Webinar 4: Why FMEA alone is not sufficient for risk analysis

FMEA is a useful technique for analyzing potential failure modes, but it is not sufficient by itself to satisfy the requirements of risk analysis according to ISO 14971.

Dear colleagues, hello! 👋

Failure Modes and Effects Analysis (FMEA) is a commonly used technique for risk analysis in the medical device industry. Although it is useful for analyzing the risk of potential device failures, it is not sufficient by itself to analyze and control the risk of harm as required by ISO 14971.

In this month’s webinar, we discuss why FMEA should be treated as only one of the techniques suitable for medical device risk analysis and that it should be complemented with other techniques for a more thorough and complete analysis.

Our discussion focused on 3 key aspects of medical device risk analysis:

1. Risk analysis in the context of ISO 14971 involves two steps: risk identification and risk estimation.

2. Risk identification occurs in the context of the intended use, reasonably foreseeable misuse and characteristics related to safety. Each risk is characterized by a specific combination of a hazard that is activated through a sequence or combination of events in a hazardous situation. Risk estimation involves combining the probability of harm with severity of that harm.

3. FMEA is a one-to-many technique, whereas hazard/harm analysis considers various many-to-one scenarios. Additionally, FMEA is not suitable for risk analysis when a device is operating in the normal mode as designed.

We illustrated the link between hazards, hazardous situations and harms using a top-down and a bottom-up approach. This concept is important to fully appreciate the limitations of FMEA when used alone for risk analysis in the context of ISO 14971. Using a case study, we illustrated the many-to-one relationship between hazards and harms. Finally, we discussed a practical technique to link an FMEA with hazard/harm analysis using applicable hazards as connecting points. This approach can be easily implemented using a relational database.

Here are some of the questions from the audience we discussed during the webinar:

1. How to identify characteristics related to safety and how to use them in hazards identification?

2. How to identify use-errors?

Watch the full recording with audience Q&A below.

Useful references:

1. Case study: When benefit-risk falls below expectations despite solid clinical data

2. Case study: Using post-market data to evaluate changes in risk level

3. ISO 14971 fundamentals: Risk management file documentation structure

4. Recorded webinar: Hazard analysis made easy

🙏Thanks for watching! Suggest topics for future webinars in comments below🔽