eTopical Precious Finds
A certificate of analysis arrives as a clean PDF: laboratory name, sample identifier, method, graph and a reported result. The file opens normally, the signature looks consistent, and every page can be printed. Yet six months later, when a researcher wants to check how a peak was measured, the information needed to reconstruct that result may be sitting in software controlled by an outside testing contractor. The PDF can be authentic while the record behind it remains inaccessible to the person relying on the conclusion.
This is an illustration of a documentation risk, not an allegation about any particular testing laboratory or supplier. It raises a specific question for research organizations: what does an analytical report preserve, and what must remain retrievable beyond the report itself?
The PDF may be genuine without being the complete record
A laboratory report is a communication product. It explains what work was requested, identifies the sample as recorded by the testing organization, and presents selected findings. Good reports make clear what was measured and which method was used. They may also display a chromatogram, a mass spectrum or a summary table. Those are valuable elements, but a formatted view does not necessarily contain the electronic records from which it was generated.
Chromatography provides a concrete example. An instrument acquires a signal over time, and the laboratory’s software uses information about the run and processing settings to turn that signal into a reported output. The resulting graph may be printed to a PDF. A reader can see the drawn peaks, but may not be able to inspect the acquisition sequence, review how the baseline was assigned, or determine whether the data were later reprocessed.
In its public guidance on current good manufacturing practice records, the US Food and Drug Administration explains that a printed chromatogram does not generally contain the complete electronic raw-data record. FDA specifically identifies supporting information such as the injection sequence, instrument method, integration method and audit trail. That statement addresses regulated drug-manufacturing and testing records under applicable requirements; it should not be presented as a universal legal obligation for every educational research test.
For nonregulated research, the same distinction is still useful. A report may faithfully reproduce a result without carrying every detail needed to evaluate how that result came into existence. Knowing how independent laboratory testing works helps separate the independence of the testing organization from the completeness of the evidence delivered to its customer.
What an instrument file can explain that a chart cannot
Consider a hypothetical sample analyzed using a chromatographic method. The final report shows a prominent signal and several smaller features. A reviewer asks how one shoulder was distinguished from the main peak. The plot itself may be too compressed to answer. The relevant information could include the original detector response, the processing parameters used to define peak boundaries, and any documented changes made during review.
Different software packages organize these details differently. Some store signal files separately from methods and result tables. Others maintain interconnected project databases. A PDF export might preserve the chosen view while excluding the underlying project structure. Even a high-resolution graph is not automatically an editable, reconstructable analytical record.
The distinction matters because a calculated result depends on more than the appearance of the plotted line. The software’s processing settings, the reference or comparison material, and the laboratory’s decisions about which signals to include may all affect the number ultimately reported. None of this means the result is wrong. It means that a serious challenge to a result cannot always be resolved by zooming into a static image.
For this reason, a research organization should know whether an outside laboratory retains original instrument files, the conditions and duration of retention, and the arrangements for retrieving them if questions arise. The precise record package should be appropriate to the method and the intended use rather than copied mechanically from another scientific field.
An audit trail is about the history of the record, not the appearance of the graph
The FDA’s December 2018 guidance, Data Integrity and Compliance With Drug CGMP: Questions and Answers, describes metadata as context necessary to understand data. Examples include when a record was acquired, which user or instrument generated it, and the related audit trail. An audit trail can document events associated with the creation, modification or deletion of an electronic record, allowing authorized reviewers to reconstruct what happened.
These records are not automatically evidence of misconduct. Revisiting an integration decision or correcting a documented error can be part of legitimate laboratory work. The issue is whether the process leaves a reliable account of the original information, the later action, who performed it and why. A history that is available for review makes ordinary scientific corrections distinguishable from changes whose basis cannot be established.
There is a real regulatory example of the stakes. In an April 24, 2020 warning letter to International Trading Pharm Lab, the FDA described computer systems used for chromatography that allowed raw files to be deleted or timestamps changed, while audit trails were disabled. The agency also explained that printing chromatograms would not preserve the dynamic electronic record. That finding concerned the facility and activities addressed in the letter. It is not evidence of similar practices at other testing contractors.
A researcher who receives only a PDF cannot infer from that omission that the contractor lacks appropriate controls. The raw records and histories may be well maintained but excluded from the standard customer deliverable. The meaningful question is whether the relevant records can be obtained and assessed when the purpose of the work requires it.
The certificate has a scope, and the raw data have their own limits
The availability of original electronic files is valuable, but it does not magically enlarge what a test measured. If an assay was designed to examine one property, access to every data point will not prove an unrelated property. Raw chromatographic data alone may not settle molecular identity, total material content, microbial contamination or suitability for a proposed experiment. Those questions require methods and evidence chosen for the particular property.
This is where certificate interpretation becomes essential. The starting questions are plain: which material was submitted, what lot or sample identifier appears on the report, what method was performed, and what do the results actually mean? A readable explanation of reading a certificate of analysis beyond its headline number reinforces that a percentage or a graph must be read alongside the identity of the tested sample and the method used.
There is also a difference between having raw data and having the tools to interpret them. A proprietary file delivered without compatible software, method information or readable export may be practically inaccessible. Conversely, a qualified laboratory might maintain a complete and reviewable archive while providing customers only a summary report unless a more detailed package is requested. The contract and technical expectations determine what is routinely supplied.
A careful reviewer avoids both extremes: treating a polished certificate as complete proof of every material property, and treating the absence of raw files in an emailed attachment as proof that the reported result is false. The defensible position is narrower. The report supports the findings within its stated method and scope; additional review may depend on records held elsewhere.
Outsourcing the analysis should not make the evidence disappear
Specialist analytical contractors perform work that a research organization may not have the instruments or expertise to conduct internally. That division of labor can improve the quality of a measurement. It also introduces an operational dependency: the organization ordering the test may not control the software, archive or retention process that produced the result.
The United Kingdom’s Medicines and Healthcare products Regulatory Agency addressed this issue in its March 2018 GxP Data Integrity Guidance and Definitions. In regulated contexts covered by that guidance, arrangements with third parties should address data ownership, governance, accessibility, archiving and continued readability. The document also discusses access to metadata and audit trails. Those expectations belong to defined regulated activities, but the questions are useful when any research team hires outside analytical services.
Before commissioning work, a customer can clarify whether the deliverable includes a summary report alone, a supporting data package, or access to original electronic records upon justified request. The parties can document which records are retained, for how long, who may retrieve them, how confidentiality is handled, and what happens if the contractor changes software or stops operating. A separate technical agreement may be appropriate for consequential research, while a simple written scope may suffice for routine exploratory testing.
Attention to access should not be confused with a demand to publish unrestricted data. Confidentiality obligations, third-party intellectual property and laboratory security can limit distribution. The objective is an agreed path for qualified review, with a clear description of what is and is not available, not indiscriminate release of every system file.
A backup copy is not automatically a usable scientific archive
A file is not meaningfully preserved merely because it exists somewhere on a server. Later reviewers need to recognize it, open it, connect it to the tested sample and understand the settings that produced the reported result. Long-term access becomes harder when software versions change, user accounts are retired or a contractor migrates data into a new system.
The MHRA distinguishes data retention and archiving from routine backup. An archive must support future retrieval and interpretation; a backup is commonly designed to restore a system after a disruption. In practice, an organization may need both. Protecting files against loss does not by itself guarantee that the analytical history can be reconstructed after an instrument workstation is replaced.
Preservation planning should therefore ask whether a future reviewer will have the data, the relevant metadata and a way to interpret their relationships. When original formats depend on proprietary applications, an agreed human-readable export may help, but it should not silently replace dynamic information that is necessary for the intended review. Decisions about format and retention must reflect the applicable regulatory framework, if any, as well as the scientific significance of the test.
These choices are easiest to make while a contract is being negotiated. After a disagreement has occurred, requesting an unanticipated format or an old software environment may be slow, expensive or impossible. The loss of access can undermine confidence even when nobody doubts the authenticity of the original PDF.
What a research organization can ask for before accepting the report
There is no universal file bundle that guarantees sound science across all analytical methods. A short written request tailored to the study is more useful than a demand for every record without a clear purpose. Questions worth settling include:
- Which exact sample and lot identifiers connect the report to the material submitted for analysis?
- What method and processing information accompanies the result, and what remains in the contractor’s controlled system?
- Can an authorized reviewer obtain relevant original data, metadata and change history if a material question arises?
- What retention period, retrieval process and readable formats have been agreed, including after software changes?
- Who is responsible for resolving discrepancies between the final report and the stored electronic record?
These questions are proportionate to the intended use of the work. A preliminary exploratory assay may need a different level of review from a regulated pharmaceutical release test. The aim is to make the meaning and limits of the deliverable explicit before a result becomes part of a larger scientific claim.
Supplier-facing pages can contribute to that transparency without taking the place of the laboratory’s underlying records. For example, USA Peptide Depot’s documentation standards describe the company’s stated approach to lot identification and the availability of analytical reports where provided. Readers still need to check the actual report for their particular material and understand what supporting data the testing organization can supply.
The missing evidence is often a question of access, not authenticity
A genuine certificate is a meaningful document, but its evidentiary weight depends on the question being asked. It may establish that a named laboratory reported a particular measurement for an identified sample on a particular date. It does not necessarily allow a different analyst to rebuild the measurement from the original acquisition record years later.
That gap is avoidable when the commissioning organization knows what the contractor retains, which electronic records support the conclusion, and how qualified people can retrieve them. The result is a more useful form of transparency: neither blanket suspicion nor blind acceptance, but a documented route from the published number back to the evidence that produced it.
