A simulation-based methodology for uncertainty modeling and analysis of clinical laboratory measurement processes

Clinical diagnostics have a vital role in many phases of the medical decision making process, and therefore knowledge of the quality of the laboratory measurement result is necessary for making the right medical decision. A statement of uncertainty about the result of a laboratory test provides this information. The clinical laboratory measurement process is conceptualized as a self-contained system with the patient sample representing the input and the measurement result being the system output, and a framework for modeling a general clinical laboratory measurement process is presented. The paper discusses how performance specifications for individual components can be used to characterize the associated uncertainty, and Monte Carlo simulation is used to integrate these individual component uncertainties into a net system uncertainty. The proposed approach is illustrated by developing a mathematical model of the serum creatinine assay analysis procedure. The output of the simulation is compared to quality control data from clinical laboratories. The uses of the model to: a.) estimate the uncertainty of the system and set quality specifications using these; and b.) estimate the contribution of each source of uncertainty to the net system uncertainty are illustrated with examples.