Engineering Answers / Compliance & Documents
"ISO 10993 certified" is not a thing a material can be. The standard is a family, and the correct test set depends on where the part touches the body and for how long.
ISO 10993 is a family of standards for biological evaluation, not a single pass/fail certificate. The test set is selected by contact category (surface, external communicating, implant) and duration (under 24 hours, 24 hours to 30 days, over 30 days). For a typical silicone part with intact skin or mucosal contact, that usually means cytotoxicity, irritation and sensitization; extractables and systemic toxicity come into play for longer or more invasive contact. Testing has to be done on the finished, cleaned part — not on a material sample from a different process.
"We need ISO 10993 certified silicone."
There is no such certificate. ISO 10993 is a family of standards — Part 1 frames the evaluation, Part 5 covers cytotoxicity, Part 10 covers irritation and sensitization, Part 11 covers systemic toxicity, Part 18 covers chemical characterisation, and so on. What a device programme holds is a biological evaluation, built from the parts of the family that apply to that device.
The evaluation is selected by two variables: where the device contacts the body, and for how long.
| Contact category | Duration | What is typically in scope for a silicone part |
|---|---|---|
| Surface device, intact skin | Under 24 h | Cytotoxicity; irritation and sensitization |
| Surface device, intact skin | 24 h – 30 days | Cytotoxicity; irritation and sensitization; sometimes systemic toxicity |
| Surface device, mucosal membrane | Any duration | Cytotoxicity; irritation and sensitization; extractables characterisation |
| External communicating, indirect | Under 24 h | Cytotoxicity; irritation and sensitization; haemolysis if blood contact |
| External communicating, direct | Over 30 days | Broader set including systemic toxicity and chemical characterisation |
| Implant | Over 30 days | Full evaluation, including local tissue response and degradation |
A CPAP mask cushion, a wearable seal, a breast pump diaphragm and a surgical drape component all sit in different rows. That is why quoting "ISO 10993" as a single line item on a purchase order is meaningless — the standard has to be quoted with the part number, the contact category and the duration.
Silicone is a good starting point for biocompatibility because platinum-cure addition systems do not use peroxide by-products, so there is no post-treatment to remove decomposition residues. But "platinum-cure" is not automatically extractable-free. The relevant leachables come from:
This is why the test has to be run on the finished part after the actual cleaning and sterilisation sequence. A material sample from a plaque pressed in a lab tells you about the compound; it does not tell you what the moulded, deflashed, washed and sterilised part releases.
A test report on its own is not a medical device supply chain. What a device programme normally needs alongside it:
Plus, at batch level: material lot traceability, a certificate of conformity, first article inspection for the tool, and a change notification agreement — because a silent change of compound or post-cure cycle invalidates the evaluation that the device was cleared on.
Testing the wrong article. A report on a slab of the same compound is not a report on the part. Surface area to volume ratio, mould release, deflashing method and cleaning all change the extractables.
Changing the colour. A masterbatch is a formulation addition. A part that passed in natural and ships in pantone 285 is not the part that was tested.
Skipping the sterilisation step. Ethylene oxide residuals are a separate test with its own limits, and they are a function of the sterilisation cycle, not the material. A biocompatibility report that assumes sterilisation without testing the residuals is incomplete.
Ask which process route, which post-cure cycle and which cleaning sequence the samples came from — and whether the supplier will notify you before any of those change. If the answer is vague, the test report will describe a part that is not quite the one you will buy.
ISO 10993 is a family, not a certificate. Decide the contact category and duration first; that selects the test set. Test the finished part, after cleaning and sterilisation, in the colour and process route you will actually buy. Keep USP Class VI as supporting evidence if your programme asks for it, and pair the report with an ISO 13485 quality system — because the report describes one batch, and the system is what makes the next thousand batches the same.
No, and the difference matters commercially. USP Class VI is a biological reactivity test originally developed for plastics and widely used as a baseline for implantable and long-term contact materials. ISO 10993 is a family of standards for biological evaluation of medical devices, with tests selected according to the nature and duration of body contact. Many device programmes require ISO 10993 evaluation; USP Class VI is often requested alongside it as supporting evidence, not as a substitute.
No. Running the full matrix on a part that only touches intact skin for a few hours wastes money and does not improve the submission. The matrix is supposed to be justified by contact category and duration — a surface device in contact for under 24 hours needs a much smaller set than an implant. Reviewers expect the justification, not the maximum spend.
Sometimes, and only if the material, the formulation (including colour masterbatch), the process route, the cleaning method and the sterilisation method are identical. A change from compression moulding to LSR injection, a different colour, or a new post-cure cycle can change the extractables profile, which is exactly what the test is measuring. Reusing a report across a process change is the fastest way to have a submission questioned.
ISO 10993 answers "is this material safe in this contact". ISO 13485 answers "can this supplier make it the same way every time, and prove it". Both are normally required: the test report describes the part, the quality system describes the process control, traceability and change notification behind it. A supplier with a test report but no quality system cannot support a device programme.