Abstract
Although the Liley spectrophotometric estimation of bile pigments in amniotic fluid [Amer. J. Obstet. Gynecol. 82, 1359 ( 1961)] is widely used, the levels of accuracy and precision necessary for clinically meaningful results have not been clarified. This paper delineates some of the important instrumental and procedural sources of error, and how each affects the final value of ΔA455: (a) A limit of error propagation is presented to show that the uncertainty in ΔA455 due to instrumental factors is considerably larger than is commonly appreciated. (b) It is desirable to use a logarithmic baseline estimation, as did Liley, instead of a linear estimation, which can introduce a serious bias. (c) Failure to establish a true zero-absorbance baseline before scanning the specimen can also result in a significant bias in the calculated ΔA455.