Validating PHMB Foam Dressing OEM Claims: A Step-by-Step Pilot and Lab Test Guide

By the end of this guide you will hold a signed-off validation file: a claim-to-test matrix, a pilot-lot sampling plan, written acceptance criteria and a record set that survives a tender review or a notified-body audit. Two limitations frame everything that follows. EN 13726:2023 sets test methods and prescribes no pass/fail limits, which is what the standard’s published scope states in its own scope sentence. And an in-vitro antimicrobial result is not a clinical claim, which is why the antimicrobial performance standard sits beside the physical-test standard rather than inside it. This guide is written for overseas medical-supply distributors, US and EU OEM procurement teams, and hospital wound-care procurement, and it treats the supplier’s claim sentence as the raw material the file has to convert into evidence.

What does EN 13726:2023 actually test, and which method backs which PHMB foam dressing claim?

EN 13726:2023 is the current European test-method standard for wound dressings, and its scope covers absorption, moisture vapour transmission, waterproofness and extensibility. The standard was approved by CEN on 17 April 2023 and consolidates four earlier parts into one document, superseding EN 13726-1:2002 (with its 2004 corrigendum), EN 13726-2:2002, EN 13726-3:2003 and EN 13726-4:2003, as the supersession list in the standard’s full annex and clause list records. The older series was four parts, not three, and a supplier citing “EN 13726-1” is citing a withdrawn document.

The annex inventory is where a claim-to-test matrix starts. Free-swell absorptive capacity (Annex B) is normative, fluid handling capacity (Annex E) is normative, MVTR in contact with vapour (Annex H) and with liquid (Annex I) are normative, waterproofness (Annex J) is normative, and extensibility and permanent set (Annex K) is normative. Fluid retention capacity (Annex C) and absorption under compression (Annex D) are informative, which is which annexes are normative versus informative in the standard’s own contents list. A buyer comparing “absorbency” numbers across two suppliers may be comparing one normative method against two non-mandatory ones.

Two corrections matter here. Conformability is not a separate annex in the 2023 revision; the mechanical assessment behind a conformability-adjacent statement is Annex K, and the standard’s rationale annex is Annex A, per the same contents list. And the antimicrobial claim is not covered by EN 13726 at all: EN 17854:2024 is the antimicrobial-specific European standard and it sets performance thresholds, which is the asymmetry the matrix has to capture. EN 13726 measures physical performance; EN 17854 sets an antimicrobial bar.

How do I turn a supplier’s claim sentence into a testable endpoint?

Take the claim sentence apart clause by clause, and for each clause name the document that would substantiate it. A typical PHMB foam claim reads that the foam contains the antimicrobial substance polyhexamethylene biguanide (PHMB, polihexanide), which kills and inhibits the growth of bacteria. That single sentence contains three separate evidence demands: a content or release assay for the PHMB loading, an EN 17854 performance test with a stated log threshold, and a claim-scope statement that stops short of clinical outcomes.

The scope limit is not optional. Under MDR Article 61, conformity with the General Safety and Performance Requirements includes a clinical evaluation, and claims about clinical performance must be supported by appropriate evidence; bench or laboratory testing may support a technical or antimicrobial property but will not by itself substantiate a clinical outcome claim, as the distinction between a bench claim and a clinical claim sets out. MDR Article 7 separately prohibits misleading claims.

Method fit is the second trap. ISO 22196:2011 is a 24-hour contact assay at 35 C in a humid environment, tested by default against Staphylococcus aureus and Escherichia coli, and its non-porous-surface scope is why it under-represents an absorbent multilayer foam, per the limits of the plastics assay. ISO 20743:2013 was withdrawn and replaced by ISO 20743:2021, which covers nonwovens. An ISO 22196 log reduction offered as evidence for a foam dressing claim is a method mismatch, not a weak result.

Verification checkpoint: for every clause of the claim sentence, you can name the document that would substantiate it, or you have identified the clause as a clinical claim that bench testing cannot carry.

How many pilot lots and how many units per lot should I ask for, and how should they be sampled?

No standards body, regulator or peer-reviewed paper states a required pilot lot count or units per lot, so treat the numbers below as what an experienced buyer asks for, with the reasoning shown, not as standard text. Practitioner guidance converges on two pilot lots per critical SKU group, one to expose setup issues and a second after material replenishment to show repeatability, plus a first-commercial-lot verification. Roughly 100 finished units per lot per critical SKU covers destructive testing, packaging and sterilization checks, shipping samples and retains, rising to 200 to 500 for high-risk sterile dressings or many test conditions.

Sampling location carries more weight than total count. Draw units from the beginning, middle and end of the run, across each machine lane, cavity and shift, so that within-lot variation is visible rather than averaged away. A single grab sample from the middle of one shift tells you almost nothing about a lot.

Two practical notes on sample custody. Ask for the sampling record to name who drew the units, when, and from which lane and shift, because a test report without a traceable sample origin cannot be tied back to a production condition. And keep the retains under the same storage conditions as the shipped product, since a retain stored differently is not a retain.

How do I write acceptance criteria when the standard gives no pass/fail limits?

The standard’s clause 4 is titled “Requirements”, but its scope is test methods only, so the acceptance limits that turn a result into a pass or fail come from the manufacturer’s specification, intended use and risk analysis, not from EN 13726. That is the sentence to hold onto when a supplier sends a test report with a number and no specification: the number is a measurement, not a verdict.

Write each criterion as a specification line with four parts: the method and annex, the unit, the limit, and the basis for the limit. The basis is what distinguishes a defensible criterion from an arbitrary one. It can be the supplier’s own design specification, a risk-analysis output, a clinically relevant performance threshold tied to intended use, or a comparison against a predicate product. What it cannot be is “the number the supplier reported”.

Two cautions on the numbers themselves. Published evaluations typically use five specimens per dressing per method, which is reported practice in peer-reviewed wound-dressing studies rather than a stated requirement of the standard text, so agree the replicate count in the protocol rather than assuming it. And EN 13726:2023’s own rationale annex notes that the Annex B to Annex E tests are based on fully saturated dressings, which is rarely the clinical situation, and that most dressings leak long before saturation. A criterion written against a fully saturated test condition should say so.

What PHMB-specific evidence should I request, and what claim scope is legitimate?

Request three documents for the antimicrobial claim: a content or release assay that states the PHMB loading and the method used, an EN 17854:2024 test report that states the log threshold and the organisms tested, and a written claim-scope statement that maps the marketing sentence to the evidence. EN 17854:2024 specifies minimum requirements and a test method for microbicidal or microbistatic activity and applies to all wound dressings claiming antimicrobial activity, per the antimicrobial performance standard. Reported thresholds are at least a 3-log reduction at the mandatory 24-hour exposure against all three specified test organisms for a microbicidal claim, and no increase in recoverable organisms for a microbistatic claim, using three test-dressing replicates and three negative-control replicates. Those threshold figures are reported by standards-catalogue and distributor listings of the standard rather than read from the standard text, so verify them against the copy you buy.

Loading is the parameter most often left unstated. A peer-reviewed review of PHMB-releasing wound dressings reports that, where manufacturers disclose the figure, loading is by impregnation from PHMB solutions of 0.1% to 0.5%, with antibacterial activity lasting up to 3 to 7 days, per what dressings actually contain. The same review’s commercial table lists Kendall AMD polyurethane foam at 0.5%, Suprasorb X + PHMB at 0.3%, McKesson PHMB hydrophilic foam at 0.8% to 1%, and Curity, Excilon, Kerlix and Telfa AMD at 0.2%, with several products leaving the loading undisclosed. A “PHMB foam” claim says nothing about loading, and 0.5% and 0.9% behave differently against an EN 17854 threshold.

Safety evidence is the gap to name plainly. The EU Scientific Committee on Consumer Safety could not assess the safety of PHMB as a cosmetic preservative at up to 0.3% because no firm conclusion on genotoxic potential could be drawn, required a properly performed gene mutation study before further consideration, applied a default 10% dermal absorption in the absence of a valid dermal absorption study, and did not recommend use in spray formulations; PHMB is a CMR substance, per the PHMB safety question behind the claim scope. That opinion addresses cosmetic preservative use, not a blanket prohibition, and it states that exposure from all possible sources is required for the safety calculation. The dressing literature reports loading concentrations and activity duration but no exposure or systemic-safety calculation for any commercial PHMB-releasing dressing, so the safety file is a question to put to the supplier, not an assumption to carry.

What should a stability plan contain, and how does it support a shelf-life claim?

Start from the duty: ISO 11607-1:2019 and ISO 11607-2:2019, both with Amendment 1:2023, prescribe no shelf life, and the manufacturer must justify the expiry period with stability evidence, per the packaging standard behind the shelf-life claim. A supplier that quotes a three-year shelf life without a stability plan has quoted a number, not a claim.

The plan itself borrows its structure from the pharmaceutical stability framework. ICH Q1A(R2) sets an accelerated condition of 40 C / 75% RH with a minimum of six months and at least three time points including initial and final, alongside long-term and intermediate conditions, per the stability framework a dressing stability plan borrows from. Applied by analogy to a dressing, that gives you time zero after sterilization and packaging, accelerated points, and real-time intervals running in parallel.

Accelerated aging is the second branch, and the arithmetic is checkable. ASTM F1980-21 is the guide for accelerated aging of sterile barrier systems and medical devices, using an acceleration factor of Q10 raised to (T AA minus T RT) divided by 10, per the accelerated-aging guide. With the typical Q10 of 2 and an accelerated temperature of 55 C against a 25 C reference, the factor is 8, so about 137 days of accelerated aging represents a 3-year claim and about 228 days represents a 5-year claim. Those durations are a laboratory’s conventional application of the equation, not a duration ASTM mandates, so present them as the calculation a buyer can reproduce.

The distinction to hold is that accelerated data support a claim while real-time data defend it. A tender reviewer or notified body will accept accelerated aging as the basis for a shelf-life statement and will still expect the real-time study to be running.

Which records must survive an audit or tender review, and who on my side asks for what?

The record set has a legal floor. Under MDR Article 10(8), the manufacturer must keep technical documentation, the EU Declaration of Conformity and applicable certificates available to competent authorities for at least 10 years after the last device covered by the Declaration was placed on the market, per the MDR articles behind the evidence file. Annex IX of the MDR is not a test list: it is the conformity-assessment route combining assessment of the quality management system with assessment of the technical documentation, and Article 10(4) requires the manufacturer to draw up and keep up to date technical documentation containing the Annex II and III elements.

On the US side, the transition date has passed. FDA’s Quality Management System Regulation makes 21 CFR Part 820 effective 2 February 2026 and incorporates ISO 13485:2016 by reference; from that date FDA no longer uses QSIT and investigators may review records including those created before 2 February 2026, per the QMSR transition date. A supplier describing its US compliance as “21 CFR 820 QSR” as if QMSR had not taken effect is working from an out-of-date framework.

The certificate-scope check is the one buyers skip most often. Read the scope line by line and note which activities are covered for which product category: a certificate that lists design and development for some dressing categories and only production and distribution for others tells you where the supplier’s own design responsibility sits. The biological-evaluation and sterilization standards named in a Declaration of Conformity are the same document types to request as test reports, so the DoC doubles as a request list. On biocompatibility, ISO 10993-1 requires biological evaluation inside a documented risk-management process rather than a fixed test menu: define the device, materials, manufacturing and sterilization residues, contact nature and duration; review existing information including material characterization and prior testing; identify data gaps; and generate only the additional data needed, per the biocompatibility evidence set. A dressing contacting breached or compromised skin draws cytotoxicity, sensitization, irritation or intracutaneous reactivity and acute systemic toxicity endpoints. Note that the underlying standard edition is changing, so refer to the current edition of ISO 10993-1 rather than pinning a year.

Common mistakes to avoid when validating PHMB foam dressing OEM claims

The recurring failures in validating PHMB foam dressing OEM claims are method mismatches, missing scope limits and records that cannot be traced back to a production condition. Each one is avoidable at the request stage.

Accepting an ISO 22196 result for a foam dressing is the most common method mismatch. ISO 22196:2011 is a non-porous-surface assay, and its scope under-represents an absorbent multilayer foam, per the limits of the plastics assay. The product-specific route now exists: EN 17854:2024 applies to any wound dressing specifically claiming antimicrobial activity, per the antimicrobial performance standard.

Treating a log reduction as proof that the dressing prevents infection is the second. Bench data supports a technical or antimicrobial property and will not by itself substantiate a clinical outcome claim, per the distinction between a bench claim and a clinical claim. A clinical evaluation report is the document that carries a clinical claim, and MEDDEV 2.7/1 rev. 4 requires it to identify residual risks and uncertainties such as medium- and long-term performance and safety under wide-spread use.

Three more, briefly. Comparing absorbency numbers without checking whether they came from a normative or an informative annex compares unlike methods, per which annexes are normative versus informative. Accepting a test report with no specification line leaves you with a measurement and no verdict, because the acceptance limits come from the manufacturer’s specification, intended use and risk analysis rather than from EN 13726. And citing a withdrawn method, whether EN 13726-1:2002 or ISO 20743:2013, dates the whole evidence file, since the 2021 and 2023 editions are the ones a current claim should be read against.

Frequently Asked Questions

Can an ISO 22196 or ISO 20743 report stand in for EN 17854?

No. ISO 22196:2011 measures antibacterial activity on plastics and other non-porous surfaces, so it describes the limits of the plastics assay rather than the behaviour of an absorbent multilayer foam. ISO 20743 addresses the textile method’s scope, which covers fabrics, not a dressing that has to manage exudate. EN 17854:2024 is the standard written for antimicrobial wound dressings, and it is the report to ask for when a supplier claims the dressing kills or inhibits bacteria.

The supplier sent a certificate but no protocol. What now?

Treat the certificate as an index entry, not as evidence. Ask for the test protocol and the raw data behind it: the method and annex used, the challenge organisms, the contact time, the sample preparation, and the laboratory that ran it. A certificate names a result; only the protocol shows what was tested and under what conditions. If the supplier cannot produce the protocol, the claim is unverified regardless of how the certificate looks.

Is EN 13726:2023 data generated under the older 2002-2003 parts still usable?

Only with care. The four replaced parts were withdrawn when the 2023 edition took effect, so reports referencing them describe methods that no longer define the current state of the art. Ask the supplier to confirm which edition the testing followed and, where the older parts were used, to state the differences that affect the result. For a new submission or a tender, request testing to the 2023 edition.

Does the FDA QMSR transition on 2 February 2026 change what a US buyer should request?

Yes, in one practical respect. From the QMSR transition date, the FDA’s quality system requirements are aligned with ISO 13485, so a US buyer can ask for the same quality management evidence a European buyer asks for instead of maintaining a separate set. The underlying expectation does not change: the manufacturer still has to show that the design inputs, verification testing and records support the claims on the label.

Can a PHMB loading figure be an acceptance criterion on its own?

No. A loading range tells you how much PHMB went in, not how much is released, how much stays active, or whether the dressing performs as claimed. Published figures for what dressings actually contain run from 0.1% to 1%, and several products do not disclose the number at all, so a loading value is a starting point for the specification rather than a pass/fail line. Pair it with a content or release assay and with the antimicrobial performance test, and set the limit from your own risk analysis.

Conclusion

The file you end with is the deliverable, not the test report. A claim-to-test matrix that names the annex and its normative status, a sampling plan that draws from the start, middle and end of the run across lanes and shifts, acceptance criteria that state their basis because the standard supplies none, and a record set that traces every result back to a production condition: that is what a tender reviewer or notified body can actually assess. The two framing limitations hold throughout. EN 13726:2023 measures physical performance and sets no limits, and an in-vitro antimicrobial result supports a technical property rather than a clinical claim.

The next step is a document request, not a purchase order. Ask the supplier for the specification sheet, the instructions for use, the certifications pack, the pilot protocol and the stability plan, and read the certificate scope line by line before you read the test reports. If the supplier can produce those five documents and explain the basis for each acceptance limit, you have a validation file. If they cannot, you have learned that before the order rather than after the audit.