Begin with the analytical target profile
Define what is measured, the matrix and sample types, reportable result and units, expected range, precision needs, decision context, and operating conditions. These choices determine which performance characteristics and study designs are relevant.
A potency procedure that compares a test article with a reference standard has different risks from a screening EC50 method or a quantitative interpolation assay. A copied validation checklist cannot resolve those differences.
Document the evidence for each performance characteristic
For every selected characteristic, document the question, experimental design, factors and levels, statistical method, predefined acceptance criterion, data set, result, and conclusion. Preserve protocol deviations and exclusions rather than rebuilding them at report time.
| Characteristic | Example planning question |
|---|---|
| Accuracy | How will agreement to an accepted value or orthogonal evidence be assessed? |
| Precision | Which within-run and intermediate sources of variation matter? |
| Specificity | Can relevant components or conditions interfere with the response? |
| Range | Where are accuracy, precision, and model behavior acceptable? |
| Robustness | Which deliberate method variations challenge routine performance? |
Software can support validation, but it cannot replace it
Versioned calculations, automated test results, controlled settings, and reproducible reports can support an organization’s validation work. The organization still defines intended use, procedures, roles, risk controls, acceptance, deployment, and ongoing change management.
A product label such as “validated” or “compliant” does not replace an evaluation of the system as it is configured and used in your laboratory.
Frequently asked questions
Does Q2(R2) prescribe one set of bioassay acceptance limits?
No. It provides a general framework. Characteristics, experimental designs, and criteria should be appropriate to the procedure’s intended purpose and scientifically justified.
Can software make an assay method Part 11 compliant?
Compliance depends on the records, controls, procedures, people, and implemented system. Software features can support controls but do not create compliance by themselves.