Biocompatibility and ISO 10993: Biological Evaluation of Medical Devices Explained

Every device that touches the body, directly or indirectly, must be shown to be biologically safe. That evidence is built through a biological evaluation under ISO 10993-1, not by running a standard set of tests and hoping for a pass. This guide explains how the process works, how contact category and duration decide the endpoints you must address, and how to write a plan and report that regulators accept.

The principle: a risk-based evaluation, not a test checklist

ISO 10993-1 treats biological evaluation as part of risk management. You start with what the device is made of, how it contacts the body and for how long. You then collect existing knowledge, chemical characterisation and prior testing, and only test what is still needed to close gaps. Regulators in the EU and US increasingly expect this reasoning to be documented, and many deficiency letters on biocompatibility say the rationale was missing rather than the test being failed.

Step 1: classify the contact

The standard classifies devices by the nature and duration of body contact. These two factors decide which biological endpoints you should consider.

Category	Examples
Surface device	Skin electrodes, bandages, mucosal contact such as a dental device
External communicating device	Blood-path devices such as IV sets and catheters, tissue or bone contact devices, circulating-blood contact
Implant device	Bone screws, stents, pacemaker leads, tissue implants

Contact duration is grouped as limited (up to 24 hours), prolonged (more than 24 hours up to 30 days) and long-term (more than 30 days). For repeated or multiple exposures, you consider the cumulative contact time.

Step 2: identify the endpoints

Depending on category and duration, the endpoints to address commonly include:

• Cytotoxicity (ISO 10993-5), a basic screen for cell damage.
• Sensitisation (ISO 10993-10), the potential for an allergic reaction.
• Irritation or intracutaneous reactivity, now covered by ISO 10993-23.
• Acute systemic toxicity and material-mediated pyrogenicity.
• Subacute, subchronic and chronic toxicity for longer contact.
• Genotoxicity, carcinogenicity and reproductive toxicity where exposure and chemistry warrant.
• Implantation effects for implants.
• Haemocompatibility for blood-contacting devices.

The list is a starting framework. You justify omitting an endpoint with evidence, for example that the material has a long history of safe use in the same application.

Step 3: chemical characterisation and toxicological risk assessment

Modern practice leans heavily on chemistry. Under ISO 10993-18, you identify and quantify what can leach or be extracted from the device, using extraction conditions that exaggerate or simulate clinical use. Under ISO 10993-17, a toxicologist then compares the exposure to tolerable limits. When chemical data and toxicological assessment show an acceptable margin, they can replace some animal testing. A strong chemical characterisation also protects you when a supplier changes a resin or additive. See supplier controls.

Step 4: write the biological evaluation plan and report

A biological evaluation plan (BEP) sets out the device description, materials, manufacturing processes and residues, contact category, endpoints, available data, gaps and the strategy to close them. The biological evaluation report (BER) records the results and a justified conclusion on whether the biological risk is acceptable. Both should be written or reviewed by competent, qualified people, and both live in the design history.

Practical points that decide success

• Test the finished, sterilised device. The final product, including coatings, adhesives, sterilisation residuals and ageing effects, is what the patient meets. Where sterilisation uses ethylene oxide, residuals need controlling to ISO 10993-7.
• Use accredited laboratories. For FDA submissions, nonclinical laboratory studies are expected to follow Good Laboratory Practice (21 CFR Part 58) or have a stated justification.
• Manage change. A new material, supplier, cleaning agent or process change can invalidate prior data. Link process validation and change control to a biocompatibility impact check.
• Align with the clinical evaluation. Under the EU MDR, the biological safety conclusion feeds the benefit-risk analysis in the clinical evaluation.
• Watch for special substances. The EU MDR has specific requirements for CMR substances and endocrine disruptors above set concentration thresholds, and for justifying them in the device.

FDA and EU context

FDA has published a guidance on using ISO 10993-1, which sets out its expectations including how to present a biocompatibility risk assessment. Under the EU MDR, biological safety is addressed in Annex I and documented in the technical documentation. See also our overviews of US FDA and EU requirements. Always check the current edition of ISO 10993-1 and FDA’s guidance before finalising a plan, as both are revised from time to time.

Common mistakes

• Pointing to a supplier’s “USP Class VI” or “ISO 10993 compliant” statement as complete evidence. It rarely covers your finished device, process and sterilisation.
• Running all possible tests without a rationale, then struggling to explain results.
• Skipping a plan and writing the report after the testing.
• Not reassessing after a material or process change.
• Forgetting that reusable devices need an evaluation after the cleaning and disinfection cycles they will see.

Frequently asked questions

Can I avoid animal testing?

Often partly. Chemical characterisation, toxicological risk assessment and existing data can reduce or replace some tests, and the standard encourages that approach. You must justify each omitted endpoint.

Do I need to repeat testing when I change suppliers?

Not always, but you must assess it. A change to a formulation, additive, process aid or sterilisation method may need new chemical or biological data. Document the assessment either way.

What is the difference between a BEP and a BER?

The plan says what you will evaluate and why, before testing. The report presents the data and the conclusion afterwards.

Is biocompatibility needed for devices that touch only intact skin?

Yes, though the endpoints are fewer. Cytotoxicity, sensitisation and irritation are typical for limited contact with intact skin, subject to your rationale.

Talk to the AptSkill team

Need help planning a biological evaluation or training your team on ISO 10993? Send us a message.