A study of six representative methods of plasma bilirubin analysis
The large number of modern methods which have been published for the determination of total bile pigments in the blood plasma indicates the difficulties of the estimation and suggests the need for a reliable standard method. In view of the present importance of plasma bilirubin levels in the diagnosis and treatment of jaundiced subjects, six analytical procedures have been studied, each chosen as far as possible, as a representative of a group of methods each embodying different principles.
A spectrophotometric method (White, Haidar, and Reinhold, 1958) and five diazo coupling procedures (Jendrassik and Gróf, 1938; Powell, 1944; Perryman, Richards, and Holbrook, 1957; Stoner and Weisberg, 1957; Lathe and Ruthven, 1958) were investigated, and the results obtained by carrying out duplicate analyses by different methods on 54 stored and 64 fresh plasmas or sera are presented and discussed in this paper.
An amount of haem pigment which can be detected visually in plasma is sufficient to cause a considerable underestimation of bilirubin by the method of Powell (1944) and overestimation by the method of Stoner and Weisberg (1957). Inferences drawn from serial estimations by these methods can therefore be misleading, and this is especially pertinent when results obtained from infant blood containing variable amounts of haem pigments are being assessed. Methods based on the original van den Bergh procedure with various preliminary steps of protein precipitation in alcoholic solutions at about pH4 give low and poorly reproducible results; losses due to the coprecipitation of bilirubin esters with protein have been confirmed. In our opinion such methods should be abandoned.
Determination of total bilirubin by colour (White et al., 1958) is reliable only for plasma known to contain no lipochrome or `directly-reacting' pigment; total bile pigment may be overestimated by this technique in plasma containing conjugated bilirubin.
We consider that the method of Lathe and Ruthven (1958) is to be preferred for use on a routine basis. It is conveniently simple, and under all conditions normally encountered gives results sufficiently accurate and reproducible for clinical use. When this or any other procedure is used, a primary standardization between different laboratories with sera containing known amounts of added bilirubin is essential if comparable results are to be obtained.