Kinetic study of the Jaffé reaction for quantifying creatinine in serum: 2. Evaluation of buffered reagent and comparison of different data-processing options.

Here we describe the evaluation of several data-processing options for the kinetic determination of creatinine by use of the Jaffé reaction. Data-processing options evaluated include initial-rate, two-point fixed-time, rate at t = k-1, and multipoint curve-fitting predictive methods. We evaluated these options for a buffered formulation of the Jaffé reagent and studied the effects of potential interferents, including glucose, acetoacetate, bilirubin, and albumin, on each option. To reduce effects of bilirubin, we evaluated the inclusion of a preoxidation step with ferricyanide. All the data-processing options gave good precision and linearity between the measurement objective and creatinine concentration. However, differences between slopes of calibration plots in aqueous and serum matrices ranged from a high of +60% for the two-point, fixed-time method to a low of -11% for the curve-fitting, predictive method. Standard additions of creatinine to sera were quantified reliably (yielding 96% to 102% of target values) by the predictive method and less reliably (62% to 102%) by the other methods. We conclude that the predictive method has the potential to yield the most reliable results for creatinine.