Abstract Urine albumin (UA) and total protein (UTP) are commonly measured to assess kidney diseases. Current clinical practice guidelines recommend fixed threshold values for making medical decisions based on UA or UTP. However, neither biomarker has standardized results from in-vitro diagnostic medical devices (IVD-MDs) from different manufacturers used by medical laboratories. In a 2013 study, 37% to 45% differences in mean values for UA measurements of 332 freshly collected patient’s samples were observed among 16 IVD-MDs. The differences in UA results among the IVD-MDs had the potential to cause erroneous classifications of chronic kidney disease patients and errors in medical decisions. Current IVD-MDs for UA use immunoassays that are highly selective for measuring the protein. However, calibration schemes for UA IVD-MDs do not have fit-for-purpose higher-order certified reference materials (CRMs) or reference measurement procedures (RMPs) available. A standardization program organized by the U.S. National Institute of Diabetes and Digestive and Kidney Diseases and the International Federation of Clinical Chemistry and Laboratory Medicine has worked with collaborators to develop higher-order CRMs and RMPs suitable for use in standardizing calibration traceability of IVD-MDs for measuring UA. Transition to using the new reference system by IVD-MD manufacturers is expected to achieve standardized UA results in the near future. Data from external quality assurance programs for UTP show differences between different IVD-MDs on the order of 50%. IVD-MDs for UTP use several different chemical reactions that have different measurement responses for the different proteins found in urine which themselves vary with the type of disease. Consequently, there is not a coherent approach to develop a CRM or RMP that is suitable for use as a higher-order reference for standardized calibration of IVD-MDs for UTP. Due to these technical limitations, there is no currently planned activity to improve agreement of results for UTP measurements.
Miller et al. (Thu,) studied this question.