Substantial Equivalence, Modified and Novel Device Classification in South Korea

1. Key takeaways

Korea sorts notified devices into three categories by their relationship to devices already on the market: substantially equivalent devices, modified devices, and novel devices. The determination method is a fixed six-element comparison covering intended use, mechanism of action, raw materials, performance, test specifications, and instructions for use. If all six elements match the comparator device, the product can be classified as substantially equivalent; if individual elements differ but can be argued with existing data, it falls into the modified device category; if a fundamental difference appears at the level of intended use or mechanism of action, it is a novel device.

The classification directly determines the documentation requirements: technical documents for substantially equivalent devices can be organized as comparisons, modified devices must add argumentation for the differing elements, and novel devices must supply complete safety and effectiveness evidence. Class 2 devices that have already been approved, certified, or notified in Korea 3 or more times may be publicly identified as recognized substantially equivalent devices and cited directly in later comparisons.

GRZAN’s view is that the comparison table matters more than the conclusion. What the reviewing body sees is not the sentence claiming equivalence, but six lines of item-by-item evidence.

2. Legal basis

The three-category framework rests on the following instruments.

Six-element comparison framework for the three device categories (per MFDS Public Notice No. 2022-52)

3. Expert insights

The six elements are parallel; there is no weighting order. No hierarchy ranks the first three elements above the rest. GRZAN has seen products whose intended use, mechanism of action, and raw materials were all identical, differing only in the test specification settings, being required to supply complete specification argumentation as modified devices. The reviewer’s logic: if any single element differs, the equivalence chain breaks at that point and needs separate argumentation. Build the comparison table flat across all six elements, without ranking marks inside it.

The real threshold for modified devices is whether each difference can be closed with argumentation. Being classified as a modified device is not a lowering of requirements; it concentrates the argumentation on the differing elements. If the difference lies in raw materials, material safety and biological evaluation data are needed; if in test specifications, the basis for the specification settings and validation data are needed. GRZAN’s experience is that when differences exceed 2 elements (practical estimate), the workload of arguing each one often approaches that of a novel device, and it is better to prepare to the higher standard in one pass than to insist on the modified device position.

The choice of comparator device decides the persuasive power of the comparison. The comparator must be a device already approved, certified, or notified in Korea. Choose first from devices under the same item and the same classification rules as the notified device, and only then consider cross-item comparisons. Even when all six elements look identical on the surface, cross-item comparisons are often pressed on the wording of intended use. Class 2 devices approved, certified, or notified in Korea 3 or more times may be publicly identified as recognized substantially equivalent devices, and citing them markedly improves comparison efficiency.

A novel device determination should be acknowledged early. A product with a fundamental difference at the level of mechanism of action will be overturned during review if its documents were compiled on the substantial equivalence route, and the upfront investment cannot be recovered. GRZAN recommends searching before documentation begins for whether any Korean precedent already exists at the mechanism level; if none does, resource the project as a novel device, including complete safety and effectiveness evidence and any necessary clinical trial arrangements (Article 10 of the Medical Device Act).

4. Practical recommendations

• Build the comparison table flat across the six elements, with no ranking marks. Give each element an identical or different conclusion, and attach the document reference number for each differing element.
• Prefer comparators under the same item and the same classification rules. Cross-item comparisons need an extra explanation of the comparability of intended use.
• Close each differing element on its own. Give every difference its own set of argumentation documents; do not gloss over it in the body text with a summary sentence.
• Prepare to a higher standard when differences are numerous. When differences exceed 2 elements (practical estimate), resource the project at the novel device level.
• Move the novel device determination forward. Complete the mechanism-level precedent search before documentation starts, to avoid a pathway-level rework.
• Register recognized substantially equivalent device status separately. For Class 2 devices that have met the 3 or more times approval, certification, or notification threshold, keep a register so they can be cited later.

5. Verification checklist

• □ Comparator qualification — verify its approval, certification, or notification status in Korea; standard: lawful market authorization obtained and the number traceable.
• □ Completeness of the six-element comparison — build the table element by element; standard: intended use, mechanism of action, raw materials, performance, test specifications, and instructions for use with no element missing.
• □ Identification of differing elements — checked against the six conclusions; standard: a clear list of differences with no hidden ones.
• □ Argumentation for differing elements — verify the documents behind each difference; standard: every difference has its own independent argumentation documents.
• □ Classification conclusion — judge substantially equivalent, modified, or novel from the six results; standard: the classification matches the comparison table.
• □ Documentation requirements match — check the document list against the classification; standard: simplified for substantial equivalence, difference supplements for modified, full allocation for novel.
• □ Recognized substantially equivalent device — verify the number of grants in Korea; standard: a Class 2 device granted 3 or more times may be publicly cited.

How GRZAN can help. GRZAN provides medical device and IVD registration, quality system, Korean licence holder and local testing support for the Republic of Korea. Our team manages MFDS approvals, certifications and notifications, prepares equivalence dossiers, and supports post-market vigilance obligations. Business enquiries: info@grzanglobal.com