Appendix I. Methods for Measuring Serum Albumin

Most laboratories utilize a colorimetric method for the measurement of the serum albumin concentration and particularly the bromcresol green (BCG) assay. If another assay is utilized, the normal range specific to that assay should be used. Research that reports the serum albumin should specify the assay used and its normal range.

Nephelometry and the electrophoretic method223 are very specific for the determination of the serum albumin concentration. However, these methods are time-consuming, expensive, and not generally used in clinical laboratories. The BCG colorimetric method is rapid, reproducible, and has been automated.224 This method uses small aliquots of plasma, has a low coefficient of variation (5.9%), and is not affected by lipemia, salicylates, or bilirubin. With values in the normal electrophoretic range of 3.5 to 5.0 g/dL, the BCG method gives values that are comparable to the values obtained by electrophoresis. The normal range for the serum albumin by the BCG method is 3.8 to 5.1 g/dL.224 The BCG method differs from the electrophoretic method by about 0.3 g/dL.223 The BCG method underestimates albumin in the high normal range and overestimates albumin below the normal range with an overall mean overestimation of approximately 0.61 g/dL.225

Some laboratories use the bromcresol purple (BCP) colorimetric method to measure the serum albumin concentration.223 Although this method is more specific for albumin and has specificity similar to electrophoretic methods, clinically it has proved to be less reliable than the BCG method. BCP has been shown to underestimate serum albumin in pediatric HD patients with a mean difference of 0.71 g/dL.226 Maguire and Price227 have demonstrated similar results in CRF patients.

Serum albumin concentrations obtained by the BCG method in HD patients were virtually identical to the values obtained using nephelometry. Values obtained by the BCP assay underestimated the nephelometric values by 19%. Agreement between BCG and BCP with the nephelometric values in CAPD patients showed less variation; however, the BCG values were not different from the nephelometric values.228

Chronic dialysis units often have little influence over the method used by their reference laboratories. If the BCG method is available, it should be requested. If the BCP method must be used, then the normal range for that laboratory should serve as the reference. Additionally, less clinical weight might be given to serum albumin concentrations measured by the BCP method and other markers of malnutrition in ESRD patients might be more heavily weighted.