Performance Evaluation (IVD)
Performance evaluation for IVDs in Japan
For IVD devices, the equivalent of the clinical evaluation report is the performance evaluation — a structured assessment of the IVD's analytical and clinical performance. This is required for IVD Shonin applications and for IVD Ninsho applications where performance testing is specified by the applicable recognised standard.
Analytical performance studies
Analytical performance studies assess the technical capability of the IVD system:
- Accuracy — comparison against a reference method or reference material
- Precision — repeatability (within-run) and reproducibility (between-run, between-day, between-site)
- Detection limit — limit of detection and limit of quantitation
- Measuring range and linearity
- Interferences — effect of potentially interfering substances (haemolysis, lipaemia, bilirubinaemia, common drugs)
- Specificity and cross-reactivity — for immunoassays and molecular tests
The specific analytical performance parameters required depend on the type of IVD and the applicable PMDA/MHLW guidance.
Clinical performance studies
Clinical performance studies assess how the IVD performs in clinical practice:
- Diagnostic sensitivity and specificity — using a well-characterised patient population
- Positive and negative predictive values — in the intended use population
- Reference range — in the Japanese healthy population (important — reference ranges established in Western populations may not apply to Japanese patients)
- Specimen type validation — particularly if multiple specimen types are claimed
Reference materials and standards
Japan uses both JCCLS (Japan Society of Clinical Chemistry) standards and international reference materials. For certain analytes, JCCLS has established Japanese-specific traceability chains that PMDA expects to be used.
Post-market performance follow-up (PMPF equivalent)
For IVDs approved in Japan, ongoing post-market performance monitoring is expected as part of the GPSP programme. This includes tracking of quality control data, proficiency testing results, and any post-market performance issues identified by users.