Kinetic and chemometric studies of the determination of creatinine using the Jaffé reaction. Part 2. Application to human serum samples: kinetic behaviour and chemometric evaluation of the determination

Abstract

The kinetic behaviour of the reaction of alkaline picrate with creatinine in human serum samples was found to be similar to that for standard creatinine solutions containing albumin. A chemometric evaluation of the kinetic determination of creatinine using the Jaffé reaction was carried out. The analysis of variance (ANOVA) method applied to the ΔA_45,180 values, obtained from two replicates of three different serum samples over a period of 10 d, showed that the between-day and between-replicate variations added a component to the total variability, the residual error (σ_R²) being 5 × 10^–5. A study of the accuracy of the determination was carried out by means of percentage recovery experiments, Youden's method and the standard additions method. Percentage recovery experiments showed that albumin has a marked effect on the results obtained. The application of Youden's method to four serum samples indicated that the method does not have a constant bias error, but, by applying the standard additions method it was concluded that the method has a proportional bias error. The recovery factor, defined as the ratio of the slope of the standard additions graph to that of the standard response graph, was also calculated for the four serum samples. The best values were obtained with different standard response graphs (7.7, 15.4 and 25.6 g l^–1 of albumin) for each sample. A modification of the routine procedure used in clinical laboratories is proposed. This modification is based on the principles of the standard additions method and gives better results for creatinine content than those obtained with the routine procedure.