Because of the inaccuracy of bilirubin (BIL) determinations, we have investigated some sources of error associated with the most commonly used methods. Inadequate standardization appears to be the most common error. BIL standards in either human serum albumin (HSA), bovine serum albumin (BSA), or pooled human serum were analyzed by these methods. Molar absorptivities (ε) of azobilirubin by the Jendrassik—Gróf procedure were practically identical in all three protein bases. The Meites—Hogg and Malloy—Evelyn methods gave substantially higher ε values with BIL in serum than with BIL in either HSA or BSA. The precision of all methods was good, but best with the Jendrassik—Gróf procedure. BIL standards can be prepared with good reproducibility (CV <0.5%). Standards deteriorate appreciably at -23 °C but are stable at -70 °C. Our data indicate a need for improved commercial BIL controls.

This candidate Reference Method for measuring total bilirubin in serum is based on the Jendrassik-Gróf principle (Clin Chem 29: 297-301, 1983). Standard Reference Material no. 916 bilirubin (National Bureau of Standards) is used as the standard. Bilirubin standard solutions may be prepared either in human serum or in 40 g/L albumin solution (human or bovine), because we found the molar absorptivity of the azopigment at 598 nm to be identical in these media. The absorptivities of the unconjugated and conjugated azopigments appear to be identical, but the conjugated azopigment is completely hydrolyzed in the final reaction mixture. Bilirubin added to serum from adults or neonates was quantitatively accounted for. Interference by hemoglobin (up to 2 g/L), ascorbic acid (up to 20 mg/L), or zinc (at physiological concentrations) is negligible. Of the therapeutic drugs we tested, only L-dopa and alpha-methyldopa interfere. We established normal adult reference values for total bilirubin and examined the intraindividual variation in 19 subjects.

We developed an enzymatic method for measuring direct-reacting bilirubin (DBIL) in serum. At pH 4.5, bilirubin oxidase (BOX) oxidizes mono-conjugated bilirubin, di-conjugated bilirubin, and most of the delta-bilirubin to biliverdin. The resulting decrease in absorbance at 460 nm is linearly related to the concentration of DBIL in serum. Mean DBIL values in the 51 patients' sera examined by the BOX method and a diazo procedure (Clin Chem 1982;28:2305) were 45.4 and 42.8 mg/L, respectively. For the same samples, mean values for DBIL and conjugated bilirubin by the Kodak "Ektachem" methods were 50.2 and 24.8 mg/L, respectively. Hemoglobin, up to 1.5 g/L, does not interfere. Unconjugated bilirubin reacts negligibly. Day-to-day CVs were 2.2% and 2.4% at DBIL concentrations of 37 and 74 mg/L, respectively.

No abstract

We used an enzymatic method for measuring total bilirubin in serum. Results by this method varied linearly with bilirubin concentrations to at least 300 mg/L. The day-to-day precision (CV) of the method ranged from less than 1% to about 11% at bilirubin concentrations of 183 and 12 mg/L, respectively. Commonly used anticoagulants, serum preparation materials, and selected drugs had no effect on the apparent bilirubin concentration, but turbidity caused a slight increase and hemoglobin concentrations of 2 g/L resulted in lower values, by as much as 17 mg/L at a bilirubin concentration of 95 mg/L. Patients' results obtained with this enzymatic method were slightly lower than those obtained with methods based on the Jendrassik-Grof principle. The largest differences, seen in samples with high "direct" bilirubin concentrations, can be decreased by measuring the absorbance at 425 nm instead of at 465 nm as recommended by the supplier of the bilirubin oxidase method.