How to Investigate Repeat Failures

How to recognize when the same device, component, model, room, or workflow keeps failing and stop treating every new work order as an isolated event

One failed battery is a repair.

Back to Biomed Basics

What This Page Explains

This page covers:

The Simple Version

When a failure repeats, stop asking only:

What is wrong with this device today?

Also ask:

What is common across all of these events?

Possible common factors include:

That shared factor may be closer to the real cause.

What Counts as a Repeat Failure?

A repeat failure is not simply:

Device broke again.

It means there is enough similarity between events to justify asking whether they are related.

Same Exact Symptom

Example:

Monitor reboots during NIBP inflation three times.

Clearly repeat.

Different Symptoms Can Share One Cause

Example:

First complaint:

Random reboot.

Second:

Screen went black.

Third:

Lost central connection.

These could be separate.

Or they could all be caused by:

Read the details.

Same Asset

Start with:

Has this exact asset had the same or similar complaint before?

Same Model

Then broaden:

Are other devices of this model doing the same thing?

Same Department

Then:

Does the pattern follow a location rather than equipment?

Each level tells you something different.

Build a Timeline

For repeat failures, create a simple chronology.

Example:

January 4 — battery replaced February 20 — unexpected shutdown March 15 — battery replaced April 2 — shutdown during transport

Now you can see that:

Battery replacement did not resolve the deeper problem.

Timeline Helps You See Failed Repairs

A repair that only lasts a week deserves scrutiny.

Same Part Replaced Repeatedly

This is a major clue.

Example:

Fan replaced three times in one year.

Possible explanations:

The Part May Be the Victim

Repeated failure often means you have not identified what is causing that part to fail.

Ask What Could Damage It

If fuses keep opening:

Find the overcurrent.

If connectors keep burning:

Find the resistance or load.

Repeat Battery Failures

Batteries deserve special attention because they are commonly replaced without deeper investigation.

Possible Shared Causes

Example

Three batteries swell in the same device.

That should immediately raise questions about:

not simply battery quality.

Repeat NPF Is Still a Pattern

“No Problem Found” does not erase the complaint.

If the same device returns:

Five times

with:

The same intermittent shutdown,

the inability to duplicate is itself a repeated event.

What Condition Are You Missing?

Ask:

Repeated NPF Should Increase Curiosity

Not decrease it.

Same Location, Different Devices

This is one of the strongest clues for infrastructure.

Example:

Three monitors in Room 12 lose network.

The devices differ.

The location does not.

Investigate:

Same Device, Different Rooms

If the failure follows the asset:

Device-side cause becomes more likely.

Location vs Asset Matrix

A useful mental experiment:

Move suspect device to known-good environment.

Move known-good device into suspect environment.

Where does the failure go?

Environment

Repeat failures can come from:

Example

Several devices in one OR overheat.

All vents are blocked by custom cabinetry.

The problem is not the fan model.

Cleaning Patterns

Repeated failures in:

may relate to:

Look Beyond Electronics

The environment may be damaging the equipment.

Accessory Patterns

A problematic accessory can move between devices.

Example:

Same SpO2 extension cable follows a patient monitor cart.

Multiple monitors appear to have intermittent SpO2 problems.

The accessory is the shared factor.

Docking Stations

A portable device may repeatedly:

because the dock is bad.

Replacing device batteries does not fix the dock.

Software-Version Patterns

Modern fleet failures may correlate with:

Example

All devices on version 3.4 freeze.

Devices still on 3.3 do not.

That is strong evidence worth escalating.

Do Not Jump to Conclusion

It still does not prove the software is responsible.

Compare:

Fleet-Level Comparison

Make a simple table if needed:

Asset | Version | Failure | Location

Patterns often become obvious.

Configuration Pattern

A failure may follow:

Replacement Board Pattern

If every replacement board requires the same setting and that step is missed:

Several “bad boards” may simply be misconfigured.

OEM Bulletins

When a failure repeats across a model, check for:

Current manufacturer information may explain the pattern.

Do Not Assume Every Repeat Failure Is a Recall

Most are not.

But repeated unusual failures justify checking.

Same Failure After Same Repair

If a problem returns after a repair, ask:

Was the original repair actually verified?

Example

Cable replaced.

Complaint returns.

Maybe the cable was not root cause.

Or Replacement Quality

The replacement itself may be defective.

Compare Old and New Parts

If possible.

Repeat Mechanical Failures

Mechanical wear patterns can reveal:

Example

Same caster repeatedly breaks.

Maybe device is routinely crossing a severe threshold.

Clinical Workflow

Sometimes the failure is connected to how equipment is used.

Example

Pump doors repeatedly damaged because carts store them with tubing clamped in the door.

This is not simply:

Staff keeps breaking pumps.

It may be:

Human Factors Matter

If the same mistake happens repeatedly, the system may encourage it.

Power Quality

If several devices in one area experience:

consider:

Network Infrastructure

Repeat connectivity complaints may trace to:

Example

Different devices all lose communication around the same time daily.

That suggests shared infrastructure or scheduled network behavior.

Time-of-Day Patterns

Yes, time can be a shared variable.

Ask whether failures happen:

Example

Devices reboot every night at 2:00 AM.

That may involve:

Pattern by Clinical Task

Failures may occur only during:

That condition may matter more than asset identity.

Use Data, Not Memory Alone

People are bad at remembering frequency.

Use:

Count Failures

Instead of:

This keeps happening,

find out:

How many times?

Over what period?

Failure Rate Matters

Five failures across:

5,000 devices

is different from:

Five failures across 10 devices.

Fleet Denominator

If you are comparing models, consider:

Avoid Overreacting to High Counts Without Context

A large fleet naturally produces more work orders.

Severity Matters Too

Ten minor printer issues are not the same as:

Two therapy interruptions.

Prioritize by Risk

Repeat-failure investigation should consider:

Hidden Failures

A failure that produces a clear alarm may be less dangerous than one that silently produces incorrect output.

Repeat Accuracy Drift

If multiple devices repeatedly drift out of calibration:

Investigate:

Could the Analyzer Be Wrong?

If every device suddenly fails the same measurement:

Check the reference.

Test Equipment as Shared Factor

This is often overlooked.

If one analyzer is biased, you can create a fleet-wide pattern that does not actually exist.

Example

All infusion pumps appear 4% high.

Second analyzer shows normal.

Shared factor was the first analyzer.

Parts Lot

Replacement parts from one production lot can also create repeated failures.

Record part identifiers when relevant.

Repair Technique

If the same repair repeatedly fails after one technician or one procedure:

Look at:

Do this carefully and without turning it into blame.

Process Improvement

The goal is to improve the system.

Build a Hypothesis

Once you see a pattern, write a testable idea.

Example:

Devices on this cart lose power because battery connector opens during transport.

Now test it.

Do Not Call Pattern the Root Cause Yet

Pattern generates hypothesis.

Testing provides evidence.

Controlled Comparison

Compare:

Affected vs unaffected.

Example:

Overheating units:

Dirty filters.

Unaffected units:

Clean filters.

Now the relationship becomes stronger.

Change One Shared Factor

When possible, change one variable and see if failure follows.

Example

Move known-good monitor into room.

It loses network.

Infrastructure becomes likely.

Escalate Fleet Patterns Early

A systemic problem can affect many devices quickly.

If several devices show unusual identical failures:

Escalate to:

as appropriate.

Build the Escalation Package

Include:

Vendor Support Needs Patterns

OEM may already know about:

Document the Investigation

Create a record that tells future technicians:

Do Not Leave the Knowledge in One Person's Head

If only one technician knows:

These fail when used with Dock B,

the organization has not captured the lesson.

Real-World Example: Repeated Batteries

Monitor receives four batteries in one year.

Review shows internal temperature runs high because intake filter is blocked.

Battery heat exposure likely accelerates degradation.

Battery replacement alone never addressed the environment.

Real-World Example: Repeat Network Complaints

Different patient monitors lose central only in hallway near radiology.

Wi-Fi survey identifies coverage gap.

The devices are fine.

Real-World Example: Repeat Touchscreen Failures

Several touchscreens fail in same department.

Inspection repeatedly shows cleaner residue at bezel.

Cleaning method becomes the shared factor.

Real-World Example: Repeated Main Boards

Three main boards replaced for reboot complaint.

Problem returns.

Power rail monitored under NIBP load and drops sharply.

Power supply was the actual shared cause.

Real-World Example: Fleet Software Issue

Multiple devices begin freezing after same firmware deployment.

Older version units remain stable.

Logs show identical application fault.

Escalated to OEM with fleet evidence.

Now you have a systemic case, not five unrelated tickets.

Common Mistakes

Treating Every Work Order Independently

History matters.

Replacing the Same Part Again Without Asking Why

Repeated part failure is a clue.

Ignoring NPF Work Orders

Repeated NPF can still reveal conditions.

Blaming Users Without Looking at Workflow

Find the mechanism.

Calling a Pattern With Too Little Data

Separate coincidence from trend.

Ignoring Test Equipment as a Shared Factor

Your reference can create false patterns.

Keeping the Pattern in Your Head

Document it.

A Useful Repeat-Failure Framework

Ask:

What is repeating?

Then:

What is shared across those events?

Then:

What differs between affected and unaffected equipment?

Then:

Can I change one variable and make the failure follow it?

Then:

Is this asset-specific, location-specific, model-specific, or system-wide?

Another Useful Question

Ask:

What if the failed part is not the root cause but the thing that keeps getting sacrificed?

That question is especially useful with:

What Did You Actually Prove?

If five devices fail after a software update:

You proved:

The failures correlate with the update.

You did not yet prove:

The software caused them.

If rollback removes the failure while everything else stays the same:

Your evidence becomes much stronger.

If one device repeatedly fails in multiple rooms while known-good devices work:

You have strong evidence the failure follows the asset.

Repeat-failure investigation is about strengthening those relationships until the pattern becomes technically defensible.

Final Thoughts for Biomeds

A repeat failure is a message.

The equipment is telling you:

The last explanation may not have been complete.

Do not automatically replace the same part again.

Zoom out.

Look at:

Then ask what every failure has in common.

Sometimes the answer is inside the device.

Sometimes it is the charger.

Sometimes it is the network.

Sometimes it is the way the equipment is used.

The important skill is recognizing when the work order in front of you is not really a new problem.

It is another chapter in the same one.

And before you close another repeat repair, ask:

What did we actually learn this time that we did not know last time?

— Jake

Important Note

Repeat-failure investigations may involve technical troubleshooting, manufacturer escalation, quality review, clinical workflow evaluation, IT or Facilities support, or formal safety processes depending on the equipment and severity. Follow current facility procedures and preserve relevant service history, logs, configuration information, and affected-part data when investigating recurring or fleet-wide problems.

Related Biomed Basics

Back to Biomed Basics