Response-surface co-optimization of reaction conditions in clinical chemical methods.

Abstract

Abstract We describe the theory and applications of the response-surface approach to simultaneous optimization (co-optimization of multiple interdependent variables. The co-optimization experiments are designed according to a selected response-surface model. Computer-assisted analysis includes fitting the data to the model, testing the resulting fit for statistical validity, and plotting contour maps of the model for simple interpretation. Co-optimizations of the reaction parameters of three methods--creatine kinase, lipase, and aspartate aminotransfrease--are discussed to illustrate the application of the approach. In contrast to the commonly used optimization strategies in which each factor is varied in turn while the others are kept constant, the response-surface approach allows study of several responses (reaction rates, sensitivities) and effects (linear, curvature, interaction) at the same time. It also allows determination of accurate optima, which is necessary for the formulation of analytically reliable clinical methods.