upper urinary tract pathology. Thus, the value of axial imaging in the diagnostic workup of hematuria is unclear due to differences in the underlying patient populations, raising concern for sampling bias. We performed a systematic review to characterize the literature and association between severity of hematuria and likelihood of upper urinary tract cancer based on axial imaging. MEDLINE, EMBASE, and Cochrane were systematically searched for all studies reporting on adult patients presenting with hematuria. We used Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) for reporting of this systematic review and meta-analysis and the Newcastle-Ottawa Scale for risk of bias assessment. Degree of hematuria was classified as "microscopic," "gross," or "unspecified." Three urological malignancies (bladder, upper tract urothelial, and renal cancer) were considered both individually and in aggregate. Random-effects model with pairwise comparisons was employed to arrive at the axial imaging diagnostic yields. Twenty-nine studies were included, of which six (20.7%) reported on patients with gross hematuria only, four (13.8%) reported on patients with microscopic hematuria only, seven (24.1%) included both, and 12 (41.4%) did not define or specify the severity of hematuria. Of 29 studies, two (6.9%) were at high-risk of bias, 21 (72.4%) at intermediate-risk, and six (20.7%) at low-risk of bias using the Newcastle-Ottawa criteria. Based on axial imaging, rates of diagnoses of renal, upper tract urothelial, and bladder cancers differed with differing severity of hematuria. Notably, rates of renal and upper tract urothelial carcinoma were higher in studies of patients with unspecified hematuria severity (3.6% and 10.4%, respectively) than among patients with gross hematuria (1.5% and 1.3%, respectively). When all urological malignancies were pooled, patients with unspecified hematuria were diagnosed more frequently (19.5%) compared to those with gross (15.3%) and microscopic hematuria (4.5%, difference=1.51%, 99% confidence interval 3.6-26.5%). Lack of granularity in the available literature, particularly with regards to patients with unspecified hematuria severity, limits the generalizability of these results and highlights the need for future studies that provide sufficient baseline information, allowing for firmer conclusions to be drawn.