The eleven terms your file turns on.

The three levels, 01

Sterile barrier system (SBS)

The minimum packaging configuration that prevents the ingress of microorganisms and allows the product to be presented aseptically at the point of use. A pouch and its lidding, a formed tray with its lid, a header bag: whichever assembly is the smallest complete one that holds sterility is the sterile barrier system.

"The packaging" is not a term, it is a shrug. It covers everything from the seal to the shipper in one word, and the requirements do not apply evenly across those layers. Microbial barrier properties, seal integrity, the sterility claim and the aging that supports it all attach to the sterile barrier system specifically. Say SBS and the sentence stops being ambiguous about which layer it means.

Scope statements, protocol titles and acceptance criteria all get sharper for naming the level they apply to. A summary report that says the packaging passed dye penetration has not said which seal on which assembly was tested, and a document that leaves that to be inferred is doing less work than the person who wrote it thinks it is.

The three levels, 02

Protective packaging

The configuration of materials that protects the sterile barrier system and its contents from damage, from assembly through handling, distribution and storage. Shelf cartons, shippers, dunnage, the thermoformed insert whose only job is to stop something moving.

Protective packaging carries no sterility claim, and adding more of it does not upgrade a marginal seal. The dependency runs the other way: thinning or removing protective packaging raises the mechanical demand the sterile barrier system has to survive, which is a change to the basis of a distribution test that has already been run and passed.

Change control, and the fact that a cost reduction on a shipper arrives looking like a purchasing decision. The question to put on it is whether the distribution evidence still describes what now ships. Where the answer is yes it takes ten minutes to write down; the cost is in not asking.

The three levels, 03

Packaging system

The sterile barrier system and the protective packaging taken together: the combination that actually leaves the building. Distribution simulation is run on the packaging system as it will ship. The sterility, barrier and seal claims are made about the sterile barrier system inside it.

The three terms are levels, not synonyms, and swapping them makes a document hard to read against the standard. A file that uses one word for all three has to be decoded before it can be reviewed, and the decoding is done by whoever is reading it rather than by whoever wrote it.

Scoping. When an engagement is written for one product family and one packaging system, that phrase means the sterile barrier system plus everything shipped around it, in one configuration. A second pouch size, a different tray or a different shipper is a second configuration and is agreed before the work starts rather than invoiced afterwards.

Performance

Aseptic presentation

The transfer of the sterile contents out of the sterile barrier system and onto the sterile field under conditions that do not compromise their sterility. It is a requirement on the opening, not only on the closure, and it is a requirement about a person performing an action under time pressure.

A stronger seal is not automatically a better seal. A seal that has grown stronger through sterilization and aging can shed fibers, tear through the porous web instead of peeling cleanly, or force the person opening it into a movement that breaks the field. A seal that has strengthened is a failure mode against this requirement, not a comfortable result.

Our position, stated as ours rather than as a requirement: seal strength acceptance criteria should carry an upper bound as well as a lower one, and aging data should be read for its trend and not only against the floor. Criteria written as a minimum only cannot detect the failure described above, because the number moves in the direction that looks like good news.

This is one of the clearest places where a packaging decision has to connect to the risk management file, which is what Amendment 1:2023 asks for. The hazard is contamination at the point of use, the control is how the package opens, and the evidence is the seal data read in both directions.

Scope

Terminally sterilized

The device is sealed inside its sterile barrier system first and sterilized afterwards, in its final packaging. The package is part of the sterilization process rather than a container for its output: it takes the ethylene oxide and the aeration, or the radiation dose, or the steam with its moisture and pressure, and it has to still be a barrier at the end of it.

Aseptic processing is the other route. There the product is sterilized and then filled or assembled under controlled conditions, and the package never sees a cycle. ISO 11607 is written for the terminal case and its title says so: "Packaging for terminally sterilized medical devices". Where a device is not terminally sterilized, the scope of the standard does not reach it, and conformity to a recognized consensus standard is voluntary in any case. That answer is sometimes the useful one.

Two things belong in the file explicitly rather than by inference. First, which sterilization method the validation covers, because the material compatibility case is method specific. Second, the maximum number of cycles the packaging is validated for, because a reprocessed load or a resterilized device is a different exposure from the one on the protocol.

Seals and integrity

Seal channel

A continuous unfused path running across the width of a seal, from inside the package to outside. It is a leak path through a seal that looks closed, and it can be narrow enough that nothing about the seal's appearance gives it away.

Seal strength, seal integrity and package integrity are three different claims and they are not shown by the same test. Seal strength (ASTM F88) measures the force needed to separate a seal, which characterizes the process window; it does not establish that no channel runs through the seal. Dye penetration (ASTM F1929) looks for a channel in the seal of a porous package. Bubble emission by internal pressurization (ASTM F2096) looks for a gross leak anywhere in the package, including through the material away from the seal.

Choosing a method is choosing which claim the result can support. A file carrying strength data alone has described how the sealer was set up and has not addressed whether the barrier held. The file should say, in words, which method establishes which claim, because that sentence is the one that makes the rest of the data legible.

Aging

Pull point

A scheduled point in an aging study at which samples come out of the chamber and get tested: time zero, any intermediate points, and the point corresponding to the end of the shelf life being claimed.

The last pull point sets the ceiling on the claim. Data pulled at the equivalent of two years supports two years, whatever the label says. And the schedule belongs in the protocol before the study starts. A pull point added after an unwelcome result, or one quietly dropped, converts a test into a search, and acceptance criteria that appear in a record after the results are, in our view, the most serious integrity failure in this work.

The pull schedule is what turns a shelf life claim into arithmetic instead of an assertion, and it sets the calendar. Once the study starts the dates are fixed. The paperwork can be compressed; the aging duration cannot, because the chamber time follows from the arithmetic and the only way to shorten it is to claim less.

Aging

Real-time aging

Samples held at documented ambient conditions for the full duration being claimed, tested at the pull points. Accelerated aging is the elevated temperature substitute that lets a claim be supported before real time has elapsed, using an Arrhenius relationship and a Q10 value you have to be prepared to defend in writing.

Accelerated data does not replace real-time data, it precedes it. The real-time study runs alongside, continues after the submission has gone in, and is what eventually confirms the claim. Our reading, labeled as ours: the Q10 is the assumption most exposed to question, because it is the one number in the arithmetic that comes from judgment rather than from the protocol.

The second distinction is the one that gets rearranged in retelling. Aging is stability. Distribution is performance. They are separate branches off the sterilized product, not a chain, and no order between them is mandated. ASTM F1980 section 1.5 puts distribution, handling and shipping events outside what the aging guide covers, which is why the two arms cannot answer each other's question. Whether to run the branches separately or combine them as a worst case is a documented risk-based decision. Combining them costs you the ability to tell an event-related defect from a time-related one, which is worth knowing before you choose.

Shelf life claims, and the schedule. It also settles a debate that circulates with confidence in both directions: there is no wrong order to point at here, so nobody should be told their sequence is a finding. What belongs in the file is the reason the branches were structured the way they were.

Documents and reviewers

Design history file (DHF)

The compilation of records showing how a device was designed: inputs, outputs, reviews, verification and validation, and the changes along the way. Packaging evidence lives there. The packaging validation plan, the protocols, the test reports, the summary report and the linkage to the risk management file are design records, not a separate binder that happens to be about packaging.

Three files get confused with one another, and they are not the same file. The design history file is how the device was designed. The device master record is how it is built: specifications, process, packaging and labeling. The technical documentation is what a European reviewer examines. FDA's quality system requirement now incorporates ISO 13485:2016 by reference, at 21 CFR 820.7(b) and 820.10(a), so the vocabulary of that standard is in circulation here for about the same body of records that the US regulation called a design history file. Naming the file you mean saves an exchange of emails at the worst possible moment.

A deliverable that has to be rewritten before it can go into the design record is a deliverable that will sit in an inbox. Our reports are written to be filed as they arrive, with the clause recorded beside every finding, so the document works as an internal record without being reformatted first.

Documents and reviewers

AI letter (Additional Information request)

The letter FDA sends during review of a 510(k) when the file does not contain enough for the review to be completed. The submission goes on hold and the review clock stops until the response goes back. The letter names what is missing, and the window is finite: FDA's published policy allows 180 calendar days from the date of the request before the submission is considered withdrawn.

An AI letter is a hold, not a refusal, and it is not a finding against the device. What it costs is set entirely by what closing it requires. If the answer is a document that already exists and needs writing up, that is days. If the answer is a rationale that has to be reconstructed after the fact, that is weeks and it is harder than doing it first. If the answer is a test with a full aging study behind it, the response time is set by the chamber, and nobody can compress it.

It is the reason the packaging section is worth reading before it is filed rather than after. Nobody can tell you what a review will ask, and nobody here will. What can be said is the arithmetic of when to spend: against a hold whose length you do not control, the cost of having the file read is small, and it is spent at a moment when the calendar is still yours.

Documents and reviewers

Notified body

An organization designated by an authority in an EU member state to carry out conformity assessment under the European medical device regulation. Where the device class requires its involvement, it audits the quality system, reviews the technical documentation, and issues the certificate that allows the CE mark to be applied. Some classes are self-certified by the manufacturer and never see one.

A notified body is not a regulator, and FDA is not a notified body. FDA reviews submissions and grants clearance or approval; in Europe the member state authority regulates and the designated third party assesses conformity and issues certificates. The term gets stretched to cover anyone who reads your file. It is worth keeping precise, because the two processes ask for the same packaging evidence in different arrangements, and a file built to satisfy one of them without thinking about the other gets rebuilt later.

Where the term appears in an engagement it means the European reviewer specifically. Our post-delivery support wording names a regulator, a notified body and a certification body separately, in those words, because the three are not the same thing and a support clause naming only one is narrower than it looks when you need it.