Device Label Durability: Testing the Actual Cleaning Exposure

What Must Survive Cleaning: Required Outcomes, Not Required Tests

A design reviewer who asks whether an exterior label or unique device identifier (UDI) carrier will stay attached and legible is not asking for a peel number. There is no universal pass/fail adhesion standard that shows a label or direct mark will survive the reprocessing in the instructions for use (IFU). Build the file in the other direction: write the claim first, copy the exposure from the validated instructions, and only then choose a test that can speak to that exposure on the real substrate.

Two legal outcomes matter, and neither one names a laboratory method:

• In the United States, 21 CFR 801.45 requires a permanent UDI marking on the device itself when the device must bear a UDI on its label, is intended to be used more than once, and is intended to be reprocessed before each use. FDA's final guidance Unique Device Identification: Direct Marking of Devices (issued November 17, 2017) is not a regulation. It interprets the triggers. "Intended to be used more than once" means repeated use on or by different patients; repeated use only on or by the same patient does not create the duty. "Intended to be reprocessed" means high-level disinfection, sterilization, or both, before each use or between uses. A device that is only cleaned, or only low- or intermediate-level disinfected, is not required to be directly marked under 21 CFR 801.45. The guidance expects the mark to last the expected service life, taking into account expected use and reprocessing in the manufacturer's instructions. It does not define "permanent" as survival of every cleaning wipe, and it does not require both plain text and automatic identification and data capture (AIDC) on the direct mark. Under 21 CFR 801.45(c), either form, or both, is allowed. The guidance also states that class I devices bearing a Universal Product Code (UPC) on the label and device packages are excepted under 21 CFR 801.40(d). The current text of 21 CFR 801.45(e), amended December 4, 2025, requires the basis for an exception in 21 CFR 801.45(d) to be documented in the design and development files required by 21 CFR 820.10(c).

• In the European Union, MDR Annex VI Part C Section 4.10 requires a reusable device to bear a UDI carrier on the device itself. Where the device requires cleaning, disinfection, sterilisation, or refurbishing between patient uses, that carrier must be permanent and readable after each process used to make the device ready for the next use, throughout the intended lifetime. The European Commission UDI helpdesk quotes that text and states two exceptions: direct marking would interfere with safety or performance, or direct marking is not technologically feasible. The provision does not name a peel method or a barcode grade. "Readable" follows the carrier that was applied.

Neither 21 CFR 801.45 nor MDR Annex VI Part C Section 4.10 mentions ASTM D3330, UL 969, or the IEC 60601-1 rub test. The binding outcome is narrower than "the label survives cleaning." In the United States the direct mark has to last the expected service life under the reprocessing in the instructions, and only when the direct-marking triggers are met. In the European Union the carrier has to stay permanent and readable after each listed process for the intended lifetime. A study is how a manufacturer evidences that outcome. It is not a method the law prescribes. IEC 60601-1, UL 969, and ASTM D3330 remain voluntary tools with stated scopes.

Where the Actual Cleaning Exposure Comes From

The exposure configuration is not a chemical menu written for the label study. It is the processing information the manufacturer already has to publish. Under EU MDR Annex I GSPR 11.2, where necessary, devices shall be designed to facilitate their safe cleaning, disinfection, and/or re-sterilisation. GSPR 23.4(n) requires the instructions for use of a reusable device to give information on the appropriate processes for allowing reuse, including cleaning, disinfection, packaging and, where appropriate, the validated method of re-sterilisation for the Member State or Member States where the device is placed on the market. The instructions must also identify when the device should no longer be reused, for example signs of material degradation or the maximum number of allowable reuses. "Where appropriate" attaches to the validated re-sterilisation method in that clause. It does not, by itself, turn every cleaning step into a named laboratory method for the label.

ISO 17664-1:2021 specifies the information to be provided for processing critical and semi-critical devices, and devices intended to be sterilised, from point-of-use treatment through cleaning, disinfection, and sterilisation. It specifies the content of the instructions. It does not define the validation methods, and it does not cover non-critical devices unless they are intended to be sterilised. FDA's March 2015 guidance, Reprocessing Medical Devices in Health-Care Settings: Validation Methods and Labeling, recommends that reprocessing instructions be formulated and scientifically validated in enough detail for a facility to reproduce them. A 510(k) exemption does not remove labeling duties or the quality-system duties that still apply to that device. ISO lists 17664-1:2021 at systematic-review stage 90.20 as of July 15, 2026, so confirm the edition in the file.

The current instructions are the exposure baseline: agents, concentrations, contact conditions, and any stated limit on cycles. If those instructions change, the durability evidence has to be checked against the change. Adjacent guides already cover how to write and validate that baseline, including audit-ready reprocessing instructions, ISO 17664 processing information, and cleaning-validation endpoints for reusable instruments. This article does not restate those methods. It uses their output as the exposure the label or mark has to survive.

The Chemistry a Label Actually Faces

Supplier phrases such as "chemical resistant" or "medical grade" do not name an agent, a concentration, a contact method, or a cycle count. They are not a durability claim. Healthcare germicides are different classes used for different tasks. A result against one class does not stand in for another, which is why the claim has to name the products in the IFU.

The CDC Guideline for Disinfection and Sterilization in Healthcare Facilities, on the chemical-disinfectants page maintained from the 2008 guideline, describes classes and typical uses. It is not a label-compatibility standard, and it does not give one wet-contact time a label protocol can copy. Contact time depends on the product, the organism, and the soil. The table is limited to statements in that guideline. The last column is the instruction for the durability protocol, not a CDC test result.

An isopropyl-alcohol exposure does not evidence a hypochlorite wipe, an aldehyde soak, or a washer-disinfector cycle. When the IFU specifies an automated washer, the durability run has to include the detergent and the thermal conditions of that validated cycle. Washer-disinfector validation and the A0 calculation are a separate subject, covered in the ISO 15883 washer-disinfector guide. This article does not assign a washer temperature.

What Each Standard Really Tests

Each voluntary standard below answers a narrower question than "the label survives reprocessing." Reading a component method as if it were finished-device evidence is the mistake the rest of this section is meant to prevent.

IEC 60601-1 Clause 7: The Short Solvent Rub Test

For medical electrical equipment, marking requirements sit in IEC 60601-1 Clause 7. The rub procedure used in test plans is subclause 7.1.3. The solvents and times below are taken from MECA's Edition 3.1 evaluation package for IEC 60601-1:2005 plus Amendment 1:2012. They are not a transcription of the Edition 3.2 consolidated text (2005 + Amendment 1:2012 + Amendment 2:2020). In that Edition 3.1 package, the marking is rubbed by hand with a cloth for 15 seconds with distilled water, 15 seconds with ethanol 96 percent, and 15 seconds with isopropyl alcohol, and the legibility check follows. Amendment 1 replaced the earlier solvent wording "methylated spirit" with ethanol 96 percent. Older summaries still say methylated spirit. The Edition 3.2 preview still places identification, marking, and documents in Clause 7. Confirm subclause 7.1.3 in the purchased Edition 3.2 text before copying it into a protocol.

The Edition 3.1 package records the pass condition as required markings that remain clearly legible during expected service life in normal use, with the rub followed by the legibility check. A separate marking summary states that adhesive labels shall not have worked loose or curled at the edges. Confirm that curl criterion in the edition you purchased before you copy it into a protocol. Three 15-second wipes are a solvent screen. They are not repeated clinical reprocessing, hypochlorite exposure, or a washer-disinfector cycle. The Edition 3.1 label checklist records clause 11.6.6 as cleaning and disinfection with the agents and process specified in the instructions for use. When IEC 60601-1 applies, that line is why the IFU sets the cleaning exposure and the rub does not. The checklist is not a substitute for the standard.

UL 969 Edition 6: Component Label System Qualification

UL 969 is the standard for marking and labeling systems. Edition 6 was ANSI-approved on July 24, 2025, with a revision dated July 28, 2025. The public scope covers labels and related products used as permanent nameplates or markings. Edition 6 section 1.2 also includes direct part markings. Systems are evaluated for specific uses and for specific surfaces that are essentially smooth, flat, and rigid, unless another surface is specified. Section 1.3 states that the acceptability of a label system in a particular application is judged under the standard covering the end product.

Do not cite Edition 6 chemical-exposure clause numbers from a free PDF. A vendor-posted file of UL 969 is the Fourth Edition, October 3, 1995, including revisions through September 19, 2014. That older text contains Section 7, Exposure Conditions, including 7.2 for specific agents and Table 7.4. Edition 6's July 28, 2025 revision note corrects paragraph 8.1.2 and a reference to Table 8.2, so the 2014 table numbers are not a safe citation for Edition 6. Take any agent-exposure table from the purchased Edition 6 text. A UL Recognized Component mark means the system was evaluated for the uses and surfaces in the recognition. It does not mean the finished device meets 21 CFR 801.45 or MDR Annex VI Part C Section 4.10.

ASTM D3330: Standard Peel Force vs. Clinical Function

ASTM D3330/D3330M-04(2018) measures peel adhesion of pressure-sensitive tape. In that text the specimen is 24 mm (1 inch) wide. Method A peels a single-coated tape at 180 degrees from a standard steel panel, or another surface of interest, at 5.0 mm/s plus or minus 0.2 mm/s. Method F is the 90-degree peel, at the same rate. Method B measures adhesion to the tape's own backing. Method D measures adhesion of the release liner. The 2018 procedures contain no disinfectant conditioning.

A peel result on an unexposed steel panel answers a quality-assurance question about that tape. It does not answer whether the device label is still attached after the IFU disinfectants. Packaging Compliance Labs describes ASTM D3330 as a 1-inch specimen peeled at 180 degrees from a stainless-steel plate, with a 90-degree variation, used to compare label materials and to watch adhesion through accelerated aging. CS Analytical describes ASTM D3330 Methods A through F and ASTM D6252, then distribution and shelf-life programs of about 4 to 12 weeks, with defect checks such as edge lift, flagging, and channeling. CS Analytical states there is no formal label-adhesion procedure for those distribution and storage conditions. Neither page tests a reusable device through the cleaning and disinfection cycles in its IFU.

The Configuration-to-Claim-to-Evidence Map

Use one matrix in the design file. Each row names the claim, the basis, the exposure that belongs in the test, the record to keep, and whether the row is a legal duty, a voluntary method, or evidence you owe only because you made the claim.

An ASTM D3330 certificate cannot stand in for the finished-device claim. D3330 is an incoming check on tape. 21 CFR 801.45, when it applies, requires a permanent direct mark that lasts the expected service life under the manufacturer's reprocessing instructions, in plain text, AIDC, or both. MDR Annex VI Part C Section 4.10 requires the reusable-device carrier to stay permanent and readable after each listed process. Neither provision requires a Data Matrix, and neither provision accepts a steel-panel peel as the showing.

Writing the Protocol: Cycles, Agents, and Pass Criteria

A protocol that a reviewer can follow has four decisions. None of them is supplied as a default number by 21 CFR 801.45 or by MDR Annex VI.

1. Substrate. A polished steel panel stands in for the device only when the labeled surface is that alloy and finish. Otherwise use production housings, or plaques of the production resin and texture. Name the material in the protocol. Examples include a polycarbonate blend, PEEK, or passivated stainless steel only when that is the actual surface.

2. Cycle count. Take the count from the service life and reuse frequency already in the design and risk files. The arithmetic belongs to the manufacturer. As an illustration only, a file that states a 2-year life and 2 reprocessing cycles a day is 2 x 365 x 2 = 1,460 cycles before any margin and before a leap day. If the protocol adds a margin, or compresses those cycles into a shorter laboratory schedule, the protocol has to justify that model. The cited regulations do not supply a 1.2x or 1.5x factor, and this article does not recommend one.

3. Application method. Match the IFU. A wipe is chemical contact plus shear. Immersion and a washer cycle are not reproduced by setting a coupon in a beaker of the same chemical. If the protocol uses a mechanical wipe, record the wipe, the stroke count, and the normal force, and say why those settings represent the instructions. This article does not set a force value. Contact time comes from the disinfectant instructions the IFU adopts, not from a generic hospital dwell.

4. Acceptance criteria, written before the test. At minimum, the label or mark is still attached, with no loose edge or curl, and the UDI form that was actually applied is still readable. Judge plain text as text. If the carrier is a printed 2D symbol, ISO/IEC 15415 is the print-quality method, and the protocol has to choose and justify a minimum grade. If the carrier is a direct part mark, use a direct-part-mark method such as ISO/IEC TR 29158. 21 CFR 801.45(c) does not require an AIDC grade when the direct mark is plain text only. Symbol meaning is a separate question, covered in the ISO 15223-1 symbol guide. ISO 15223-1 does not set illumination or a barcode grade. For medical electrical equipment, the Edition 3.1 legibility check in the MECA package (clause 7.1.2) uses a stated observer and ambient luminance of 100 lx to 1,500 lx at 1 m. That check belongs to IEC 60601-1. It is not a universal UDI rule.

flowchart TD
  A["Validated IFU processes"] --> B["Write the claim: mark, substrate, cycles"]
  B --> C["Copy agents, concentration, method, and count"]
  C --> D["Prepare production-substrate samples"]
  D --> E["Run the named process"]
  E --> F{"Criteria set before the test"}
  F --> G["Still attached: no loose edge or curl"]
  F --> H["Still readable in the form that was marked"]
  G --> I["Design and development file"]
  H --> I

ISO/IEC 15415 grades parameters of a printed 2D symbol, including symbol contrast, modulation, axial non-uniformity, grid non-uniformity, and unused error correction. The overall grade is only relevant if an AIDC carrier was applied, and only against the minimum the protocol justified. A production print target is not a reprocessing pass/fail rule in 21 CFR 801.45 or MDR Annex VI, and a grade does not by itself prove a hospital scanner will read the symbol at end of life. Where the direct mark feeds device identification records, see how to verify a UDI in AccessGUDID and global UDI timing.

When a Label Is the Wrong Answer: Direct Part Marking and Exceptions

Steam sterilization, an alkaline washer-disinfector cycle, or a long liquid soak may be more than an adhesive label can be shown to survive. That is a device-specific judgment. It is not a rule that every reusable device must abandon the label. When the durability claim cannot be shown for the adhesive construction, change the marking method or use an exception that the file actually supports.

FDA's November 2017 guidance does not prescribe a marking technology. The examples it gives are etching, a permanent plaque on durable equipment, a permanent tag such as RFID, and, in some cases, a sticker or other item that is designed to last the expected service life and that bears the UDI as 21 CFR 801.45(b) and (c) require. Laser annealing, dot peening, and electrochemical etching are industry methods for making a mark. None of them is exempt from the service-life check against the IFU processes, and none of them is evidence of durability until that check is done. A mark that would interfere with safety or effectiveness is a reason to use the 21 CFR 801.45(d)(1) exception, not a reason to mark anyway.

The exception paths are specific. Using one of them does not remove a label UDI that 21 CFR 801.20 still requires.

• 21 CFR 801.45(d)(1). Any type of direct marking would interfere with the safety or effectiveness of the device. The 2017 guidance expects the rationale to be scientifically justified by analysis, testing, or both. A file might, for example, show that every available method creates a crevice the cleaning process cannot address, or that it changes an optical surface. The regulation requires that basis. It does not require a particular example.

• 21 CFR 801.45(d)(2), (d)(3), and (d)(4). Direct marking is not technologically feasible; or the device is a single-use device subjected to additional processing for one further single use; or the device has already been marked under 21 CFR 801.45(a). The guidance says (d)(1) and (d)(2) can lose their basis as marking technology changes, and that labelers should reassess them. The same guidance reads (d)(3) as applying to a device the original labeler intended for one use, or for one patient during one procedure, even if someone later reprocesses it for another single use.

• MDR Annex VI Part C Section 4.10. The direct-marking requirement does not apply where any type of direct marking would interfere with safety or performance, or where direct marking is not technologically feasible. Those are the two exceptions stated on the Commission UDI helpdesk page.

Under the current text of 21 CFR 801.45(e), amended December 4, 2025 (90 FR 55979), a labeler that uses a 21 CFR 801.45(d) exception must document the basis in the design and development files required by 21 CFR 820.10(c). The November 2017 guidance still tells labelers to put that basis in the design history file under the former 21 CFR 820.30(j). The regulation's current citation controls. The guidance also says to indicate in GUDID that the device is subject to direct marking but excepted. A separate route, 21 CFR 801.55, is a request to FDA for an alternative. It is not a substitute for a 21 CFR 801.45(d) exception that already applies.

Records: Design Files, Risk File, and Submission Evidence

Reviewers look for a trace from the claim to the design and development records and the risk management file. Label content, UDI structure, and translation control sit in the labeling and UDI guide and the label translation and symbol-release guide. Those guides do not set the cleaning exposure for this test.

Keep these records with the design and technical documentation:

• Protocol and report. Record the substrate, the lot, the IFU process that was run, the cycle count and why that count was chosen, the acceptance criteria set before the test, and the before-and-after condition of the mark. Include an AIDC grade only when an AIDC carrier was part of the claim.

• Risk management file. Where loss of the identifier or of a safety marking could lead to misuse, that hazardous situation belongs in the ISO 14971 risk management file. The durability result is one risk control. FMEA is one way to record the analysis. ISO 14971 does not require FMEA by name.

• GSPR cross-reference. For the EU file, show where GSPR 11.2 applies and how the design facilitates reprocessing, and show that the durability protocol used the processes named under GSPR 23.4(n). Do not describe cleaning and disinfection in 23.4(n) as if the clause called every one of them a validated method. The validated-method wording in that point is attached to re-sterilisation, where appropriate.

• Supplier data. Facestock, adhesive, surface, and any UL 969 recognition for that surface can support incoming consistency. The recognition file is not the finished-device durability report.

Limits and Engineering Decisions You Still Own

While the configuration-to-claim-to-evidence methodology provides an audit-ready framework, design engineering teams must manage several unresolved operational variables that cannot be solved by a generic testing template:

• Cycle compression. Running the full claimed cycle count in a much shorter laboratory schedule can change how a pressure-sensitive adhesive responds, because clinical use usually includes drying time between cases. If the protocol compresses the schedule, it has to say what that compression does and does not represent. There is no default compressed protocol in the sources for this article.

• Agents outside the IFU. A facility may use a disinfectant the instructions do not list. That use is outside the validated claim. The risk file should say how field reports of label damage will be handled, and what change to the IFU list or the label material would follow. This article does not estimate how often off-IFU products are used.

• Edition control. ISO 17664-1:2021 was in systematic review, stage 90.20, as of July 15, 2026. UL 969 Edition 6 carries a July 28, 2025 revision. ASTM lists a 2025 reapproval of D3330. 21 CFR 801.45(e) was amended on December 4, 2025, while the direct-marking guidance issued November 17, 2017 still cites the former design-history section. The guidance has not been treated here as withdrawn. Cite the edition and the regulatory text you actually used.

The decision the file has to support is specific: which mark, on which substrate, after which IFU processes, for how many cycles, with the acceptance criteria written first. 21 CFR 801.45 and MDR Annex VI Part C Section 4.10 fix the outcome only inside the triggers and exceptions above. Peel force, a UL 969 recognition, and the IEC 60601-1 rub are tools with smaller jobs.