Skip to main content

IVD Devices

IVD devices in South Korea are regulated under the In Vitro Diagnostics Act (in vitro 진단용 의료기기법) with a separate Grade A–D classification system based on the risk and severity of an incorrect test result to patient health. Key IVD-specific requirements:

Formal performance evaluation (not clinical evaluation) — analytical validation (sensitivity, specificity, accuracy) and clinical validity in the intended patient population Reference laboratories — for Grade C–D devices, results may require confirmation by MFDS-designated reference laboratories Companion diagnostics coordination with drug approval — IVDs used to select patients for specific drugs must align with drug approval timelines

  • Formal performance evaluation (not clinical evaluation)
  • Reference laboratories for certain confirmatory testing
  • Companion diagnostics coordination with drug approval

IVD classification system ​ IVDs are classified A–D based on the risk of producing incorrect results: ​ • Grade A: minimal risk (general screening tests) • Grade B: low-to-moderate risk (most routine laboratory tests) • Grade C: moderate-to-high risk (confirmatory tests, pregnancy tests, blood grouping) • Grade D: highest risk (blood screening, HIV/hepatitis testing) ​ Higher grades require more robust performance evaluation and may require reference laboratory validation.

Clinical validity evaluation

Clinical validity differs from analytical validation. While analytical validation demonstrates the test's ability to detect the target analyte (sensitivity, specificity, accuracy), clinical validity establishes that the test result correctly predicts or diagnoses the clinical condition in the intended patient population. Clinical validity requires evidence that the test result correlates with the disease state, is relevant to patient management decisions, and demonstrates appropriate positive predictive value and negative predictive value in relevant patient cohorts. For companion diagnostics (Grade C–D), clinical validity data must align with approved drug labelling and clinical trial outcomes.