This Familiar, Seemingly Harmless Table is Quietly Undermining Effective Risk Management. Here's How.

This Table has Nothing to do with the Actual Requirements for Accepting Risk

ISO 14971:2019 defines four criteria for determining whether residual risk is acceptable:

1. Were risk controls implemented in the required priority order? (Inherently safe by design → protective measures → information for safety) (ISO 14971:2019 A.2.7.1)

2. Was the effectiveness of those controls verified? (A.2.7.2)

3. Has risk been reduced as far as practicable? (A.2.7.3)

4. Does the benefit outweigh the residual risk? (A.2.7.4)

That’s it. Those four statements, written directly into the risk management plan, fully satisfy both regulatory expectations and the ethical obligation of medical device manufacturers: minimize harm while delivering meaningful therapeutic benefit.

Anything beyond this—especially arbitrary numerical thresholds—creates decisions that are hard to defend and even harder to sleep on.

This Table Undermines “State of the Art” and Risk Reduction

When organizations adopt quantitative or semi‑quantitative “lines in the sand,” they unintentionally block the very concept of state‑of‑the‑art risk reduction.

If a table says a probability of 1 in 100,000 is “acceptable,” then no amount of new technology, engineering insight, or improved controls will matter. The table becomes the ceiling.

Consider a typical probability scale used in many SOPs.

When combined with a standard risk acceptance table, the following "limits" are derived.

• Catastrophic harm must be driven below 1 in 1,000,000 uses.

• Serious harm becomes “acceptable” at 1 in 100,000 uses.

This can lead to absurd conclusions.

An Example of Absurdity

Imagine a manufacturer of a life‑sustaining hemodialysis machine. The therapy is high‑benefit, and the device is considered safe if its overall residual risk is comparable to other machines on the market. Key inherent hazards include:

• blood loss

• air infusion

• microbial contamination

Through strong engineering controls, catastrophic harm is estimated at <1 in 1,000,000 therapies. Now consider a different hazard: a thermal event. The casing and circuit boards are flame‑retardant per ISO standards, so uncontrolled flames are not expected. The severity is therefore “serious,” not “catastrophic.” The machine delivers ~1,248 therapies per year, and 10,000 units are in the field—about 1.25 million therapies annually. Under the table‑driven logic, the manufacturer could receive 120 reports of devices catching fire per year and still declare the risk “acceptable.”

Hopefully, common sense would intervene far before this conclusion, but this is the risk with these tables.  They do not effectively drive application of appropriate controls and can become a sanctuary for complacency. 

There Is No Equation That Makes Harm “Okay”

How would a manufacturer defend the exact numerical thresholds they chose? Why is this number of harmed patients acceptable, but not more or less?

A far stronger, more defensible position is:

“We deliver an important therapeutic benefit and use state‑of‑the‑art technology to reduce risk as far as practicable. Here is our engineering analysis and our post‑market evidence demonstrating that performance.”

That is what regulators want to see. That is what ISO 14971 requires. That is what ethics demands.

How to Do Better

Notified Bodies often like seeing a risk acceptability matrix in an SOP. It's ok to keep the table in the SOP as a general guideline if it smooths an MDSAP inspection but don't plan on using for decisions or in the actual risk acceptability documentations for the product. Using it for decision‑making leads to problems and drives a check the box mentality to compliance that degrades quality and safety.

Instead:

• State the four true risk‑acceptance criteria in each product’s risk management plan.

• Ensure engineers use risk management to identify hazards, evaluate and implement controls, and verify effectiveness.

• Favor rigorous engineering analysis over checkbox compliance.

• Treat risk files as living documents, not static archives.

This is how the medical device industry delivers safe, innovative therapies that genuinely protect public health.

Looking for support in this area? A‑Z Continuous Compliance, LLC provides Risk Management process development, training, and creation of Risk Files. We can also develop culture change strategies that engages the organization cross-functionally through Risk Management value propositions.

Contact