Combines an assay's bias and imprecision into a single total analytical error and compares it against an allowable total-error goal, with a sigma metric.
Bias (systematic offset) and imprecision (random scatter) both push a result away from the truth. Total analytical error (the Westgard model) rolls them into one worst-case figure — TAE = |bias| + z·SD — that a result can plausibly deviate from the true value. Comparing TAE against an allowable total error (TEa) tells you whether the assay is fit for its clinical purpose.
The sigma metric, (TEa − |bias|) / SD, expresses the same idea on the Six-Sigma scale: how many standard deviations of headroom sit between the process and the specification. Sigma ≥ 6 is world-class; < 3 needs tighter QC or method improvement.
A test-method column and a paired reference/target column (bias and imprecision are taken from the differences). Optionally an allowable total error.
Bias, imprecision (SD), the total error TAE = |bias| + 1.65·SD, the sigma metric, and an acceptable / exceeded verdict against the allowable total error.
A method can have excellent precision yet fail on total error because of bias, or vice versa — TAE is what the patient ultimately experiences. The sigma metric maps directly to how much QC is needed: high sigma tolerates simple QC, low sigma demands multi-rule QC and more frequent checks.