Why Potential Effects of Failure and Severity Should Start with Clinical Intent, Not Failure Modes

FMEA 4.0: Why Potential Effects of Failure and Severity Should Start with Clinical Intent, Not Failure Modes

In the earlier article, DFMEA and PFMEA Are Not Separate Worlds, I described how design and process FMEAs are two views of the same risk reality and why Potential Effects of Failure and Severity (S) must be shared across both.

There is a deeper implication behind that idea.

If Severity belongs to the effect, and the effect belongs to the patient or user, then those effects do not need to wait for design or process failure modes to exist.

In fact, waiting is often the mistake.

Potential Effects of Failure Exist Before Failure Modes

Potential Effects of Failure are not technical artifacts. They are outcome statements.

They describe what could happen to:

• A patient

• A clinician

• A user

• The clinical workflow

These outcomes are driven by intended use, clinical environment, and human interaction, not by specific design details.

You do not need to know the architecture of a device to identify outcomes such as:

• Over delivery of medication

• Under delivery of therapy

• Delayed treatment

• Incorrect clinical feedback

• Loss of critical monitoring

Those effects exist the moment intended use is defined.

Severity Is a Clinical Judgment First

Severity is often treated as an engineering estimate. In reality, it is primarily a clinical judgment.

Severity answers a single question:

If this effect occurs, how serious is the impact on the patient or user?

The most qualified people to answer that question are not CAD designers or process engineers. They are:

• Physicians

• Nurses

• Clinicians

• Experienced healthcare professionals

These subject matter experts understand:

• Real patient harm

• Clinical consequences

• Workflow disruption

• Use error amplification.

Engineers play a critical role in translating these insights into design and process controls. But the definition of the effect and its Severity should originate with clinical expertise.

Concrete Example: Infusion Pump

Consider an infusion pump early in concept development.

Before any design architecture exists, clinicians can already identify a critical Potential Effect of Failure:

Potential Effect of Failure:
Over-delivery of medication to patient

Effect Description:
Excess medication delivered over a short period, potentially leading to acute patient harm.

Severity (S):
9

This assessment does not require knowledge of:

• Control algorithms

• Sensors

• Software architecture

• Manufacturing processes

It only requires understanding of clinical therapy, patient physiology, and real-world use.

Later in the lifecycle:

• A DFMEA failure mode such as an incorrect flow rate algorithm may link to this effect.

• A PFMEA failure mode such as a calibration error may link to the same effect.

Different failure modes.
Same clinical outcome.
Same Severity.

Decoupling Effects from Failure Modes Is a Maturity Step

Traditional FMEA approaches often tie Potential Effects of Failure directly to failure modes.

This creates two problems:

• Effects are discovered late.

• Severity shifts as designs evolve.

In FMEA 4.0, effects are treated as stable, reusable risk elements.

They can be:

• Identified early

• Assessed clinically

• Stored independently

• Reused across DFMEA and PFMEA

Failure modes then become what they should be: implementation-specific contributors to known outcomes.

Traditional FMEA Timing vs FMEA 4.0 Timing

Traditional FMEA Timing

• Design concepts are developed first.

• Failure modes are brainstormed.

• Potential Effects of Failure are inferred from failure modes.

• Severity is assigned late and often revised repeatedly.

• DFMEA and PFMEA develop separate effect definitions.

FMEA 4.0 Timing

• Intended use and clinical context are defined early.

• Potential Effects of Failure with clinical impact are identified upfront.

• Severity is assessed once by clinical SMEs.

• Effects become stable risk anchors.

• DFMEA and PFMEA link to these effects as failure modes appear.

This shift does not reduce rigor. It increases it.

Linking Early Effects to DFMEA and PFMEA Later

Once Potential Effects of Failure are defined, they do not sit in isolation.

As the design matures:

• DFMEA failure modes are linked to existing effects.

• PFMEA failure modes are linked to the same effects.

• Severity stays consistent.

• Traceability strengthens automatically.

The result is a risk model that reflects reality rather than document boundaries.

Why This Matters

When Potential Effects of Failure are found late:

• Clinical risks are underestimated.

• Severity discussions become technical instead of patient focused.

• DFMEA and PFMEA drift apart

• Audit defensibility weakens.

When effects and Severity start with clinical intent:

• Risk discussions become clearer.

• Design decisions improve.

• Manufacturing risks stay aligned with patient impact.

• Lifecycle traceability becomes natural.

The Takeaway

Potential Effects of Failure do not belong to DFMEA or PFMEA.

They belong to the patient and the clinical use of the device.

Severity does not depend on how a failure happens. It depends on what happens when it does.

As organizations mature their risk management approach, many discover that treating effects as first-class, reusable elements naturally leads toward more structured, connected ways of managing FMEA data across the lifecycle.

That structural shift often matters just as much as the methodology itself.

As organizations mature their risk management approach, it becomes clear that this way of working is not reliably achievable with document-based FMEAs. Treating Potential Effects of Failure and Severity as first-class, reusable risk elements is best implemented using a structured FMEA database, such as the NoioMed FMEA 4.0 Database Schema described in an earlier article. That schema intentionally separates Potential Effects into their own authoritative structure and allows them to be linked, one-to-many, to DFMEA and PFMEA failure modes across the lifecycle. This is what makes early clinical Severity definition, cross-FMEA consistency, and long-term traceability practical rather than aspirational.