sigma metric

A sigma metric expresses analytical performance relative to a selected allowable error goal. A common laboratory calculation is sigma = (TEa − absolute bias) / CV, when total allowable error, bias, and coefficient of variation are all expressed on the same percentage basis. With absolute concentration units, the denominator must instead be a compatible standard deviation. Mixing units makes the calculation meaningless.

The result depends strongly on the chosen performance goal, the concentration examined, and the quality of the bias and imprecision estimates. It is not an intrinsic score that remains constant across all uses of a method. A negative numerator indicates that estimated bias already exceeds the selected allowance. Use the metric alongside clinical risk and control-system characteristics rather than as a stand-alone quality verdict.

Sigma calculations can support a reasoned discussion of control-rule selection and frequency during accreditation. Keep the source of the error goal and the periods used for bias and imprecision visible. Avoid presenting a favorable metric as proof of competence or a guarantee of patient safety; the broader examination process still requires independent controls.