Reprocessing Validation for Reusable Devices: Applying ISO 17664

A reprocessing IFU is a set of claims, not just a block of text

Teams working through their first reusable device registration often treat the reprocessing instructions as a writing exercise: find a similar product's IFU, swap the product name, add an exploded view, hand it to regulatory affairs for layout. By the time the file reaches technical review, or an overseas technical documentation assessment, the questions almost always land in the same place. Every "the device may be processed this way" sentence in the IFU is a technical claim the manufacturer makes to the user, and claims need evidence behind them.

That is exactly what the ISO 17664 series is for. It sets out what reprocessing information the manufacturer has to provide to the user and how far that information has to go. It does not decide for you how the device should be cleaned, disinfected or sterilised; what it does is put the "this must be stated" list on the table. Once a processing route is written into the IFU, the obligation to validate it comes along with it. Read the other way round: the more routes you write in, and the more "complete" the IFU looks, the more validation you have bought. This is the line item companies most often miss at budgeting time. One extra sentence saying "may also be processed in an automated washer-disinfector" usually means one more full round of cleaning validation.

A word on citation practice before we go further, so you do not lose time hunting for the wrong document. ISO 17664 exists as numbered parts, ISO 17664-1 and ISO 17664-2 among them, and the parts do not cover the same objects or the same information requirements. Where this article refers to the framework as a whole it says "the ISO 17664 series"; the moment a specific clause is at stake you have to go back to the text of the relevant part. Which part your device falls under, and the current scope and status of each part, are governed by the current valid version of the standard text. Do not search on a part-less number, and do not let one stand in as a specific basis in your technical file. The same rule applies to the GB/T 16886 series: where biological evaluation is mentioned here, the series as a whole is meant; which part specifies a given endpoint and which edition applies must be confirmed against the standard text once the evaluation route is settled.

One more common misconception is treating reprocessing validation as if it were sterilisation qualification. Sterilisation qualification is only the tail end of the chain. A device coming back from the point of use passes through point-of-use preparation, disassembly, cleaning, rinsing, disinfection, drying, inspection and functional checks, and packaging before it ever reaches sterilisation and storage. If any earlier step is weak, the qualification at the end is dragged down by its own inputs. If bioburden has not come down, no set of sterilisation parameters, however elegant, will hold up.

Decide the route first: three questions that set the scope

Answer these three questions before you ask for a quotation or a schedule, and the scope of validation is essentially fixed.

The first is how the device contacts the body. Intact skin, mucous membranes, or entry into sterile tissue or the vascular system. The deeper the contact, the higher the end-point requirement: non-critical devices usually stop at cleaning and low-level disinfection, semi-critical devices call for high-level disinfection or sterilisation, and critical devices call for terminal sterilisation. This decides whether the IFU ends in "disinfect" or "sterilise", and therefore whether process qualification to ISO 17665 or ISO 11135 has to be brought in.

The second is accessibility of the geometry. Does the device have lumens, blind holes, hinges, threads, porous surfaces, telescoping members, long flexible tubing, sealed cavities that cannot be opened? These features are where cleaning validation actually gets difficult, and they are the basis on which sampling locations are designed.

The third is who reprocesses the device, and to what level. Is it the hospital sterile processing department working from your IFU, or does the device return to the manufacturer for professional refurbishment? The first means the IFU has to be written at a granularity an ordinary operator can reproduce, and validation has to be run under conditions close to hospital reality rather than laboratory ideal. The second is easier to control but places heavier documentation demands on the processing site and on release decisions.

Which claims have to be evidenced, one by one

The table below is a working tool for the project kick-off meeting. On the left are statements that may appear in the IFU; on the right is the validation obligation each one automatically triggers. Run through it before the IFU is frozen and delete the routes you cannot afford to validate. That is far cheaper than adding the work afterwards.

All specific parameters and acceptance limits involving temperature, time, concentration and cycle count referred to in the table are governed by the current valid version of the standard text; this article does not restate them at the numerical level.

Choosing the worst-case configuration: an engineering analysis

Nobody validates cleaning on every size in the range. The mainstream approach is to group devices into families and select the hardest-to-clean representative within each group, the worst case. The rationale for that selection has to be written down; it is queried during assessment more often than almost anything else.

Reasoning from mass transfer and interfacial behaviour gets you most of the way. Flow of cleaning solution inside a lumen has a boundary layer at the wall; velocity falls as you approach the wall and mechanical scouring decays with it. The narrower and longer the lumen, the greater the driving pressure difference needed for effective scouring, and the slower the cleaning agent inside the lumen is renewed. Blind ends, side ports, thread roots and hinge gaps are flow dead zones. The bulk flow does not really enter them, residues sit there held by diffusion and capillary action, and removal depends more on soak time and chemistry than on mechanical action. Porous coatings and rough surfaces add real contact area and more retention sites.

There is a materials-side mechanism that is easy to overlook: proteins in blood and body fluids denature when heated and bind more firmly to the oxide layer on metal surfaces. That means rinsing with warm water at the point-of-use step can actually fix the soil onto the device and make later cleaning harder. In the same way, the longer a device sits after use, the more likely the soil is to dry and biofilm to form, and the harder removal becomes.

The above is an engineering analysis based on mass transfer paths and material interface behaviour. Its purpose is to help identify the hardest-to-clean configuration and the sampling locations. It is not a statistical conclusion and it does not replace measured data. The final choice still has to be confirmed the other way round by actual cleaning validation results: if the representative configuration comes out clearly cleaner than the others in its group, the grouping or the choice of representative needs to be revisited.

Simulated-use soiling and recovery: the two foundations

Plenty of failed projects fail not because the cleaning process is poor but because the validation method itself does not stand up.

The first foundation is simulated-use soiling. Running a brand new, clean device through the cleaning process, measuring residues and getting a beautiful result produces data that carries no weight in assessment. The correct approach is to use a test soil containing protein, blood components and polysaccharides, apply it to the device in the prescribed manner and particularly to those flow dead zones, then simulate the post-use hold and drying so that the soil is in an unfavourable state, and only then run the IFU process. Where the soil is applied, how, and under what drying conditions all have to be recorded, because they directly determine how demanding the validation is.

The second foundation is recovery. After cleaning, residues have to be taken off the device to be measured, and no sampling technique, whether extraction, lumen flushing, sonication or swabbing, recovers everything. Without a recovery study first, the residue figures cannot be interpreted: is a low number a clean device, or a failed extraction? Recovery data is what converts "amount detected" into "amount present". Without it, the conclusion of the entire cleaning validation report is floating free.

The methodology for those two foundations does not live in the ISO 17664 series

This is the point this article most wants to flag, and the root cause of a good many protocols being sent back at review. The ISO 17664 series answers the question "what must the manufacturer tell the user". It will not tell you what the test soil is made of, how it is applied, how dry it has to be to count as unfavourable, which residue indicators to measure or how to judge the result, and it does not cover how the performance of a washer-disinfector itself is to be qualified.

That methodology comes from a different set of documents. One family addresses washer-disinfectors, specifying equipment performance requirements, test methods and the process qualification framework. Another addresses the determination and acceptance of residue indicators, giving the measurement approach and acceptance logic for protein, total organic carbon and similar markers. This article deliberately does not list numbers for those two families, because the transposed versions adopted differ by target market, and the numbering format and current validity status have to be checked line by line with the testing laboratory when the protocol is set up. A wrong number in a technical file is more trouble than a temporarily missing one. The principle worth remembering fits in one sentence: the claims live in the ISO 17664 series, the evidence methods live elsewhere. List the two separately in your technical file, each with its own basis; do not let a single ISO 17664 series designation stand in for the methodological source of an entire cleaning validation report.

At the indicator level, practice generally combines several dimensions. Protein, haemoglobin and total organic carbon are common, with endotoxin assessment added where relevant. A single indicator is easily skewed by the process: some detergents remove protein residues well without a matching drop in overall organic load, so a combination supports the conclusion better than a single point. Bioburden after cleaning can be determined by the methods of ISO 11737-1 and used to support the input conditions for subsequent sterilisation process qualification. This step is easy to squeeze in the schedule, but it is precisely the interface between the cleaning evidence and the sterilisation evidence; squeezing it usually means rework at the sterilisation qualification stage.

The acceptance approach has to be fixed at protocol stage, not chosen once the data is in. Three acceptance logics are common: compare against an uncleaned positive control and look at the order of magnitude removed; compare against a clean negative control and see whether the device returns to background; or compare against pre-defined acceptance criteria. Whichever you use, state where the criteria come from and why, and state the detection capability of the method. If the limit of detection sits above the acceptance criterion, the data set does not hold together logically. Recovery correction has to be shown explicitly in the report: state the sampling method, how the recovery study was run, what came out of it, and by what relationship the final residue figure was derived from the detected amount. What assessors care about is usually not whether the numbers look good, but whether that conversion chain is unbroken.

If the IFU claims processing in an automated washer-disinfector, there is one more layer that is easy to miss. Qualifying the equipment and qualifying your device's process on that equipment are two different things. The first shows that the machine runs its programmes as intended and that its performance is reproducible; it rests on equipment-side standards and is usually provided by the equipment manufacturer. The second has to show that your device, on this machine, with this programme, this loading and this set of connectors, actually gets clean. That is the device manufacturer's own obligation, and nobody can discharge it for you. The classic rejection is a submission containing the equipment vendor's performance report and no validation at all of the company's own loading arrangement and lumen connections.

Material tolerance and functional retention: the section that gets skipped

If the IFU states a maximum number of reprocessing cycles, there has to be data after repeated cycles to support it. And the acceptance criterion here is function, not appearance.

What usually needs looking at: whether clearances and operating forces of moving parts have drifted; whether seals have aged to the point of leaking; whether coatings have flaked or blistered; whether markings and scales are still legible; whether cutting or gripping surfaces have dulled; whether threads and snap fits have loosened after repeated assembly; and, for devices with electrical parts, whether insulation and liquid ingress protection claims still hold.

A recurring trap in real projects is that the company runs appearance and dimensional checks and never defines functional acceptance criteria. Asked at review "you say it still works, against what criterion?", nobody has an answer. Write functional criteria into the validation protocol in measurable, reproducible form and state where they come from. Company-defined criteria are perfectly acceptable, provided the rationale is sound and the method repeatable. Repeated processing cycles can also change the extraction behaviour of materials; if the device materials or surface treatment change over the cycles, the risk analysis has to address whether the existing conclusions of the biocompatibility evaluation still apply, with supplementary assessment per the relevant part of the GB/T 16886 series where needed. Which part specifies what, and which edition applies, is confirmed against the current valid version of the standard text and the chosen evaluation route.

Interfaces with sterilisation, packaging and biological evaluation

Reprocessing validation is not an isolated project; it has to connect upstream and downstream. The table below sets out what each standard contributes to the chain, and where projects commonly misuse it.

The interface with sterilisation validation in particular needs planning ahead. One of the required inputs to sterilisation qualification is post-cleaning bioburden, so scheduling sterilisation qualification before cleaning validation has produced results is a reliable way to end up repeating it. On packaging validation: even though the hospital does the packaging for a reusable device, the manufacturer still has to state the recommended class of sterilisation container or packaging material and its suitability, and remains responsible for the packaging description in the IFU.

One boundary is worth drawing. This article deals only with the reprocessing chain. If your product is a surgical instrument supplied sterile, the requirements on the sterile-supply side, meaning primary packaging form, sterile barrier and accompanying information, are a separate topic. The starting point for that judgement is whether the product is supplied sterile or non-sterile and whether reprocessing instructions are supplied with it; evidence from the two chains is not interchangeable.

What to prepare before submitting for testing

Experience says the more complete the up-front material, the fewer rounds of protocol review. We suggest having the following ready before you place the order; the detailed list can also be checked against the submission requirements.

• The draft reprocessing IFU, with the complete sequence of steps for every route, including the classes of detergent and disinfectant intended. State the class and formulation type; brand names are neither necessary nor advisable.
• The device family grouping rationale and the justification for the hardest-to-clean configuration, including the structural comparison it rests on.
• Disassembly and reassembly instructions, the special tool list, and the numbering scheme for small parts.
• Samples covering the selected representative configurations, in quantities that leave headroom for repeat validation and retesting, plus sacrificial samples for repeated processing cycles and destructive examination.
• Accessories, and the recommended sterilisation container or packaging material.
• The bill of materials and a description of surface treatment processes, including coatings, platings and adhesives.
• The intended combination of residue indicators, the acceptance logic and the source of the criteria. If you already have a preferred sampling method in mind, say so, so the recovery study can be arranged in parallel.
• If automated washer-disinfector processing is claimed, the programme class of the intended equipment, the loading scheme, the lumen connectors, and any existing equipment-side performance documentation.
• Functional acceptance criteria and how they are measured, with the rationale for any company-defined criteria.
• Existing risk analysis documents, so validation items and risk control measures can be lined up against each other.

Common reasons for rejection, and how to fix them

Grouping the issues we have seen over the past few years, the recurring ones are roughly these: two cleaning routes in the IFU and only one validated; validation run on new samples with no simulated-use soiling; no recovery data, so residue results cannot be interpreted; the source of the acceptance criteria unclear, or the detection capability of the method mismatched with the criteria; the entire methodological basis hung on the ISO 17664 series, with no traceable source for the test soil or the residue indicators anywhere in the report; device families that force structurally very different products into one group; drying not validated at all, leaving residual moisture inside the pack and affecting sterilisation; a reprocessing cycle limit with no cycle data behind it; disinfection claims disconnected from material compatibility data; sterilisation qualification loads that do not match what hospitals actually do.

The remedy is always the same shape. Go back to the IFU, mark up every claim sentence by sentence, and attach evidence to each one. Where you cannot, either add the validation or change the IFU. Most projects find that once the routes nobody uses and nobody can validate are deleted, total cost goes down and the IFU becomes more executable.

What SUNGO Lab can do

SUNGO Lab is a third-party medical device testing laboratory accredited by CNAS, CMA and IAS (USA), with laboratories in Shanghai and Hefei. To be explicit: accreditation marks only demonstrate that the laboratory holds the corresponding technical competence within its accredited scope; they are not a commitment as to market access outcomes in any target market.

Around reprocessing validation for reusable devices, we can support IFU claim mapping and validation protocol design, device family grouping and worst-case configuration assessment, simulated-use soiling and recovery studies, confirmation of the residue indicator combination and acceptance criteria, cleaning and disinfection effectiveness validation, bioburden determination per ISO 11737-1, sterilisation process qualification linked to ISO 17665 and ISO 11135, and functional and material assessment after repeated processing cycles. Where specific parts and methodological references are involved, we list them line by line against the current valid versions at protocol confirmation, so that the technical file does not end up citing a superseded edition. Project schedules and test combinations can be confirmed item by item against device geometry and target market. Our wider scope is set out under testing services.

To work out which reprocessing route applies to your device and which validation items you need, call our engineers on +86 132 4819 8029 or request a quote.