How to Document an Intermittent Problem

Published September 9, 2026 · Revised September 14, 2026

How to write a work order that captures what actually happened, what you tested, what you could not reproduce, and what evidence the next technician needs

Intermittent problems are some of the hardest failures in biomed.

Back to Biomed Basics

What This Page Explains

This page covers:

The Simple Version

Keep four categories separate: what staff reported, what you directly observed, what conditions and tests you used, and what the evidence supports. For example, document that staff reported two shutdowns during battery-powered transport; then separately state that no shutdown occurred during 90 minutes on battery with ECG, SpO2, and repeated NIBP cycles.

List the battery-capacity check, power-source transitions, connection inspection, and log review you completed, including duration and relevant accessories. Conclude narrowly: “Unable to reproduce the reported shutdown; applicable tests passed at the time of evaluation.” That record is honest about the unresolved risk and gives the next technician something reproducible. “NPF” hides the complaint, effort, and limitations.

Start With the Original Complaint

The work order should preserve the actual reported failure as accurately as possible.

Do not immediately rewrite the complaint into your diagnosis.

If the clinician says:

The monitor shuts off when we roll it down the hall,

do not document:

Battery failure.

You have not proven that yet.

A better entry is:

Staff reports monitor intermittently shuts down during transport.

That preserves the symptom without inventing the cause.

Reported vs Observed

One of the most useful habits in corrective-maintenance documentation is distinguishing:

Reported

from:

Observed.

That may sound minor, but it changes the quality of the record significantly.

Suppose staff says:

NIBP pump stops working after several readings.

You test the monitor for an hour and cannot reproduce it.

A good note might say:

Reported: NIBP would not inflate after repeated patient measurements. Observed: Completed 25 repeated NIBP cycles on simulator with no failure.

Now anyone reading the history can tell which information came from staff and which came from your bench testing.

Do Not Convert a Clinical Description Into a Technical Diagnosis

Clinicians often describe problems using practical language.

Examples:

Those descriptions may be correct.

They may also be interpretations.

Try to capture the underlying observation.

Instead of:

Battery bad.

Write:

Staff reports unit shuts off immediately when AC power is disconnected.

That gives the next technician something testable.

Record the Conditions

Intermittent problems often depend on conditions that disappear once the device leaves the clinical area.

Important conditions may include:

The more specific the complaint, the easier it becomes to reproduce.

Example

Weak:

Monitor drops connection sometimes.

Better:

Staff reports monitor loses central connection only during transport from ICU Room 8 to CT. Local monitoring remains active during event.

That immediately changes the troubleshooting path.

Now the next person knows this may involve:

rather than a general monitor failure.

Time Matters

If the complaint is intermittent, document when it occurred if known.

Useful timing includes:

Example:

Staff reports reboot occurred at approximately 14:32 after device had been operating for about 45 minutes.

That timestamp may allow correlation with:

“Sometimes” Is Not Enough

Try to convert vague frequency into something more useful.

Instead of:

Happens sometimes.

Ask whether it happens:

Even an estimate is better than nothing.

Accessories Matter

If the problem involved an accessory, document whether the actual accessory came with the device.

Examples include:

This matters because the device may test perfectly with a known-good accessory while the original problem followed the accessory.

Example

Original SpO2 extension cable involved in complaint was not provided with device. Monitor tested with known-good cable and sensor; unable to reproduce dropout.

That limitation is important.

Without it, a future technician might incorrectly assume the original cable was evaluated.

Document What You Could Not Test

This is one of the most underrated parts of good work-order documentation.

If you could not reproduce the original conditions, say so.

Examples:

These are not excuses.

They are boundaries on what your testing actually proved.

Test Duration Matters

If the complaint is:

Reboots after an hour,

and you test for:

10 minutes,

that is not equivalent.

Document how long you actually tested.

Example:

Operated unit on battery for 2 hours with continuous ECG/SpO2 simulation and NIBP cycling every 5 minutes. No shutdown observed.

Now the next technician knows the depth of the test.

Do Not Hide Short Testing

If you only had time to test for 15 minutes, document that honestly.

Do not write:

Extended test passed

unless the test was actually extended enough to mean something.

Record the Test Conditions

A useful intermittent-failure note should often include:

Example:

Unit operated on battery with ECG simulator connected, SpO2 active, and NIBP cycling every five minutes for 90 minutes.

That is much more informative than:

Tested okay.

Document Logs

If the device has event history, record what you found.

Useful entries include:

Example

Event log reviewed. Two unexpected shutdown entries were recorded at 14:31 and 14:34, matching the approximate time reported by staff.

That is strong evidence even if you cannot reproduce the problem on the bench.

No Log Is Still Worth Documenting

You might write:

No power, thermal, or watchdog faults present in available event history during the reported period.

That tells the next person the logs were checked.

Do Not Clear Logs Before Recording Them

If logs are relevant, capture:

before clearing or resetting anything.

For serious incidents, follow evidence-preservation procedures.

Record Physical Inspection Findings

Intermittent faults often involve:

Document both positive and negative findings when relevant.

Example:

Inspected battery contacts and internal connector. No corrosion, looseness, or physical damage noted.

Negative Findings Have Value

A good work order does not only say what failed.

It can also record what was ruled out.

For example:

Known-good AC cord and battery produced same behavior.

That removes two variables from future troubleshooting.

Do Not List Every Test You Can Think Of

Documentation should be useful, not bloated.

Record the tests that matter to the complaint.

For a network dropout, recording an electrical safety test may be required by procedure, but it does not explain the communication failure.

Prioritize complaint-related evidence.

Use Specific Language

Weak:

Checked device.

Better:

Verified AC input, battery operation, charger function, and AC-to-battery transfer.

Weak:

Network okay.

Better:

Ethernet link present, correct static IP confirmed, gateway ping successful, and application connection restored.

Specific words create useful history.

Avoid Vague Conclusions

Phrases like:

are weak.

They do not tell the next technician what was actually evaluated.

Better:

Unable to reproduce intermittent shutdown. Device completed two-hour battery load test without reset and passed manufacturer functional checks.

“No Problem Found” Needs Context

There is nothing wrong with an NPF conclusion when appropriate.

The problem is an NPF note with no evidence behind it.

A defensible NPF note explains:

Example

Weak:

NPF. Returned.

Strong:

Staff reported intermittent touchscreen freeze after approximately 30 minutes of use. Unit operated for 90 minutes with repeated touchscreen interaction and no failure. Touch diagnostic grid passed all regions. Event logs showed no touchscreen controller or system reset faults. Unable to duplicate reported condition.

Now “unable to duplicate” actually means something.

Do Not Overstate Certainty

If you cannot reproduce an intermittent problem, avoid language such as:

Device repaired.

unless you actually identified and corrected a specific defect.

A better phrase may be:

Unable to reproduce reported condition. All applicable testing passed at time of evaluation.

That is accurate.

When You Do Find Something

If you identify a likely intermittent cause, document the reproduction.

Example:

Intermittent shutdown reproduced by flexing battery connector during battery operation. Connector latch found damaged. Connector replaced. Unable to reproduce shutdown after repeated movement test and 90-minute battery operation.

That gives a complete failure-and-verification chain.

Document the Trigger

A strong note includes not only the failed part but how you proved it.

Example:

ECG lead-off reproduced when trunk cable flexed approximately 2 inches from connector. Known-good trunk cable eliminated failure.

That is powerful evidence.

Do Not Document Guesswork as Fact

Weak:

Software glitch caused reboot.

If the only evidence is:

It worked after restart,

you did not prove software caused it.

Better:

Unit rebooted prior to arrival. No hardware faults identified. Event history showed application watchdog reset at reported time.

Now you have evidence for the software/process side without overstating the cause.

Repeated Intermittent Problems Need History

The real power of documentation appears when the device returns.

Suppose the device has three work orders:

January

Staff reports shutdown during transport. Unable to reproduce. Battery passed.

March

Staff reports shutdown during transport. No fault found.

May

Staff reports shutdown while moving device through hallway. Reproduced loss of power by moving battery connector.

The January and March notes are not wasted if they clearly documented the conditions.

Together they reveal the pattern.

CMMS History Is a Diagnostic Tool

When a device returns with an intermittent complaint, review previous work orders before starting from zero.

Ask:

Patterns across months can reveal what one service visit cannot.

Write for the Next Technician

Imagine you are not writing the note for yourself.

You are writing it for a technician who has never seen the device before.

Would they know:

If yes, the documentation is doing its job.

Write for Your Future Self

Six months later, you may not remember:

Good documentation saves you from relying on memory.

Document Known-Good Substitution

If substitution was used, say what happened.

Example:

Original NIBP hose reproduced pressure leak. Known-good hose passed leak test on same monitor. Failure followed original hose.

That is far better than:

Replaced hose.

Document Service-Mode Evidence

If you used diagnostics, record useful findings.

Examples:

Do not dump 50 irrelevant diagnostic values into the note.

Record what supports the repair.

Network Intermittent Example

Weak:

Network checked. Works now.

Better:

Staff reported monitor intermittently missing from central. Bedside monitoring remained normal. Ethernet link remained active during testing. Device had correct IP and could reach gateway and central server. Central connection dropped twice during cable movement; known-good patch cable eliminated failure. Replaced patch cable and verified 60-minute central connection.

That note tells a story.

Battery Intermittent Example

Weak:

Replaced battery.

Better:

Staff reported unit shut down during transport despite battery showing approximately 70%. Battery capacity test failed at 28% of rated capacity. Known-good battery supported normal operation for 90 minutes and AC-to-battery transition passed repeatedly. Replaced battery and verified charging and runtime.

Temperature Intermittent Example

Weak:

Temp probe checked.

Better:

Staff reported temperature reading intermittently dropped to 20–22°C. Reading reproduced by flexing probe cable near connector. Monitor temperature input remained stable with simulator and known-good probe. Replaced patient temperature probe.

Touchscreen Intermittent Example

Weak:

Touchscreen okay.

Better:

Staff reported touchscreen becomes unresponsive after approximately 45 minutes. Unit operated for 75 minutes until failure reproduced. Physical hard keys remained functional while touch input stopped. Service diagnostics showed touch controller offline. Reboot restored function. Escalated for display/touch controller repair.

Unable to Duplicate Does Not Mean Staff Was Wrong

If a clinician reports something you cannot reproduce, document it respectfully.

Do not write:

User could not demonstrate problem.

unless that is specifically relevant.

Better:

Reported condition could not be reproduced under bench test conditions.

That keeps the focus on the equipment and evidence.

Serious Event Documentation

If the device is associated with a serious patient-care event, routine CM documentation may not be enough.

Follow facility procedures for:

Do not alter evidence unnecessarily.

Keep Technical Notes Technical

Avoid blame, emotion, or unsupported conclusions.

Bad:

Nurse probably unplugged it.

Better:

No loss of internal power reproduced. AC cord and battery transfer tested normally. Cause of reported shutdown not identified.

Common Mistakes

Writing Only “Could Not Duplicate”

It does not explain what was actually done.

Writing the Diagnosis Instead of the Reported Symptom

Preserve the original complaint.

Leaving Out Test Duration

Time-dependent failures need time-dependent testing.

Leaving Out Missing Accessories

Future technicians need to know the original setup was incomplete.

Documenting a Reboot as a Repair

A restart is not automatically root-cause correction.

Failing to Record Logs

Logs may be the only evidence of an intermittent event.

Overstating Certainty

If you do not know, say you do not know.

Writing So Little That Every Future Technician Starts From Scratch

The CMMS should accumulate useful knowledge.

A Useful Documentation Framework

For intermittent failures, use this mental structure:

Reported

What did staff say happened?

Conditions

When, where, and under what operating state did it occur?

Observed

What did you personally see?

Tested

What functions, accessories, and conditions did you evaluate?

Evidence

What did logs, measurements, or substitution show?

Limitations

What could you not reproduce or evaluate?

Conclusion

What is the most defensible technical outcome?

A Practical Example

Staff reports unit rebooted twice during transport while operating on battery. Original battery was installed at time of complaint. Inspected battery and power contacts; no visible damage noted. Event history showed two unexpected power-loss events near the reported time. Battery failed capacity test at 31% of rated capacity. Unit operated for 90 minutes on known-good battery with repeated movement and NIBP cycling without reboot. Replaced battery and verified charge function, AC/battery transition, and normal operation.

That note tells another technician almost everything they need to understand the repair.

What Did You Actually Prove?

If you tested a device for 20 minutes and it did not fail:

You proved:

The failure did not occur during those 20 minutes under those conditions.

You did not prove:

The intermittent failure cannot occur.

If you reproduce the problem and then eliminate it through a controlled component substitution:

You have much stronger evidence.

Good documentation should reflect the strength of that evidence.

Final Thoughts for Biomeds

Intermittent problems are difficult partly because the evidence disappears.

Your documentation is how you preserve what you learned.

Do not reduce a complicated complaint to:

NPF.

Capture:

A good work-order note can turn three separate mysterious service calls into one obvious pattern.

The goal is not to write a novel every time a cable flickers.

The goal is to leave enough technical evidence that the next person does not have to start over.

When you finish the work order, ask:

If this happens again, what did I actually prove today?

Then put that answer in the record.

— Jake

Important Note

Documentation requirements vary by healthcare organization, CMMS, manufacturer, incident type, and regulatory environment. Serious patient-safety events may require separate incident reports, equipment quarantine, log preservation, and formal investigation procedures beyond routine corrective-maintenance documentation.

Related Biomed Basics

Back to Biomed Basics