Comparing Results from Different Analytical Laboratories
Align the question before comparing numbers
Begin by identifying the measurand in each report. A chromatographic area percentage, a target mass fraction and a dry-basis assay are different quantities even if each appears under a purity heading. Numerical closeness is not meaningful until the definitions and units are compatible.
Also establish that both tests refer to the same material and relevant sampling history. A shared product name does not establish a shared batch or subsample. Record identifiers, analysis dates and any intervening handling that could affect the comparison before attributing a difference to laboratory performance.
Compare the full procedure
Chromatographic conditions, detector settings, reference assignments and processing rules can change what a result means. Different selectivity may reveal an impurity in one method while leaving it unresolved in another. Neither headline percentage can explain that difference without the supporting analytical details.
ICH Q14: Analytical Procedure Development concerns scientific approaches to analytical procedure development. ICH Q2(R2): Validation of Analytical Procedures addresses validation. In a laboratory transfer or comparison, these frameworks support a practical question: what evidence demonstrates that the receiving procedure produces suitable results for the intended purpose under its own working conditions?
Design a useful comparison
Where a material difference matters, agree the comparison plan before collecting more data. Define sample identity, the quantities to report, relevant method information and the basis for judging agreement. This prevents a later discussion from shifting its criteria merely because the first pair of numbers looked inconvenient.
Replicate measurements can reveal variability, but repeated injections alone do not describe every source of uncertainty. Sampling, preparation, calibration and interpretation may also contribute. Consider which stages the comparison actually challenges and avoid treating excellent repeatability at one stage as proof that the entire measurement process is equivalent.
Investigate discrepancies without averaging them away
Retrieve relevant reports through the COA library and preserve both original results. Ask each laboratory to explain its reporting basis and method scope. An arithmetic average of unlike quantities does not resolve their disagreement and can conceal the very issue the comparison was intended to investigate.
If the results remain inconsistent after aligning definitions, identify a targeted next step. That may involve a shared reference, a controlled sample comparison or an alternative method addressing a suspected interference. Record the eventual explanation and its limits so future batch decisions can use the learning rather than repeating the same unresolved numerical argument.
Related research reading
- Assay Versus Purity: Specifying Peptide Content
- Orthogonal Testing: Using Independent Evidence for Identity
Explore the research blog and original batch-document library for further context.