GuidelinesGMP2026.10.02
Planning Medical-Device Process Validation Before IQ, OQ and PQ
Separate processes adequately verified by inspection from those needing validation, then build evidence for installation, operating limits and actual production variation.
Identify what inspection may miss
The first question is not whether the company needs reports called IQ, OQ and PQ. Examine whether subsequent inspection can adequately verify each output, whether it is destructive and whether hidden defects can be detected. Place quality characteristics and inspection methods beside the process map.
A bond may look acceptable while internal bonding or long-term retention cannot be checked on every unit. Other outputs may be adequately verified by measurement. GHTF guidance provides process-validation concepts; establish Korean obligations from the applicable KGMP and product requirements.
Write acceptance and failure handling into the plan
Define the process, product family, site, equipment, materials and activities, with reasons for exclusions. Connect important quality characteristics, influential process variables, acceptance criteria and measurement methods. Following a work instruction alone does not demonstrate consistent output.
Approve the plan before execution. Assign decisions on sample segregation, investigation, repeat testing and plan changes if limits are exceeded. Retrofitting acceptance criteria to results removes the original question. Align process-risk priorities with the risk-management evidence.
Give IQ, OQ and PQ different questions
This is a practical distinction between the stages. Evidence appropriate to the product and process matters more than document titles.
| Stage | Question | Example records |
|---|---|---|
| IQ: installation qualification | Were equipment and supporting conditions installed as intended? | Asset identification, installation, utilities, calibration |
| OQ: operational qualification | Are outputs acceptable across the defined operating range? | Upper/lower settings, alarms, challenge conditions and results |
| PQ: performance qualification | Is consistency maintained in actual production? | Operator, lot, shift and time variation with quality results |
Supplier installation records may support IQ, but do not automatically demonstrate suitability using the actual materials and product. Reconcile the OQ operating range with production instructions.
Explain sampling and worst-case choices
Do not select batches and samples solely because “three runs is customary.” Identify the variability to include, confidence sought, destructive testing and product-family representation. Quantitative measurements and pass/fail attribute data may need different analysis methods.
The largest product is not always worst case. Small bonding area, thick material, heat sensitivity or complex loading may challenge different quality characteristics. Document why a representative model was chosen and the similarities and differences of excluded models. Do not extend results beyond the evidence.
Preserve deviations and their disposition
Identify affected tests and samples when criteria fail, equipment stops, samples are missing or measurement errors occur. Preserve original data and connect investigation, actions, additional work and approvals. Keeping only the successful retest makes the actual process harder to explain.
The conclusion should state the acceptable operating range, product scope and remaining restrictions. Transfer validated conditions into work instructions, training, routine checks and monitoring. As the ISO 13485 and KGMP guide explains, certificates and operational evidence serve different roles.
Define what triggers reassessment
Changes in material suppliers, equipment parts, location, software or process settings can alter the original assumptions. Neither blanket full revalidation nor blanket exemption because the equipment is unchanged is a sound decision rule. Determine the work from impact and risk, retaining its approval rationale.
Start with the process map, inspection specifications, equipment/material inventory and recent changes. Bring them to a process-validation scope review to separate outputs verified by inspection from those needing process evidence, and resolve planning gaps before testing.
Sources checked: 2026-09-28. The tables and preparation steps are practical suggestions; confirm the legal submission scope and applicable standards for the particular product.
Frequently asked questions
- Q. Must every manufacturing process be validated?
- Not identically. Assess whether subsequent inspection can adequately verify the output, its limitations and the risk of failure before selecting validation and other controls.
- Q. Does PQ always require three batches?
- There is no universal number for every product. Consider process variation, risk, statistical rationale and applicable requirements, and document the batch and sample choices.
- Q. Can we use equipment-supplier acceptance records?
- They may support the work, but confirm whether they cover the installation environment and actual products and processes. Address gaps with additional evidence.