Date-Stamping Aliquots and Opened Vials: The Minimum Record That Makes a Result Traceable

The minimum record is five fields on the container and one line in a log: what the material is, which lot it came from, what happened to it, the date that happened, and the initials of whoever did it. Everything else — barcodes, inventory software, colour codes — is refinement on those five. A tube carrying them traces backwards to a receiving record and forwards to a result. A tube carrying a felt-tip letter is anonymous the moment the person who wrote it steps away, and every measurement made from it is uninterpretable afterwards.

What follows treats the label and the log as the deliverable rather than as housekeeping, written for a laboratory in the United States or Canada handling research-use-only material. The compounds referred to are not medicinal products, and nothing here concerns anything other than identifying, dating and recording laboratory containers.

What has to be on the container

There is a regulatory floor, and it is lower than most people assume. Under the United States Good Laboratory Practice regulations, each storage container for a test article must be labelled by name, chemical abstract number or code number, batch number, expiration date if any, and, where appropriate, the storage conditions necessary to maintain identity, strength, purity and composition ^[1]. That is the whole requirement for a container as received.

Notice what is missing: no date of anything. A sealed container as supplied has one state, and its expiry or retest date already encodes the time information. An aliquot is a different kind of object — it has a parent, a moment of creation and a diluent, none of it recoverable from the material itself. So the aliquot label carries what the source label never needed: the event, its date, and a pointer back to the parent so the chain does not break at the split.

The storage condition belongs there too, in the terms the source documentation uses rather than as a habit — which conditions a lyophilised solid should be held under is a separate question from recording the answer where the next handler will see it. USP General Chapter <659> supplies the vocabulary the rest of the supply chain is written in — cold, controlled room temperature, and the excursion allowances built into each — so the next person reads the same sentence you did ^[7].

Date of reconstitution and date of first puncture are two different clocks

These routinely share one box on a label, and they are not the same event. Reconstitution is the moment a lyophilised solid became a solution; first puncture is the moment the closure was breached. They coincide when a solid is made up and split immediately, and they diverge constantly: a vial can be punctured on four separate occasions, and a container can be punctured to withdraw diluent with nothing reconstituted at all.

"Date of first puncture" exists as a named field because a puncture starts a clock in compounding practice. USP General Chapter <797>, official since November 2023, caps an entered multiple-dose container at its assigned beyond-use date or twenty-eight days from puncture, whichever is shorter, where antimicrobial effectiveness supports the longer figure ^[6]. That standard governs sterile compounding in a clinical setting and is not a rule for research material — but it is where the convention comes from, and anyone who has worked near a pharmacy will expect the field to be there.

Purely analytical practice reaches the same place from another direction. WHO's good practices for pharmaceutical quality control laboratories expect reagents and prepared solutions to be labelled with what they are, the date prepared or opened, their shelf-life or validity, the storage conditions and the person who prepared them ^[8]. No clinical setting anywhere in that document — just the observation that a container of unknown age is a container of unknown quality.

The instruction is unglamorous: separate fields, and neither overwrites the other. A label reading "opened 14 Mar" cannot say whether the material spent March as a powder or as a solution — and material in solution ages on a different schedule from the same material still dry — so that one ambiguity destroys most of the value of having written a date.

The label is a pointer; the log is the record

A label is a few square centimetres of abradable surface in a cold, wet place. It cannot hold a history. Its job is to carry an identifier unique enough that exactly one row in a log matches it; everything else lives in the log, where there is room to be complete and where corrections leave a trace.

ISO/IEC 17025 sets what those technical records must achieve: enough information to identify the factors affecting a result, and enough to repeat the activity under conditions as close as possible to the original. It also requires amendments to be traceable to the previous version, with the date of the change and the person who made it ^[4]. That second clause is the one quietly broken, by correcting a log with a fresh sheet and no history of what it replaced.

The discipline has a name worth knowing, because it is the vocabulary an auditor will use. The MHRA's data integrity guidance defines records as needing to be attributable, legible, contemporaneous, original and accurate — ALCOA — extended in practice to complete, consistent, enduring and available ^[3]. Contemporaneous fails first, and it fails invisibly: a log filled in on Friday for a split performed on Tuesday looks identical to a correct one and is not.

The Good Laboratory Practice regulations add the movement half. A handling system must ensure proper storage, preclude contamination or deterioration in distribution, maintain identification throughout, and document the receipt and distribution of each batch including the date and quantity distributed or returned ^[2]. Read that with an aliquot in mind and the mapping is exact: every time material leaves a parent container, that is a distribution event carrying a date and an amount.

The access count and the freeze–thaw log

Access count is the field most often estimated at the end and the easiest to keep exactly, because it is a tally rather than a measurement. Every removal from storage counts, including those where nothing was withdrawn: a tube taken out, warmed in a gloved hand while a rack was searched, and put back has had the same physical experience as one that was used. Counting as it happens costs a pen stroke; reconstructing it in November for work done in June is guessing, and it guesses in the direction that flatters the sample.

• Date and time out, and the initials of the person removing it.
• Whether material was withdrawn, and roughly how much of the container was consumed.
• Whether it returned to storage or was consumed and discarded.
• The running access number after this event — one integer, incremented, never recalculated.
• Where it went if it left the room: an aliquot that travelled to another bench and back has left the chain of custody the record exists to maintain ^[2].
• Any deviation worth a sentence — a warm cabinet, a delayed return, a closure that did not reseal cleanly.

What that count is for is a separate subject; the damage repeated freezing and thawing does, and how aliquots are sized against it, is treated elsewhere rather than here. The point of the field is metadata. A count is a covariate you can put in a methods section and use to explain an outlier that would otherwise have no candidate explanation to test. Repository practice treats this as basic, expecting each container to carry a unique identifier linked to a record of its handling history, precisely so later analysis can ask whether handling accounts for what was seen ^[9]. The international biobanking standard makes the same demand in more abstract language, requiring traceability through every processing step ^[5].

Be honest about the size of the problem this addresses. In a widely cited survey of 1,576 researchers, more than seventy per cent reported having tried and failed to reproduce another scientist's experiment, and more than half had failed to reproduce one of their own ^[10]. That is self-reported opinion, not an audit of laboratory records, and should not be quoted as a measured failure rate. Nobody has shown that better aliquot labelling fixes reproducibility. What labelling does is make the question answerable — a smaller claim, and a true one.

The best access log is the one never kept. Splitting into single-use aliquots at first preparation makes the count trivially one for every tube, collapses the whole problem into a single documented event, and deletes an entire category of record-keeping error. Where the work allows it, that is a record design decision rather than a storage one.

Labels that survive the container they are on

A perfect record on a label that falls off in a freezer is not a record. The physical failures are boringly predictable and worth designing against once.

• Apply the label while the container is dry and at room temperature. Adhesive applied to a frosted surface bonds to the frost, and the frost leaves.
• Use a solvent-resistant marker or labels rated for the storage temperature. Ballpoint and many inkjet inks lift under alcohol, and alcohol is what wipes down a bench.
• Label the container itself, not only the bag, box or rack. Racks get emptied and re-sorted; the tube must be self-describing alone.
• Do not cover the manufacturer's label on a source vial — the lot number underneath anchors the whole chain. Use a second label or a tail wrap.
• Add a machine-readable code where the identifier is long enough to be mistyped ^[9].
• Assume condensation. A tube out of cold storage acquires a water film in seconds, which is exactly when someone tries to write on it.

Electronic logs have their own version of the same trap. A spreadsheet anyone can edit, with no history of who changed what and when, is a convenient index rather than a record. The regulatory expectation for electronic records is a secure, computer-generated, time-stamped audit trail independently recording the date and time of entries and of actions creating, modifying or deleting a record ^[11]. Few small laboratories need a validated system to that standard, but the gap should be a conscious decision: a dated paper sheet that cannot be silently rewritten is stronger evidence than an unversioned spreadsheet, and cheaper.

What a complete entry looks like

Two properties make this work, and both are cheap. Identifiers are hierarchical, so the aliquot ID contains the parent ID and a person holding one tube can name its parent without consulting anything. And the log holds one row per object and one row per event, never a cell edited in place. A record that grows by appending can be audited; a record that grows by overwriting cannot, whatever it is written on ^[3].

The failures this actually prevents

1. The unassignable tube: two similar names, one abbreviation, and a box that now holds an unknown. It is not recoverable as an identified article.
2. The lot question you cannot answer. A supplier reports a problem with a lot; without that number on every derived container, the only safe response is to quarantine material that was never affected ^[1].
3. The undatable result — no way to say whether the material behind it was prepared that week or the previous quarter, which makes it unrepeatable as a method ^[4].
4. The unattributable correction: a log tidied afterwards with no record of what it said before, turning a documentation slip into a credibility problem covering the whole file ^[3].
5. The silent freeze–thaw history — an anomalous result with no handling record to interrogate, so the one free hypothesis never gets tested ^[5].
6. The orphan found behind a rack carrying only a compound name, indistinguishable from correctly stored material and useless for anything that has to be defended.

None of this makes material better than it was. It makes material identifiable, which is the precondition for every other claim anyone might make about it: that this tube came from that lot, that it was prepared on a known date by a known person, that it has been out of storage a known number of times. Traceability is not a quality attribute of the compound but of the record — and it is the only one entirely within the laboratory's control. The compounds discussed here are supplied for research use only. They are not medicinal products, they are not intended for administration to humans or animals, and nothing above should be read as guidance for anything other than labelling, storing and documenting laboratory containers.