FMEA or FTA? Why Choosing the Method Will Not Save Your Risk Analysis

In many MedTech projects, the same pattern can be observed: as soon as a hazard analysis or risk analysis is due, the question of which method to use comes up very quickly. FMEA? FTA? Or several in parallel right away? The question is understandable, but it falls short. In practice, the quality of the analysis is not determined by the method, but by the fit between the question, the method, and integration of the results.

Methods Are Not Yet a Risk Analysis

Techniques such as PHA (Preliminary Hazard Analysis), FMEA (Failure Mode and Effects Analysis), FTA (Fault Tree Analysis), ETA (Event Tree Analysis), HAZOP and HACCP are well-established tools in risk management. What is often overlooked: they are not synonymous with the hazard analysis, and certainly not with the complete product risk analysis required by ISO 14971:2019.

Each method answers a different question:

• PHA structures hazards and initial harm scenarios early and systematically.
• FMEA analyses failure modes, their effects and their causes at the component or process level.
• FTA makes complex causal relationships and combinations of causes visible by means of the fault tree.
• ETA shows how initiating events develop along protective barriers.
• HAZOP uses guide words to systematically challenge deviations from documented intended operation.
• HACCP focuses on the identification and monitoring of critical control points.

The decisive point: a failure mode is not automatically a hazard. A top event is not automatically a risk. The output of these techniques is an intermediate result, not a finished entry in the risk management file.

The Real Hurdle Lies at the Interface

The typical problem in practice is not carrying out the method. It is the question of how its results are actually translated into the risk analysis. Between the method output and the risk analysis there is almost always a necessary step: the analytical translation into the logic of ISO 14971, that is, into hazard, hazardous situation, sequence of events, harm, risk controls and evidence of effectiveness.

ISO 14971 requires a continuous chain from the hazard through the hazardous situation to the harm, including evaluation and demonstrated control of the risk. An FMEA provides building blocks for this, such as failure modes and their effects, but it does not automatically place these into precisely this chain. This is exactly where most weaknesses arise. Not because methods are applied incorrectly, but because their results are not carried through consistently in the file.

A recurring pitfall: the FMEA table is neatly filled in, but no one can show during the audit which failure mode leads to which hazardous situation and which risk control specifically addresses that situation. The method was correct, but the traceability to the risk management file is missing.

More Methods Do Not Automatically Mean Better Analyses

A second misconception: that the quality of the hazard analysis increases with the number of methods used. On the contrary. In many projects, combining several techniques creates typical knock-on problems:

• Overlapping results and duplicated effort. The same hazards are described several times without any additional insight being gained.
• Inconsistent terminology. If failure mode, hazard and harm are named differently across FMEA, FTA and PHA documents, later traceability suffers.
• Artificial complexity. A larger number of documents does not lead to more clarity, but to less, particularly during internal audits and regulatory inspections.

The decisive question is therefore not "Which method do we use?", but: "Which question do we want to answer, and which method actually delivers an insight for it?" An FTA makes sense when the interplay of several causes needs to be understood. An FMEA fits when failure modes of individual components or process steps are to be worked through systematically. If both are used, their results must converge on the same risk management file, not be documented side by side.

How to Recognise a Robust Hazard Analysis

A good hazard analysis is not evident from the number of documented methods or the level of detail of individual tables. It is evident from four characteristics:

• Visibility: Critical risks and hazardous situations become visible early and are addressed during the design process, not only at final verification.
• Traceability: Cause pathways and chains of failure are logically comprehensible to third parties, such as auditors, and documented without gaps.
• Prevention: Risk controls are derived in a targeted manner and proportionate to the actual hazard, in line with the residual risk evaluation required by ISO 14971.
• Substantiation: Evaluation decisions on severity and probability are technically and, where relevant, clinically justified, not set out of habit.

In short: it is not methodological breadth but analytical coherence that is decisive. Methods should not be applied mechanically but deliberately, derived from the question to be answered.

What Needs to Be Done

The most effective lever does not lie in another method, but in the clean integration of the method output into the risk management file in line with ISO 14971. In practical terms, this means: a common terminological framework across all techniques used, a continuous link from failure mode through hazardous situation and harm to the risk control and its evidence of effectiveness, and a deliberate choice of method based on the specific question.

Entourage helps MedTech manufacturers transfer existing FMEA, FTA or PHA results consistently into an audit-ready risk management file and align method selection with the respective question, instead of running methods in parallel. The underlying relationships are explored in depth in the detailed white paper on transferring method output into the product risk analysis to ISO 14971.