510(k) Biocompatibility Requirements: ISO 10993 for Medical Devices

Biocompatibility assessment is required for any medical device that contacts the patient — directly or indirectly. FDA's expectation is that biocompatibility be evaluated using the ISO 10993 series of standards. Biocompatibility gaps are among the most common reasons 510(k)s receive AI requests, and they are almost entirely avoidable with thorough preparation.

Does your device need biocompatibility assessment?

Any device that contacts the patient — including indirect contact through circulating blood or bodily fluids — requires biocompatibility evaluation. This includes devices that contact intact skin, mucosal membranes, breached skin, blood, tissue, bone, or dentin. External communicating devices that contact blood or body fluids externally also require evaluation.

Non-contacting devices (e.g. imaging equipment that doesn't touch the patient, software-only devices) do not require biocompatibility assessment.

The ISO 10993-1 risk-based approach

ISO 10993-1 (updated in 2018) requires a risk-based approach to biocompatibility. Rather than automatically running every test in the matrix, you must first perform a biological evaluation — a documented assessment of your device's materials, manufacturing processes, and contact characteristics — and use that assessment to determine which tests are needed, which can be justified by existing data, and which can be waived.

FDA's guidance on the use of ISO 10993-1 (issued 2016, updated 2023) aligns with this approach and specifically states that manufacturers should not default to running all tests in the matrix. Over-testing without a documented rationale is as problematic as under-testing.

When literature justification is acceptable

If your device uses materials with an established biocompatibility history — such as materials that have been used in many cleared predicate devices — you may be able to justify some endpoints with literature or existing data rather than new testing. This is particularly common for:

• Standard medical-grade polymers (silicone, polyurethane, PEEK) with extensive published biocompatibility data
• Metals with established biocompatibility records (titanium, stainless steel, cobalt-chrome alloys)
• Materials that have been used in cleared predicate devices with the same contact type and duration

Literature justification must be documented and specific — citing a general reference that the material is "commonly used" is not sufficient. The reference must cover the relevant endpoints for your contact type and duration.

Common biocompatibility AI triggers

• Missing endpoints — failing to address one or more endpoints required for your contact type and duration
• Wrong contact duration — testing for limited contact when your device is actually prolonged or permanent use
• Processing residuals not addressed — manufacturing aids (mold release agents, lubricants, cleaning agents) that may leave residuals must be addressed even if the base material is biocompatible
• Colorants and additives not evaluated — colorants, stabilizers, and other additives in plastics require separate biocompatibility consideration
• New materials not tested — citing a predicate device's biocompatibility when your device uses different or additional materials
• Outdated test standards — testing to superseded versions of ISO 10993 part standards