Abstract
The automation of the prothrombin time using Low Turbidity Australian Reference Thromboplastin on the Cobas Fara II centrifugal analyser is described. Initially this test was programmed as a 2-step technique to mimic the reference manual method. However, the wide variation of Cobas Fara normal control values obtained with several different batches of the reagent was unacceptable. Investigation of 1-step techniques resulted in dramatic changes in the normal control values. These changes would consequently alter reference values and affect International Sensitivity Index (ISI) and International Normalized Ratio (INR) determinations. This paper stresses the necessity to fully assess all methodology variables when evaluating new reagents.