Assesses whether a reference or QC material behaves like a patient sample, by checking if it falls inside the prediction interval of the patient-sample method-comparison regression.
Reference materials and controls are only trustworthy if they behave the way real patient samples do across two methods — a property called commutability. A non-commutable material can show method differences that patient samples never would (a matrix effect), invalidating its use for calibration or QC harmonization.
The engine fits the patient-sample regression (method X vs method Y) and its 95% prediction interval, then checks each reference/QC material's (X, Y) point. A material inside the interval is commutable; one outside behaves differently from patient samples.
Patient-sample measurements by method X and method Y (paired), plus the reference material's measurements by method X and Y.
The patient-sample regression + residual SD, and for each material its predicted value, the 95% prediction interval, and a commutable / not-commutable verdict.
A non-commutable material must not be used to transfer calibration or to judge agreement between the two methods — the observed difference is an artifact of the material, not a real method bias. Assess commutability before drawing method-comparison conclusions from processed materials.