Standardization of immunochemical determinations of serum albumin.

When we compared four commercially available preparations of human serum albumin with the human serum standard IFCC 74/1 by radial immunodiffusion, by immunoprecipitation turbidimetry, by laser nephelometry, and by "rocket" immunoelectrophoresis, three of the preparations gave almost "theoretical" results with the immunoprecipitation turbidimetric method. Results by the other three methods tended inconsistently to be low. Four pools of normal human serum were also analyzed for albumin by these four immunochemical methods, again with IFCC 74/1 as the reference standard. The results were virtually identical with those obtained by fractionation with 1.8 mol/L sodium sulfate and determination of protein nitrogen in the filtrate. We suggest that a combination of (a) fractionation of a pool of normal human serum in this way and (b) critical comparison with selected commercial preparations of human serum albumin will permit standardization of the serum albumin determination.