Source of Certificate of analysis

{{Infobox concept | name = Certificate of analysis | subtitle = Quality document | image = coa-sheet.svg | caption = A certificate reports what was measured, on what, by what method, and by whom. Fields beyond those four are optional and frequently absent. <!-- Identification --> | Abbreviation = CoA, COA | Issued by = The supplier, manufacturer or repackager | Scope = One named lot <!-- Typical content --> | Core fields = Product, lot, test, specification, result, method, release date | Frequently absent = Water content, peptide content, counterion, endotoxin | Standing = Self-published primary document }} {{hatnote|This article is about the document. For results obtained by a laboratory unconnected with the supplier, see [[Third-party testing]]; for the sampling problem that limits both, see [[Blind sampling]].}} A '''certificate of analysis''' is a document issued with a consignment of material, reporting the results of specified tests carried out on a specified [[Batch number|lot]]. In regulated pharmaceutical supply it is the record by which a quality unit releases material against a written specification, and it is retained as part of the [[Lot traceability|batch record]].{{r|usp1503}} Outside that framework — which is where all material labelled [[Research use only|for research use only]] sits — no obligation to issue one exists, no format is prescribed, and no authority audits what is written on it. The document's evidential standing follows from who issues it. '''A certificate is a claim by its issuer about material it is selling.''' That is not a criticism: a claim made in writing, against a named lot, by an identifiable party, is a substantially stronger thing than no claim at all, and it is checkable in ways a verbal assurance is not. It is simply not the same as an independent measurement, and the two are frequently conflated.{{r|iso17025}} This article describes what the document normally contains, which fields carry information and which are decorative, how a certificate relates to an independently obtained result, and what a purchaser can reasonably ask for beyond it. == Fields and what each establishes == A certificate that is useful for anything carries at least seven fields. Each of them answers a question that cannot be answered from any of the others. | Field | Answers | Comment | |---|---|---| | Product identity | What the material is claimed to be | Should give a name a reader can look up, not a catalogue code alone | | Lot identifier | Which material this document describes | Must match the vial label; see [[Lot traceability]] | | Test performed | What was measured | "Purity" alone is not a test; a method makes it one | | Method | How it was measured | Without this, two results cannot be compared | | Specification | What the result was measured against | A result with no specification is a number, not a release decision | | Result | What was found | Should carry its units and its significant figures honestly | | Release date and signatory | When and by whom the lot was released | A signature records release, not verification | A specification is a list of tests, the analytical procedures referenced by them, and the acceptance criteria applied to the results, so a figure reported with no specification against it has not been the subject of a release decision at all.{{r|ich}} The distinction in the last row is the one most often misread. A signed certificate establishes that somebody at the issuing organisation accepted responsibility for releasing that lot. It does not establish that the tests were performed by an accredited laboratory, that the signatory saw the raw data, or that the material in the purchaser's hand came from the lot named.{{r|iso17025}} == The fields most often missing == Across the certificates described in reports reaching this wiki, four determinations are absent far more often than they are present, and their absence has the same consequence in each case: the purity figure cannot be converted into a statement about mass.{{r|usp1503}} * '''Water content.''' A lyophilised peptide routinely holds several per cent water. Determined by [[Karl Fischer titration]] or by [[Loss on drying|loss on drying]], which are not interchangeable. * '''[[Peptide content|Peptide content]].''' The mass fraction of the vial that is the peptide itself, as opposed to water, counterion and residual solvent. * '''Counterion identity and content.''' Trifluoroacetate is the commonest and can account for a large fraction of the mass of a basic peptide. * '''[[Bacterial endotoxin test|Endotoxin]].''' Reported for a small minority of material; unlike purity, it speaks to the process rather than to the shelf life, because endotoxin does not decrease on storage. {{note|A certificate reporting 99% purity and nothing else is consistent with a vial containing anywhere from roughly two thirds to essentially all of its nominal peptide mass. The purity figure is a ratio among the species that eluted; it is not a fraction of the vial. See [[Area percent purity]].}} == Relationship to an independent result == Where a purchaser obtains an independent determination on material from a lot and compares it with the certificate, agreement between the two establishes something narrow and worth being precise about: two measurements of the same sample agreed. It rules out a certificate written without measurement for that sample. It does not establish that the sample represents the lot, because the sample was chosen by the purchaser rather than drawn under a documented procedure — see [[Blind sampling]].{{r|iso17025}} Disagreement is equally easy to over-read. Two laboratories running different gradients on different column chemistries, integrating valley-to-valley rather than by drop-line, and calibrating against different [[Reference standard|reference standards]], will produce different numbers for the same vial and both may be honest. This is why the '''method''' field matters more than its position on the page suggests: without it, a comparison between a certificate and a report is a comparison between two unspecified procedures.{{r|usp1503}} === What to request beyond the certificate === A purchaser wanting a certificate figure to be comparable with an independently obtained one needs four further items, all of which exist in the issuer's records whether or not they appear on the document: # the gradient and column chemistry the determination was run on; # the detection wavelength, since ultraviolet response is not uniform across impurities; # the water determination, with the method that produced it; # the [[Peptide content|peptide content]], or the counterion identity and content from which it can be derived. Requests of this kind are made against the lot identifier printed on the vial. The supplier articles on this wiki each record, under §Ordering and availability, which of the four that organisation supplies as standard and which it supplies on request.{{r|reports}} == Forgery, alteration and the limits of the document == Because a certificate is an unregulated document with no prescribed format, it can be altered or fabricated, and instances of both are documented — see [[Test report forgery]] and [[Purity claim inflation]]. Three observations follow, none of which requires assuming bad faith anywhere. First, a certificate that cannot be tied to a physical vial is worth very little: the check a purchaser can perform without any equipment at all is that the lot identifier on the vial, the certificate and the invoice are the same string. Second, a certificate issued at manufacture describes the material at manufacture. Where material is held in a regional warehouse and dispatched months later, the interval between the certificate date and the dispatch date is information the certificate does not carry, and it has to be asked for separately. See [[Beyond-use date]] and [[Temperature excursion]]. Third, a scannable code on a certificate establishes that the document was issued from the system it points at. It says nothing about the content of the document or about the material.{{r|iso17025}} == References == {{reflist}} <ref name="usp1503">United States Pharmacopeia, General Chapter <1503>, "Quality Attributes of Synthetic Peptide Drug Substances" (informational). USP–NF, current revision. Describes the determinations that together characterise a synthetic peptide, and why purity alone does not.</ref> <ref name="iso17025">ISO/IEC 17025:2017, ''General requirements for the competence of testing and calibration laboratories''. International Organization for Standardization. The standard against which a testing laboratory's competence is assessed; self-description is not assessment against it.</ref> <ref name="reports">Independent test reports and purchaser submissions held by PeptidePedia, 2024–2026. Individual reports on individual samples; they are not a sample of any organisation's output. See [[Blind sampling]].</ref> <ref name="ich">International Council for Harmonisation, ''Q6A: Specifications — Test Procedures and Acceptance Criteria for New Drug Substances and New Drug Products: Chemical Substances''. Defines a specification as a list of tests, references to analytical procedures and acceptance criteria; a result reported without one is not a release decision.</ref> == Further reading == * [[Project:Sourcing guidelines]] — how this wiki cites certificates, test reports and community tallies. * [[Project:Verifiability]] — why a photograph of a certificate is a primary document of uncertain provenance. == See also == * [[Third-party testing]] * [[Area percent purity]] * [[Peptide content]] * [[Blind sampling]] * [[Lot traceability]] * [[Batch number]] * [[Test report forgery]] {{DEFAULTSORT:Certificate of analysis}} [[Category:Certificates of analysis]] [[Category:Testing methodology]] [[Category:Traceability]] [[Category:Quality and testing]]