Validation Engineer Interview Questions for Medical Devices (With Answers)

Validation engineers are asked to prove that processes, equipment and software work consistently. Interviews check your grasp of IQ, OQ and PQ, statistics and documentation discipline, and how you handle a failed run. This guide has 8 model questions and two case studies with ideal answers for process validation, sampling and software validation roles.

What to prepare

Have one validation you can explain from protocol to report: what the process was, how you set acceptance criteria and sample sizes, what failed and what you did. Revise process validation, sampling plans and software validation.

Eight common questions with model answers

1. When is process validation required?

Model answer: “When the output cannot be fully verified by later inspection or test, for example sterilisation, sealing, welding or moulding, or when testing is destructive. ISO 13485 clause 7.5.6 requires it. If every unit can be fully and non-destructively verified, verification may be enough.”

2. Explain IQ, OQ and PQ.

Model answer: “IQ confirms correct installation. OQ shows the equipment works across the operating range, including worst-case limits. PQ shows the process consistently makes conforming product under normal production conditions with trained operators.”

3. How many PQ runs do you need?

Model answer: “No regulation sets a number. Three consecutive runs is common practice, but I justify the number by risk, variability and prior knowledge. A high-risk, variable process would need more.”

4. How do you choose a sample size?

Model answer: “From the risk of failure. For attribute data I use confidence and reliability. With zero failures allowed, 95% confidence and 95% reliability needs 59 samples; 95% and 90% needs 29. For variable data I use tolerance intervals after checking normality. I document the rationale before testing.”

5. What do you do if a PQ run fails?

Model answer: “Stop, document the deviation, investigate the root cause, and do not simply repeat until it passes. Decide on a corrective action, then repeat the affected stage under an approved protocol amendment or new protocol, and record everything including the failure.”

6. When is revalidation needed?

Model answer: “When there are changes to equipment, materials, software, parameters, suppliers or location, when trends or complaints show drift, or at a periodic review interval defined by risk.”

7. How do you validate software used in the quality system?

Model answer: “Define intended use, assess risk of each function, assess the supplier, then match assurance activities to risk: scripted testing for high-risk functions, lighter methods for low risk. Record the results and conclusion, control changes and review periodically. This follows ISO 13485 and FDA’s risk-based assurance approach.”

8. Tell me about a validation you led.

Model answer (STAR): “We had to validate a new heat sealer (situation). I wrote the protocol and owned execution (task). OQ at the temperature and time limits showed a narrow window at high speed, so I proposed a tighter operating range and ran PQ in that range with three lots (action). Seal strength passed with a capability index above our target, and the process went live on schedule (result).”

Case study 1: protocol written after the data

Scenario: A colleague says the OQ data looks good and asks you to write the protocol now so the report can be finished.

Ideal answer: “I would decline. An approved protocol must exist before execution, with acceptance criteria set in advance. Writing it afterwards makes the criteria look chosen to fit the data and is a serious data integrity issue. I would tell the quality manager, record a deviation and repeat the work under an approved protocol.”

Case study 2: supplier says the process is already validated

Scenario: A contract sterilizer provides a validation report and says no further work is needed.

Ideal answer: “I remain responsible. I would review whether the report covers our product, packaging and loading pattern, whether it follows the relevant standard such as ISO 11135, 11137 or 17665, whether acceptance criteria meet our specification, and how revalidation and changes are controlled by contract. If gaps exist, I would require additional work before release.”

Mistakes to avoid

• Quoting 3 runs or 30 samples with no justification.
• Describing validation as paperwork rather than evidence.
• Not mentioning failures. Interviewers prefer honest examples of what went wrong.

Frequently asked questions

Do I need statistics skills?

Yes, at least confidence and reliability, capability indices and measurement system analysis. Practise explaining them simply.

Is software validation different from process validation?

The principle is the same, but methods differ. Software uses intended use, risk-based testing and change control, while processes use IQ, OQ and PQ.

What documents might they ask to see?

Sometimes a redacted sample protocol or report. Remove confidential data before sharing anything.

Model answers are examples; adapt them to your real experience.

Talk to the AptSkill team

Preparing for a validation or quality engineering interview? Send us a message.