（一）Rule 1 → Class D (Highest Risk)Three scenarios directly result in Class D classification:- Screening of blood, cells, tissues or organ donations to assess suitability for transfusion or transplantation, for detection of transmissible pathogens;
- Detection of highly transmissible, life-threatening pathogens (HIV, HBV, HCV, Ebola, SARS‑CoV‑2, etc.);
- Viral load monitoring for life-threatening infectious diseases where results directly guide patient treatment decisions.
Common pitfall：The same pathogen assay may not be Class D if intended for routine patient diagnosis rather than blood donor screening.（二）Rule 2 → Class C / D (Blood Grouping and Tissue Typing)- ABO, Rh(D), Kell, Kidd and Duffy critical blood group typing: Class D
- Other blood group typing and HLA tissue typing: Class C
- Covers testing of donor and recipient samples, including fetal RhD genotyping kits.
（三）Rule 3 → Class C (Large Class C Category, Widely Used by Chinese Manufacturers)Products falling under any subparagraph (a)‑(m) and not covered by Rule 1 or Rule 2 are Class C. This category includes many mainstream IVD products:- (a) Sexually transmitted pathogen detection (Chlamydia, Gonococcus, HPV, Syphilis, etc.)
- (b) Direct detection of pathogens in cerebrospinal fluid or blood that are not highly transmissible
- (c) Pathogen assays where incorrect results may cause death or severe disability to the test subject, fetus or offspring
- (d) Prenatal screening of infectious pathogen immune status in pregnant women
- (e) Infectious or immune status testing where false results may lead to life-critical clinical decisions
- (f) Companion Diagnostics (CDx) (Highlight!)
- (g) Disease staging where false results carry life-threatening risks (tumour staging excluded, covered under (h))
- (h) Screening, diagnosis and staging of neoplastic diseases (including pre-malignant lesions)
- (i) Human genetic testing (carrier screening, prenatal testing, newborn screening, pharmacogenomics, DTC genetic testing)
- (j) Monitoring of drugs or biomarkers where false results may lead to fatal outcomes
- (k) Testing for management of patients with confirmed life-threatening conditions
- (l) Screening for congenital defects in embryos or fetuses
- (m) Newborn screening for congenital disorders (PKU, congenital hypothyroidism, thalassemia, etc.)
4-step official logic for CDx determination (aligned with MDCG flowchart)- Is the assay essential to determine the safety and efficacy of a medicinal product with an INN? → If no, not CDx.
- Is the assay only used to monitor drug concentration within the therapeutic window? → If yes, not CDx.
- Is the assay only used to calculate dosage for patients already eligible for treatment? → If yes, not CDx.
- Does the assay, before or during treatment, identify patients most likely to benefit OR at increased risk of serious adverse reactions? → If yes, it is CDx and Class C.
Frequent manufacturer mistake: Therapeutic drug monitoring or genetic polymorphism testing solely for dose calculation is NOT companion diagnostics.（四）Rule4：Self-Testing and POCT4(a) Self-testing devices: Default Class C. Only pregnancy, ovulation, total cholesterol, urine glucose/erythrocyte/leucocyte/bacteria assays are Class B.If a kit combines Class B and Class C analytes, the whole product is classified as the higher risk Class C. HIV self-tests are Class D.4(b) POCT does not automatically upgrade risk class. POCT reagents are classified under Rules 1‑3 based solely on their intended use; they share the same risk class as laboratory-based versions. POCT setting alone does not raise classification, a common misconception among manufacturers.General laboratory consumables, non-critical accessories, buffers, wash solutions, general culture media, sample collection containers; instruments specifically designed for IVD testing.
Note: Products fall under IVDR only if the manufacturer explicitly states intended use for in vitro diagnostic purposes. Generic laboratory equipment is outside IVDR scope.
All products not covered by Rules 1–5 are Class B. Examples: routine electrolyte assays, certain common pathogen assays (HAV antibody, rotavirus, Candida etc.), some autoantibody assays.
Incorrect results generally cannot directly cause death or severe permanent disability.
Unassigned controls (values to be determined by end users) are uniformly Class B. Assigned controls follow the risk class of the corresponding reagent.
四、Practical Compliance Tips from MDCG 2020‑16- Intended use wording is the core of classification. Statements in IFU, technical documentation and marketing materials must be fully consistent. Ambient claims may lead notified bodies to assign a higher risk class. To exclude high-risk applications, clearly state limitations in the IFU, e.g. This kit is not intended for blood donor screening.
- Do not assume bundled products inherit the host instrument’s class. Reagents, software, calibrators and controls shall be individually classified per the implementation rules.
- Software classification: Standalone IVD software is classified based on its own intended use. Instrument embedded control software follows the class of the instrument/reagent, refer to MDCG 2019‑11 software guidance.
- For CDx: The assay must be linked to a medicinal product with an INN. Full CDx justification shall be prepared, and notified bodies will seek opinions from EMA or Member State competent authorities during assessment.
- Prepare a complete Classification Justification document, citing IVDR Annex VIII and MDCG 2020‑16 rule by rule. This is the primary document reviewed by notified bodies.
Under IVDR, many reagents classified as low-risk under the old IVDD regime have been upgraded to high-risk Class C/D, bringing significantly higher compliance costs. Classification is not a simple lookup exercise; it requires rule-by-rule assessment combining clinical intended use and risk scenarios. Manufacturers are advised to complete classification determination at the early project stage, to avoid rework after product development and documentation are finalised due to wrong classification.
