Analytical goals for haematology tests.

All analytical methods can be defined in terms of their practicability and reliability performance characteristics. Desirable standards of performance, or analytical goals, are required for these, particularly for imprecision and inaccuracy. Goals for imprecision have been set using a variety of methods including reference values, opinions of clinicians, views of individuals, and data on biological variation. The last approach is currently favoured; desirable imprecision is equal to or less than one-half of the biological within-subject variation. If this goal is met, total variability of test results is increased by less than about 10% due to analytical variability. Valid estimates of within-subject variability are available for the complete blood count. The goal for inaccuracy is that methods should have no bias so that results are comparable over time and geography; goals based on biological variation should be viewed and used, therefore, as goals for total error. In current practice, some of the goals cannot be met; they should be considered as targets worthy of achievement, not as inflexible criteria of acceptance or rejection of methods.