Legion Diagnostics

Quantitative Method Comparison

The Quantitative Method Comparison module estimates the systematic error (bias) between two measurement procedures by testing the same patient samples on both methods. One method is designated the reference method (X) — the established or gold-standard procedure — and the other is the test method (Y) — the candidate or new procedure being evaluated. By analyzing how results from the two methods relate across the clinically relevant measurement range, the module quantifies whether the test method's bias is acceptable relative to the analyte's Total Allowable Error (TEa).

Method comparison studies use two complementary analytical approaches: regression analysis (to characterize the linear relationship between methods and estimate bias at any concentration) and Bland-Altman difference plots (to visualize systematic and random differences between methods across the measurement range). Together, these provide a complete picture of method agreement and support defensible pass/fail determinations at clinically important decision levels.

When to Use This Module

A quantitative method comparison study is required or strongly recommended in the following situations:

  • New instrument or analyzer implementation — Before a new analyzer is placed into routine use, bias relative to the current method must be assessed to confirm that patient results will not be meaningfully affected by the transition.
  • Method or reagent system change — When switching to a new reagent system, platform, or manufacturer, a method comparison study documents whether the new method produces equivalent results within the allocated TEa.
  • Method verification — End-user laboratories verify that a manufacturer-validated method performs as claimed in their specific environment, including confirming that bias against a reference method is within published specifications.
  • Accreditation requirements — CAP, The Joint Commission, and CLIA require documented method comparison data as part of the method validation/verification record for any test placed into routine use.