Accuracy Evaluation of a Tetrabromophenolphthalein Ethyl Ester Colorimetric Assay for Urinary Albumin

Abstract

Background The tetrabromophenolphthalein ethyl ester (TBPE) assay has been used to quantify urinary albumin in point-of-care devices. We assessed the accuracy of this TBPE assay for urinary albumin through comparison with an established immunoturbidimetric method (ADVIA 1800 Chemistry System, Siemens). Methods We developed a TBPE assay protocol to quantify albumin in the range associated with microalbuminuria (0–200 mg/L). The Jaffe reaction and a 3-dimensional (3D) surface were used to compensate for creatinine interference. Spiked simulated urine samples and patient samples were used to compare the TBPE assay with the immunoturbidimetric method. Multiple linear regression was used to analyze factors that could account for discrepancies between the 2 methods. Results We found that creatinine interfered with the TBPE assay. To compensate, a 3D surface was successfully used to quantify albumin in spiked deionized water and simulated urine samples. In spiked simulated urine samples, the immunoturbidimetric method underestimated the albumin concentration by 2 to 45 mg/L, and the TBPE assay overestimated it by 9 to 82 mg/L. In patient samples, the albumin concentrations measured with the TBPE assay and the immunoturbidimetric method differed by an average of 184 mg/L. Conclusions The TBPE assay is a function of the creatinine concentration, and a 3D surface can be used to provide accurate albumin concentrations for standard samples. The corrected TBPE method and the immunoturbidimetric method deviated from known concentrations of spiked samples. Further investigation and comparisons with a third albumin measurement method, such as LC-MS/MS, are necessary before conclusions on the accuracy of the TBPE assay can be made.

A Variable Height Microfluidic Device for Multiplexed Immunoassay Analysis of Traumatic Brain Injury Biomarkers

“…The limit of detection (LOD) and coefficient of variation (CV) for each QLISA were calculated as described in Krausz et al [ 25 ]. Briefly, the LOD for each assay was calculated as the value of the blank plus three standard deviations (SD) ( n = 10).…”