Guideline References

Measurement Procedure Comparison and Bias Estimation Using Patient Samples

The primary standard governing method comparison study design and analysis. EP09C specifies minimum sample requirements (≥40 patient samples, preferably 100+), data collection practices, regression techniques (Deming, Passing-Bablok, OLS), difference plots, outlier detection, and bias estimation at clinical decision points. The third edition introduced confidence interval computation for all parameters and CLSI-compliant outlier detection using the Extreme Studentized Deviate (ESD) method.

CAP Checklist Requirement

CAP requires that laboratories document a comparison of patient results between the new and old (or reference) method prior to placing a new method into service. Comparison data must be retained as part of the laboratory's quality and validation records.

Standard: Establishment and Verification of Method Performance Specifications

CLIA requires that laboratories verify or establish method performance specifications — including accuracy relative to a comparison method — for all non-waived tests before reporting patient results.