Biological variation in protein C, protein S and antithrombin concentrations in plasma of healthy subjects.

The analytical, intra- and inter-individual components of biological variation were estimated for protein C, protein S and antithrombin over a period of 6 weeks in a cohort of 17 apparently healthy subjects. Expressed as percentage activity (protein C and antithrombin) and as percentage concentration in normal human plasma (protein S), the mean values for men and women show no significant differences (p > 0.05) for any of the analytes studied. Calculated analytical goals (CV, %) for precision required for optimal patient care are: protein C, 2.9; protein S, 2.9 and antithrombin 2.7. A single numerical index, called 'index of fiduciality', was also calculated to demonstrate that the analytical performance of our method was satisfactory. The generally applicable differences (% activity or % concentration) required for two results to be significantly different (p ≤ 0.05) were calculated as: protein C, 17; protein S, 16 and antithrombin, 16. The usefulness of critical differences as guidelines for the interpretation of changes in serial results was investigated using an 'index of heterogeneity' of intra-individual variation. The marked degree of individuality demonstrated for all the quantities indicates that, if conventional population-based ranges are used uncritically, major changes in analyte concentration may not be correctly identified for some patients, because observed values continue to lie within the reference range. The utility of conventional population-based reference intervals was determined by calculating a single numerical index, called 'index of individuality'. For protein C, protein S and antithrombin we found that five of a patient's specimens are required to achieve a 95% chance that the mean is within ± 5% of the true value.