When to Trust the Device's Internal Self-Test

What a passing self-test really proves — and what it does not

Modern medical equipment checks itself constantly.

Published August 13, 2026 · Revised September 6, 2026

Back to Biomed Basics

What This Page Explains

This page covers:

The Simple Version

A self-test answers a limited question: did the device detect a problem in the functions it knows how to exercise and observe? A pass supports that those internal checks behaved as expected at that moment. It does not automatically prove that every clinical function is accurate, every accessory works, or the device is ready for use.

Use the service manual to learn the test boundary. A defibrillator may verify internal charging yet still require an analyzer to confirm delivered energy through the therapy cable. After a repair or safety-relevant complaint, combine self-test results with the required external functional, performance, alarm, and safety checks.

What Is a Self-Test?

A self-test is a diagnostic procedure performed by the medical device itself.

Depending on the equipment, it may check:

The exact scope varies enormously.

Power-On Self-Test

Many devices perform diagnostics during startup.

This may be called:

The device may check whether major internal components respond correctly.

If everything looks normal, startup continues.

If something fails, you may see:

A Startup Pass Means Something

If the device boots successfully, that does give you useful information.

It suggests that certain required startup checks passed.

For example:

But it does not prove every clinical function is accurate.

Self-Test Is Defined by the Manufacturer

This is one of the most important concepts.

A device tests only what its manufacturer programmed it to test.

Imagine a monitor startup routine that checks:

It may know the NIBP module exists.

That does not necessarily mean it knows whether the module is measuring pressure accurately across its full range.

Presence and performance are different things.

Internal Self-Test vs External Verification

An internal test asks the device to evaluate itself.

External verification compares the device against an independent reference.

Example:

A monitor says:

NIBP self-test passed.

External test:

Connect NIBP analyzer.

Set known pressure.

Compare displayed result.

Those tests answer different questions.

Think About the Reference

Every measurement needs some reference.

If a device checks itself using an internal reference, you are trusting:

What happens if the reference itself is wrong?

The system may not always be able to detect that.

This is why independent verification exists.

The Ruler Checking Itself Problem

Imagine a ruler trying to verify its own length markings.

If all of the markings expanded together, the ruler might still appear consistent with itself.

An external reference could reveal the error.

Medical devices can have a similar problem.

An internal check may confirm consistency.

External calibration or performance testing can verify accuracy against something independent.

Self-Test Can Be Excellent at Detecting Hard Failures

Internal diagnostics are often very good at finding:

These are clear failures.

The system expected a response.

It did not get one.

Self-Test May Be Less Powerful With Drift

Drift is harder.

Suppose a pressure sensor should read:

100 mmHg.

It now reads:

104 mmHg.

The sensor still responds.

The signal still looks reasonable.

Communication still works.

The device may have no reason to think anything is wrong.

An external accuracy check may catch it.

Passing Does Not Equal Calibrated

This is especially important.

A self-test can pass while a device is outside calibration.

Example:

Temperature circuit:

Self-test:

PASS.

External reference:

Actual 37.0°C.

Device displays:

38.2°C.

The system functions.

It is just wrong.

That is why calibration and verification are separate concepts.

Pre-Use Checks

Some equipment performs more extensive checks before clinical use.

Common examples include:

These checks may evaluate:

These can be much more comprehensive than a simple startup self-test.

Pre-Use Tests May Include External Components

This makes them especially useful.

A ventilator pre-use check may evaluate not just internal hardware, but also:

That means the test can sometimes catch setup problems that an internal startup test cannot.

But Pre-Use Checks Still Have Limits

A successful pre-use check does not necessarily prove:

Again:

Know what the test actually covers.

Continuous Self-Monitoring

Some devices monitor themselves during normal operation.

They may watch:

If a value moves outside acceptable limits, the device may generate:

These systems are extremely important.

But they still rely on sensors and logic that can themselves fail.

Watchdog Systems

Processors can freeze.

A watchdog is a safety mechanism designed to detect when the main software stops operating correctly.

The processor must periodically signal:

I am still alive.

If that signal stops, the watchdog may:

If you see repeated watchdog resets, do not assume:

Device rebooted and fixed itself.

The watchdog may be telling you there is an underlying software or hardware problem.

Plausibility Checks

Some devices compare measurements against what should physically make sense.

Example:

Sensor A says:

Pressure = 100.

Sensor B says:

Pressure = 0.

The device may decide those signals cannot both be correct and generate a fault.

This is a plausibility check.

Multiple sensors can improve self-diagnostics.

Redundancy

High-risk medical equipment may use redundant systems.

Examples:

One system can monitor another.

That improves fault detection.

But redundancy is still not a replacement for appropriate external verification.

Self-Test Failed

A failed self-test deserves investigation.

Do not treat it as:

Annoying startup message.

The test failed for a reason.

Record:

Then follow manufacturer troubleshooting.

Do Not Just Keep Rerunning It

This happens all the time.

Test fails.

Run again.

Fails.

Run again.

Fails.

Fourth time:

Pass.

Good enough.

Maybe not.

Ask:

Why did it fail three times?

The intermittent nature may be the most useful clue.

Possible causes:

A single pass does not erase repeated failures.

Repeated Passes Matter Too

On the other hand, repeating a test under controlled conditions can provide useful confidence.

Example:

You repaired a flow sensor connection.

Pre-use test passes.

You cycle equipment and repeat the test several times.

Passes consistently.

Now you have stronger evidence that the repair addressed the issue.

The difference is whether you understand what changed.

Failed Self-Test After Part Replacement

Suppose you replace a board.

Device now boots.

But calibration self-test fails.

Do not ignore it because:

At least it turns on now.

The repair may require:

Complete the manufacturer's required return-to-service process.

Internal Diagnostics Can Help Isolate Subsystems

Service mode may let you test individual components.

Examples:

These can be very useful.

If the service menu can activate the pump directly, you may be able to determine whether the pump itself functions independently of normal software control.

Direct Component Test vs Full System Test

Suppose:

Service mode:

Motor runs.

Normal operation:

Motor does not run.

That suggests the motor may be good.

Look toward:

This is a great example of using internal diagnostics logically.

“Test Passed” Messages Can Be Misleading Without Context

Imagine a monitor says:

Battery Test Passed.

What battery test?

Did it check:

You need to know.

A five-second internal check probably cannot prove a three-hour runtime requirement.

Battery Example

Internal check:

Battery detected.

Voltage acceptable.

Result:

PASS.

Clinical complaint:

Monitor only runs 15 minutes.

Both statements can be true.

The battery passes presence and voltage checks but fails capacity.

External runtime testing reveals the problem.

Alarm Example

Self-test checks:

Speaker produces a tone.

Pass.

Does that prove all alarm behavior?

No.

You may still need to verify:

Self-test may only have checked the speaker hardware.

Display Example

Startup displays every color briefly.

Pass.

That may help verify:

It may not detect:

Test the actual complaint.

Network Example

Device internal diagnostics:

Network adapter = PASS.

Yet:

Device cannot reach integration server.

The self-test may only know:

Network hardware is present.

It may not know:

A healthy network interface does not prove healthy communication.

Sensor Example

Device says:

Flow sensor recognized.

That proves communication or identification may be working.

It does not necessarily prove:

Recognition is only one level.

Error Logs and Self-Tests Work Together

Suppose the current self-test passes.

But error history shows:

Do not ignore that history just because the device behaves now.

The logs may reveal intermittent conditions.

Use Self-Test as One Piece of Evidence

A strong troubleshooting conclusion might be:

Device completed startup and internal diagnostics without errors. NIBP accuracy verified externally at multiple pressures and alarm functionality tested per procedure. All results within specification.

That is much stronger than:

Self-test passed.

You combined internal and external evidence.

When a Self-Test May Be Enough

Sometimes the manufacturer specifically states that a certain self-test is the required verification for a particular repair or function.

If the official service procedure says:

  1. Replace module.
  2. Run System Test X.
  3. If System Test X passes, return to service.

then follow the manufacturer's procedure and your facility requirements.

Do not invent unnecessary testing.

The point is not:

Never trust self-tests.

The point is:

Know what the manufacturer says the self-test proves.

Avoid Testing Just to Feel Better

There is another side to this.

Do not perform ten unrelated tests because:

More testing must be safer.

Tests should have a purpose.

If the manufacturer defines a validated verification procedure, follow it.

Good testing is not about generating the largest possible number of passing results.

It is about proving the required functions.

Real-World Example: Ventilator Passes Startup but Fails Pre-Use Check

Startup:

PASS.

Ventilator boots normally.

Pre-use check:

Flow calibration fails.

The startup test proved enough hardware worked for the system to boot.

It did not prove the flow measurement system was ready for patient use.

The pre-use check found the deeper problem.

Real-World Example: Monitor Passes Internal Test but NIBP Is Wrong

Monitor startup:

No errors.

NIBP module:

Recognized.

External analyzer:

Known 120 mmHg.

Monitor:

143 mmHg.

Internal diagnostics did not catch the accuracy failure.

The external reference did.

Real-World Example: Defibrillator Self-Test Says Ready

Device displays:

READY.

That status may represent a collection of internal checks.

But your scheduled PM or post-repair verification may still require testing:

The readiness indicator and the PM answer different questions.

Real-World Example: Battery Passes Internal Check

Transport monitor shows:

Battery health good.

On runtime test:

Device shuts down after 18 minutes.

The internal test did not prove capacity.

Now you understand the limitation.

Common Mistakes

Assuming Pass Means Everything Is Good

Know what was tested.

Ignoring a Self-Test Failure Because the Device Seems to Work

The diagnostic failed for a reason.

Rerunning Until It Passes

Investigate repeated failures.

Assuming Sensor Recognition Means Accuracy

Recognition and measurement are different.

Using Self-Test Instead of Required External Verification

Follow the service procedure.

Over-Testing Beyond Requirements

More testing is not automatically better.

Ignoring Logs Because Current Diagnostics Pass

Intermittent faults may only appear historically.

A Useful Question

Whenever a device says:

PASS

ask:

What exactly just passed?

That question prevents a lot of assumptions.

Then ask:

Is that the same thing I actually need to prove?

Sometimes yes.

Sometimes no.

Another Useful Question

Ask:

What would this self-test be unable to detect?

Possible answers:

That tells you what additional testing may be needed.

What Did You Actually Prove?

A successful self-test proves:

The device did not detect a failure in the conditions and functions evaluated by that self-test at that time.

That may be a lot.

It may be very little.

The meaning depends entirely on the test.

Do not claim more than it proves.

Final Thoughts for Biomeds

Internal self-tests are incredibly useful.

Use them.

They can save time.

They can isolate failures.

They can identify problems you would never find with visual inspection alone.

But understand their limits.

The device cannot necessarily prove everything about itself.

A passing startup test may not prove measurement accuracy.

A recognized sensor may not be calibrated.

A battery may be detected but have terrible capacity.

A network adapter may pass while communication still fails.

The right question is not:

Do I trust self-tests?

The better question is:

What does this particular self-test actually prove?

Once you understand that, internal diagnostics become much more useful.

Not as a replacement for troubleshooting.

As evidence inside the troubleshooting process.

— Jake

Important Note

Internal self-tests, diagnostics, pre-use checks, and readiness indicators vary significantly by equipment manufacturer and model. Follow current manufacturer service documentation, required performance-verification procedures, facility policy, and your authorized service scope when determining whether equipment is ready for clinical use.

Related Biomed Basics

Back to Biomed Basics