Luke Webster a , Brett Larive b , Jennifer Gassman b , Alexander Bullen a , Steven D. Weisbord c , d , Paul M. Palevsky c , d , Linda F. Fried c , d , Kalani Raphael e , f , Tamara Isakova g , Joachim H. Ix a , h , i , *
09 March 2020
Background: Accurate assessment of urine flow remains challenging in both inpatient and outpatient settings. We hypothesized we could derive an equation that would accurately estimate urine flow rate (eV) through derivation from other existing equations commonly used in nephrology clinical practice. Methods: The eV equation was derived using the Cockcroft-Gault and the measured creatinine clearance (CrCl = U<sub>Cr</sub>V/P<sub>Cr</sub>) equations. Within the African American Study of Kidney Disease and Hypertension (AASK; n = 570) and COMBINE ( n = 133) clinical trials, we identified participants with concordant estimated and measured creatinine excretion rates to define a subset with highly accurate 24-h urine collections, to assure a reliable gold standard. We then compared eV to measured 24-h urine flow rates in these trials. Results: In AASK, we found a high correlation between eV and measured urine flow rate (V; r = 0.91, p < 0.001); however, Bland-Altman plots showed that eV was 9.5 mL/h lower than V, on average. Thus, we added a correction factor to the eV equation and externally validated the new equation in COMBINE. eV and V were again highly correlated ( r = 0.91, p < 0.001), and bias was improved (mean difference 5.3 mL/h). Overall, 80% of individuals had eV that was within 20% of V. Conclusions: A simple equation using urine creatinine, demographics, and body weight can accurately predict urine flow rate and may have clinical utility in situations where it is difficult to accurately measure the urine flow rate.