Test report forgery (revision 27)

Test report forgery is the fabrication of an analytical report, or the alteration, misattribution or reuse of a genuine one. It is the reason a document cannot by itself establish anything about a material: a document is an assertion, and assertions can be manufactured as easily as the things they describe.^[1]

Four distinct practices are usually grouped under the term. A wholly fabricated report describes work never done. An altered report changes figures on a genuine document. A misattributed report is a genuine report on other material presented as covering this one. A reused report is a genuine report on an old lot presented as current.^[2]

The last two are the most common and the hardest to detect, because the document is genuine and will survive any inspection of the document itself. What distinguishes them is the relationship between the document and the material, which is a question about traceability rather than about the paper.

Only the last two are detectable from the document plus the vial, and only if the vial carries a batch number to check against. This is why the presence of a lot identifier on both is more informative than any figure on the certificate.^[2]

Fabrication and alteration are detectable in principle by asking the named laboratory whether it issued the report. Whether that is possible depends on the laboratory's practice; some issue reports with a verifiable reference and some do not. A report that cannot be checked with its issuer is not thereby forged — most are not — but it is unconfirmable.^[3]

Some fabrications are visibly defective, and the defects are usually arithmetic or procedural rather than typographic.

• Dates out of order: testing before manufacture, release before testing, retest before release.
• A stated method incapable of producing a reported determination — a purity figure attributed to a method that does not separate, or a water figure attributed to a method that does not measure water.
• Determinations reported to more significant figures than the method supports.
• Related-substance figures that do not sum consistently with the reported main-peak area.
• A chromatogram whose visible integration does not correspond to the stated figure.

None of these establishes forgery. Each establishes that the document is internally defective, which is a weaker and more defensible statement — and, in practice, a sufficient reason to treat the figures as unusable.^[1]

The absence of such signals establishes nothing at all. A competent fabrication is internally consistent by construction.

This wiki records what documents assert and attributes those assertions to the documents. It does not hold samples, does not verify documents with issuing laboratories, and does not treat a certificate as evidence of the property it reports.^[1]

The practical position that follows is not that documents are worthless. A supplier that issues lot-specific certificates naming its methods, retrievable years later against a lot identifier, has built something that is expensive to fake consistently over time — and consistency over time is a stronger signal than any single document. That is a statement about a records system rather than about any one certificate.^[4]

The reverse also holds: a single unverifiable document proves little in either direction, and treating its absence as evidence of wrongdoing is as unsound as treating its presence as evidence of quality. See Purity claim inflation for the adjacent failure in which the document is genuine but the figure is presented misleadingly.

• Certificate of analysis
• Purity claim inflation
• Batch number
• Lot traceability
• VendorInvestigate
• Counterfeit peptide