Linearity and method comparison are not just graphs

What CLIA is asking the laboratory to establish

For an unmodified FDA-cleared or FDA-approved test system, the laboratory must verify performance before reporting patient results. The required work includes accuracy, precision, reportable range, and verification that the manufacturer's reference intervals are appropriate for the laboratory's patient population.

That is an operating responsibility, not a software checkbox. The laboratory needs a written plan, suitable materials, traceable results, defined acceptance criteria, documented review, and authorized approval before implementation.

• Accuracy: do results agree closely enough with an expected or comparative value?
• Precision: can the method reproduce results under the study conditions?
• Reportable range: can the laboratory support results across the interval it intends to report?
• Reference interval: is the manufacturer's interval appropriate for the local patient population?

Linearity and comparison answer different questions

A linearity study examines whether measured results remain proportional to expected values across a stated concentration interval. A method-comparison study examines the relationship and bias between a candidate method and a comparative method using appropriate samples.

The same scatterplot can appear in both studies, but the study design and interpretation are not interchangeable. Linearity is not established by R-squared alone. Method comparison should assess bias, the regression model, clinically important decision points, sample distribution, outliers, and uncertainty.

• For linearity, show the expected-versus-observed relationship, predefined allowable deviation, study coverage, replicate design, exceptions, and the supported interval.
• For comparison, show candidate-versus-comparator results, an appropriate regression approach, a difference or bias view, and bias at medically important concentrations.
• Use the current manufacturer instructions and the laboratory's approved SOP. Do not let a generic template silently choose the final acceptance rule.

Why current laboratory situations become difficult

A new analyzer may arrive before vendor applications support is available. A reagent formulation, calibrator lot, software version, or reportable-range claim may change. The laboratory may have the instrument, samples, and deadline but no single owner for the protocol, calculations, exception review, and final evidence package.

That gap consumes supervisor and director time. It also creates avoidable risk: poorly distributed samples, unverified limits, incomplete traceability, a graph with no decision rule, or a conclusion that overstates what the data support.

The executive question is not 'Did we make a graph?'

The better question is: can qualified leadership see what was tested, why it was acceptable, what remains uncertain, who approved it, and what monitoring is required after go-live? A defensible package protects the patient, the laboratory director, the implementation timeline, and the organization's investment.

AssayIQ can structure the study plan, analyze approved deidentified data, produce review-ready tables and visuals, document limitations, track exceptions, and prepare the final package for qualified client review. This returns technical leaders to the work that only they can authorize and oversee.