Not every lot change is a revalidation trigger, and pretending otherwise

A lot change on a vial septum took our 5890 GC from 0.4% RSD on octanol area counts to 2.8% in a single afternoon, with no other change on the bench. Same method, same analyst, same inlet temperature at 250 °C. The septa were the same part number, same catalog page, same nominal 11 mm diameter. What moved was the lot, and with it the silicone formulation and the bleed profile. That is the kind of event that forces the question this article answers: when a chromatography products supplier or lot changes, what actually has to be revalidated, what only needs verification, and how do you write the change-control entry so an inspector signs off instead of writing a 483.

I have run this drill more times than I care to count, on both sides of the bench: as the lab manager defending a method and as the supplier whose lot change triggered the call. The short version is that most consumable changes do not require full revalidation, and treating every lot change as a revalidation event is its own compliance failure, because it trains everyone to rubber-stamp. The decision has to be risk-based, documented, and specific. Below is how I draw the lines.

Not every lot change is a revalidation trigger, and pretending otherwise erodes your change control

The instinct after a supplier notification is to open a revalidation protocol for everything. Resist it. A revalidation is a formal re-demonstration of method performance against the acceptance criteria in your validation report. It costs analyst time, instrument time, and standards, and it produces a document that has to be reviewed and approved. If you fire that process at a change that only touches a polypropylene autosampler vial body with no contact with the analytical path, you have spent a week proving nothing and taught your reviewers to approve without reading.

The better frame is a three-tier decision. Every supplier or lot change lands in one of three buckets: no action beyond a documented assessment, verification (a targeted check that the change did not move a critical performance attribute), or revalidation (re-execution of the relevant validation parameters). The tier depends on one question: does the change touch a part of the consumable that is in the analytical path, in contact with the sample, or in contact with the detector signal?

That question is what separates a vial cap color change from a septum formulation change. It is also what separates a column lot from a filter lot. Get the tier right and the rest is bookkeeping.

Which consumable changes force revalidation, which need only verification

Here is the decision table I use. It is built around what the part does in the method, not what the supplier calls the change.

The pattern is simple. If the change touches the sample, the mobile phase, the carrier gas, or the light path, it goes to revalidation. If it touches only the housing, the closure, or the packaging, verification or assessment is enough. The trap is the middle column. A vial body that looks cosmetic can become a contact surface the moment the liner is thinner than the wall tolerance, and then it is not cosmetic at all.

Two changes deserve a note because they get misclassified constantly. A column lot change with the same part number is not “no change.” Modern vendors control phase chemistry tightly, but retention time on a new lot can drift by 0.1 to 0.3 minutes on a 30 m GC column, and that is enough to move a late-eluting peak into a co-elution. Treat it as targeted revalidation every time. The second is a septum lot change. The septum is in the inlet, it is heated, and it bleeds. A formulation change that raises siloxane bleed by even a few percent will raise baseline noise and can cost you a low-level detection limit. That is a revalidation event, full stop.

The verification tier is the one that gets skipped, and it is where most failures hide

Verification is the tier between “do nothing” and “revalidate everything,” and it is the one most labs skip because it does not fit a protocol template. Verification is a targeted check: you run the specific test that would catch the failure the change could cause, and you document the result. It is not a full validation. It is not nothing.

For a syringe filter lot change on the same membrane material, verification is a recovery check at the working concentration plus a blank through the filter. If recovery on your representative analyte stays within 95 to 105% of the unfiltered control and the blank shows no peak above the method’s reporting limit, the lot is verified and you move on. That is a 30-minute check, not a 3-day revalidation.

For an autosampler vial body resin change at the same dimensions and same liner, verification is a dimensional check and one system suitability run. Measure the neck inner diameter against the GPI 8-425 or GPI 13-425 spec your closure requires, confirm the crimp or thread engages, and run the system suitability. If the needle seats and the run passes, you are done.

The failure mode I have seen in this tier is the analyst who runs the check but does not write down what would have counted as a fail before running it. A verification with no pre-defined acceptance criterion is not a verification, it is an anecdote. Write the criterion in the change-control entry before the bench work starts.

How to write the change-control entry an inspector will accept

This is where most revalidation packages fall apart, and it has nothing to do with the science. An inspector reading your change-control record is asking three questions: did you know about the change, did you assess its impact, and did you act on the assessment before the change reached the bench? If the record answers those three in order, you pass. If it reads like a retrospective justification, you get a finding.

Here is the skeleton I use. It fits on two pages and it holds up.

• Change description: what changed, the old and new identifiers (part number, lot, formulation, supplier), and the date the supplier notified you or the date you detected it.
• Impact assessment: which method or methods use the part, and which validation parameters could be affected. Name them: system suitability, resolution, tailing factor, detection limit, recovery.
• Tier decision and rationale: revalidation, verification, or assessment only, with the reason stated in one sentence. “Septum formulation change; septum is in the heated inlet and affects bleed; revalidation required for baseline noise and detection limit.”
• Pre-defined acceptance criteria: the numbers, written before the bench work. RSD, resolution, tailing factor, recovery range.
• Execution record: the run, the analyst, the date, the raw data reference, and the result against each criterion.
• Disposition: released, released with a note, or rejected, and the effective date for the new lot or supplier.
• Approval: the person who owns the method signs, not just the person who ran the test.

The single most common finding I have seen is a change-control entry that names the change but never ties it to a validation parameter. “Supplier notified us of a vial lot change; no impact” is not an assessment. It is a sentence that tells an inspector you did not think about it. Name the parameter, state the tier, and show the number.

The second most common finding is timing. The change-control entry has to be opened before the new lot is used in a regulated run, not after a failure forces you to write it backward. If your supplier notification arrives after the lot is already on the shelf, your entry needs a line explaining how the affected lot was quarantined until the assessment closed. Inspectors accept a late notification if the quarantine is documented. They do not accept a late assessment.

Supplier changes and lot changes are different events, and the paperwork differs

A lot change within an approved supplier is a routine event. A supplier change is a qualification event. Confusing the two is how labs end up with a qualified supplier list that does not match the shelf.

For a lot change, the trigger is the supplier’s certificate of analysis plus your own incoming check. For a supplier change, the trigger is a new vendor qualification, and that means the full package: the certificate of analysis, the specification, a sample run, and a documented comparison against the incumbent. If you are moving from one vial supplier to another, that is a supplier change even if the part number is identical, because the manufacturing site and the resin source are different. I have watched a lab switch vial suppliers on the strength of a matching drawing and then discover the septum durometer was 5 Shore A softer, which changed the puncture force and started coring. The drawing matched. The part did not.

This is why I recommend that any supplier change get the revalidation tier by default unless you can show the manufacturing site, the resin, and the liner are identical. In practice they almost never are. If you are evaluating a new source, start with the dimensional and material data before you commit to a bench comparison. The vial supplier side of the market is where the differences live, and a supplier who cannot give you a lot-specific certificate with the resin grade and liner spec is telling you something.

Two field cases where the tier decision was the whole ballgame

Case one: the septum lot that moved the detection limit

Symptom: a GC method for trace volatiles started failing its detection limit check on a new septum lot. The peak was there, but the signal-to-noise on the low standard dropped from about 18:1 to 9:1. The analyst’s first move was to bake the column and recondition the inlet, which cost a day and changed nothing.

The measurement that identified it: a bleed blank run with the inlet at 250 °C and the column held at 40 °C for 10 minutes. The new lot produced a baseline rise of roughly 3 pA above the old lot, concentrated in the siloxane region. That is a formulation difference, not a contamination. The supplier confirmed the silicone grade had changed at the molding site.

The fix: revalidation of baseline noise and detection limit on the new lot, with the acceptance criterion set at the original validation’s S/N of at least 10:1. The lot passed the bleed blank after a 30-minute inlet purge but still sat at 12:1, so the method was released with a note that the low standard be re-prepared on each new lot. The change-control entry named the septum formulation, the bleed measurement, and the release condition. That entry is what an inspector wants to see.

Case two: the column lot that co-eluted a critical pair

Symptom: an HPLC method for a two-component assay started showing a resolution of 1.3 on a new column lot, against a validation criterion of not less than 1.5. Same part number, same phase, same dimensions. The analyst assumed the method was drifting and started chasing mobile phase pH.

The measurement that identified it: retention time on the new lot was 0.22 minutes later than the historical mean for the late-eluting peak, while the early peak was unchanged. That pattern says the phase selectivity shifted, not the mobile phase. A lot comparison against the retained reference column confirmed it.

The fix: targeted revalidation of resolution and retention time on the new lot, with the acceptance criterion tied to the validation report. The lot failed at 1.3, so the method was not released on that lot. The supplier replaced it with a lot that met the criterion. The change-control entry recorded the failed lot, the resolution value, and the replacement. That is a clean record: the change was assessed, the criterion was pre-defined, the lot was rejected, and the disposition is documented.

Both cases share a lesson. The tier decision was not the hard part. The hard part was measuring the right thing before touching the method. A bleed blank and a retention-time comparison are cheap. A week of method development chasing a phantom pH shift is not.

A working checklist for the change-control entry

Before you close a change-control entry on a chromatography products change, run this list. If any line is blank, the entry is not done.

• Is the change described with old and new identifiers, including the lot and the manufacturing site?
• Is every affected method named, and is the affected validation parameter named for each?
• Is the tier decision stated with a one-sentence rationale?
• Are the acceptance criteria written before the bench work, with numbers?
• Is the raw data reference attached and traceable to the run?
• Is the disposition recorded, including any release condition?
• Is the effective date set, and is the affected lot quarantined until that date?
• Has the method owner signed, not just the analyst?

The dimensional and material data you need to fill these lines usually comes from the supplier’s certificate and specification. That is why I tell labs to keep a specification file for every consumable in a validated method, not just the certificate. When the lot changes, you compare against the file. If you are building that file from scratch, the chromatography products catalog is the starting point, and the technical data sheets from vendors like Agilent, Thermo Fisher, and Restek are where the phase and bleed numbers live. For a supplier you are qualifying, Aijiren publishes lot-level data that makes the comparison possible. And when the change is on the vial side, the dimensional and liner specs from vials with caps are what you measure against. If the method is tied to a regulatory framework, the container and holding-time questions sit in SW-846, and the supplier’s own data package should reference the same specifications.

What to do the next time a supplier notification lands in your inbox

Open the change-control entry the day the notification arrives, before the new lot goes on the instrument. Classify the change by whether it touches the analytical path, the sample, or the signal. If it does, revalidate the affected parameters with pre-defined criteria. If it touches only the housing or the closure, verify with a targeted check or document an assessment. Write the rationale in one sentence, name the parameter, and get the method owner to sign. The decision rule is this: the tier follows the contact surface, the acceptance criteria come before the bench work, and the entry closes only when the disposition and the effective date are both recorded. Everything else is a variation on that rule, and a record that follows it will survive an inspection without a rewrite.

References & further reading

Related resources from our lab consumables network:

• Chromatography Products at chromatographyselling.com
• Aijiren at aijirentech.com
• Vials With Caps at vialswithcaps.com
• Vial Supplier at vialsupplier.com

External references and standards:

• Agilent chromatography consumables and columns
• Thermo Fisher chromatography and sample handling
• U.S. EPA SW-846 test methods for waste and materials
• Restek technical resources for chromatography