When an instrument returns from calibration with an out-of-tolerance (OOT) result, an important question is whether previous measurements could have been affected.
An OOT result means the instrument failed to meet the applicable calibration acceptance requirements. However, this does not automatically mean that previous measurements were incorrect or that affected products failed to meet their requirements.
The possible impact depends on several factors. These include the affected measurement function and range, the size and direction of the error, how the instrument was used, and the requirements of the application.
A structured investigation can help identify which measurements need review, whether previous decisions may have been affected, and what actions are needed.
The investigation should also document the evidence and reasoning used to reach its conclusions.
Key Takeaways
- An out-of-tolerance result does not automatically invalidate previous measurements. The potential impact depends on the affected function or range, how the instrument was used, and the application’s requirements.
- The as-found result does not establish when the problem began. Use calibration history, intermediate checks, usage records, and other evidence to establish a justified review period.
- Evaluate whether the error could have changed a previous decision. Consider its size and direction, relevant measurement uncertainty, and the original acceptance requirements.
- Document the evidence, assumptions, and limitations. Explain what supports a conclusion of identified impact, no identified impact, or insufficient evidence.
- Returning the instrument to service does not resolve its previous-use impact. Adjustment, repair, and successful calibration address the instrument’s condition, while earlier measurements require a separate evaluation.
The following considerations provide a framework for evaluation. The scope, sequence, and depth of the investigation should reflect the application, available evidence, and applicable quality-system requirements.
- Understand What the OOT Result Means
An instrument is generally considered out of tolerance when one or more as-found calibration results fail to meet the applicable acceptance requirements. This includes the agreed decision rule when one is used.
As-found results show how an instrument performed before any adjustment or repair.
For example, consider a digital voltmeter tested at 10 V DC. The instrument has an allowable error of ±0.020 V, but calibration finds an error of +0.030 V.
The instrument fails its tolerance requirement at that test point.
However, this result alone does not tell us whether previous measurements were affected.
The investigation must consider how the instrument was used. An error that exceeds the instrument’s calibration tolerance may have little effect on one application but could change an important decision in another.
The calibration result identifies the problem. How the instrument was used determines the possible impact.
Review the Specific Calibration Findings
Before reviewing previous measurements, identify:
- The affected measurement function, range, and test points.
- The instrument tolerance or acceptance requirement.
- The size and direction of the reported error.
- The reported measurement uncertainty.
- The decision rule used to determine whether the instrument passed or failed, if applicable.
- Any relevant notes about the instrument’s condition, settings, or configuration.
An OOT result at one test point does not necessarily mean the instrument performed incorrectly across every function or range.
Likewise, passing results at other test points do not prove that the instrument performed correctly at points that were not tested.
Understanding the specific OOT finding helps determine which applications may need further review.
- Control the Instrument and Preserve the Evidence
When an OOT condition is found, the instrument should be controlled according to the organization’s procedures to prevent unintended use.
The organization should also keep the information needed to understand the instrument’s as-found performance.
Important records may include the calibration certificate, as-found readings, instrument settings, relevant accessories, and notes about damage or unusual conditions.
This information is especially important because adjustment or repair can change the instrument’s performance.
If more information is needed, the customer should contact the calibration provider promptly.
For example, the customer may need details about the reported error, measurement uncertainty, or the instrument’s condition when it arrived for calibration.
Keeping this information helps ensure that the investigation is based on measurement evidence rather than assumptions.
- Determine Which Previous Measurements May Have Been Affected
An important first step is determining whether the affected measurement function or range was actually used.
Many instruments have several functions, ranges, and operating modes. A failure in one area does not automatically mean that every measurement made with the instrument needs review.
Identify How the Instrument Was Used
Start by identifying the instrument’s actual applications.
Useful information may include:
- Measurement functions and ranges used.
- Inspection, testing, or production procedures.
- Recorded measurements and acceptance limits.
- Equipment users and responsible departments.
- Jobs, products, lots, or processes involving the instrument.
- Equipment usage and inspection records.
Consider a torque wrench used only at 50 lbf·ft. Its calibration shows an OOT result at 75 lbf·ft.
If records confirm that the wrench was used only at 50 lbf·ft, the investigation should evaluate the available evidence supporting its performance at that value. If the evidence supports a conclusion that previous measurements were not affected, further review may not be needed. Otherwise, additional evaluation may be necessary.
This is why the review should focus on how the instrument was actually used rather than assuming that every use created a measurement problem.
Establish the Investigation Period
The investigation must also consider when the instrument may have been performing outside its requirements.
A previous passing calibration provides evidence of the instrument’s performance at that time. It does not tell us exactly when a later OOT condition began.
Information that may help establish the review period includes:
- Previous calibration results and performance trends.
- Intermediate checks or independent measurements.
- Maintenance and repair records.
- Known damage, overload, or handling problems.
- Environmental conditions.
- Changes in how the instrument was used.
For example, a documented event or a reliable intermediate check may help support a shorter review period.
However, the investigation should not assume that an instrument began drifting halfway through its calibration interval. It should also not assume that the error increased at a steady rate.
If the evidence does not support a shorter review period, the organization may need to review measurements made since the last acceptable verification or another justified starting point.
The goal is to establish a review period supported by the available evidence.
Use Reverse Traceability to Identify Potentially Affected Work
Reverse traceability means tracing an instrument’s previous use to find measurements, inspections, products, or process decisions that may have been affected.
For example, an organization might follow this path:
Instrument ID → Equipment usage records → Inspection results → Acceptance decisions → Affected products or processes.
This helps identify which work needs further evaluation.
It can also help exclude work that did not involve the affected measurement function or range.
Missing or incomplete usage records may make the review more difficult. In those cases, the organization may need to review more work or find other supporting evidence.
- Evaluate Whether the Error Could Have Changed a Previous Decision
Once the affected applications and measurements have been identified, the next step is to determine whether the OOT condition could have changed previous results or decisions.
This requires comparing the calibration finding with the requirements of the original application.
Consider the Size and Direction of the Error
Both the size and direction of the error matter.
For example, if an instrument reads higher than the actual value, it may cause a measurement to appear larger than it really is.
Whether this matters depends on the application’s acceptance limits.
A measurement close to an acceptance limit may need more review than one well within the acceptable range.
However, being far from an acceptance limit does not automatically prove that a previous decision was correct.
The evaluation should consider the possible measurement error, uncertainty, and available evidence about the instrument’s past performance.
Consider Measurement Uncertainty
Measurement uncertainty characterizes the spread of values that could reasonably be attributed to the quantity being measured, based on the available information.
The uncertainty reported on a calibration certificate applies to the calibration measurement under the stated conditions. It does not necessarily represent the uncertainty of measurements previously made with the instrument.
Previous measurements may have been affected by other factors, such as temperature, the measurement method, instrument resolution, repeatability, and how the measurement was performed.
The organization should consider the uncertainty information that applies to the original measurement and any decision rule used to determine acceptance.
A complete uncertainty analysis of previous measurements may not be needed for every OOT investigation.
The level of review should depend on the application, the available evidence, and the possible consequences of an incorrect decision.
Review Other Supporting Evidence
Other records may help determine whether previous decisions were affected.
Examples include independent measurements, intermediate checks, retained samples, process monitoring records, and measurements made with other suitable instruments.
However, each type of evidence has limits.
For example, measuring a retained part again may provide useful information about its current condition. It may not prove what the part’s condition was during the original inspection.
Likewise, a passing intermediate check provides evidence of instrument performance at the time of the check. It does not necessarily prove that the instrument performed correctly throughout the entire period.
Supporting evidence should be used only for what it can reasonably establish.
- Example: How the Same OOT Result Can Have Different Impacts
Consider a fictional digital voltmeter identified as VM-017.
During calibration, the following results are obtained:
| Calibration information | Result |
| Instrument | Digital voltmeter VM-017 |
| Test point | 10.000 V DC |
| Indicated value | 10.030 V |
| Instrument tolerance | ±0.020 V |
| As-found error | +0.030 V |
| Expanded calibration uncertainty | 0.004 V (k = 2) |
| Conformity result | Fail, using simple acceptance |
For this example, simple acceptance means the measured error is compared directly with the instrument tolerance, without guardbanding. The reported uncertainty is not added to or subtracted from the tolerance limits.
VM-017 fails its calibration requirement at the 10 V test point.
Now consider two applications involving this same instrument and OOT finding.
Both applications have the same acceptance limits, but the previous measurements are different.
To show how the error could affect previous measurements, assume the meter had the same +0.030 V error during earlier use and across the measurement values shown. An actual investigation would need evidence to support these assumptions.
Situation A: Previous Measurements Well Within Application Limits
VM-017 was used to check a nominal 10 V DC power supply. The acceptable voltage range was 9.50 to 10.50 V.
Complete usage records show that all measurements during the review period were between 9.90 and 10.10 V.
If the meter had the same +0.030 V error during those measurements, subtracting that error gives:
- Lowest recorded reading: 9.90 − 0.030 = 9.87 V.
- Highest recorded reading: 10.10 − 0.030 = 10.07 V.
Both values remain within the application’s acceptance limits.
This comparison suggests that the identified error may not have changed the previous acceptance decisions.
However, the investigation must still consider whether the instrument could have had a different error during the review period. Relevant measurement uncertainty and application conditions must also be considered.
If the available evidence shows that possible past measurement errors would not have changed the acceptance decisions, the investigation may support a conclusion of no identified impact.
That conclusion would apply only to the measurements reviewed. It would not mean that VM-017 was in tolerance or suitable for continued use.
Situation B: A Previous Measurement Near an Acceptance Limit
Now consider VM-017 being used for another nominal 10 V DC power supply check.
The acceptance limits are still 9.50 to 10.50 V.
During a previous inspection, the meter displayed 9.52 V. The power supply was accepted because the reading was within the specified limits.
Calibration later found that VM-017 was reading 0.030 V high at the 10 V test point.
If the same error existed during the earlier inspection, the corresponding voltage would have been:
9.52 − 0.030 = 9.49 V.
This value is below the application’s lower acceptance limit of 9.50 V.
The comparison shows that the same OOT error could change a previous acceptance decision when the recorded reading is close to an acceptance limit.
However, the final as-found error does not prove that the meter had the same error during the earlier inspection.
The investigation would need to consider past performance, measurement uncertainty, application conditions, and other available evidence.
The 9.52 V reading identifies a measurement that needs further review. It does not, by itself, prove that the power supply failed to meet its requirements.
What the Two Situations Demonstrate
Both situations involve the same instrument, the same +0.030 V error, and the same application acceptance limits.
In Situation A, the recorded measurements were far enough from the limits that the identified error alone would not change the previous decisions.
In Situation B, the recorded measurement was close enough to the lower limit that the same error could change the acceptance decision.
Neither comparison proves what the instrument’s actual error was during previous use. That requires supporting evidence.
These examples show why an OOT investigation should focus on how the instrument was used and whether the error could have affected previous decisions.
- Document the Evidence Supporting the Conclusion
An OOT investigation should produce a record showing what was reviewed, what evidence was available, and how the conclusion was reached.
A statement such as “no impact” provides little useful information without an explanation.
A stronger record connects the calibration finding to the actual applications and explains why the evidence supports the conclusion.
What Should the Investigation Record Include?
Depending on the organization’s requirements and the investigation, useful records may include:
| Record element | Information to capture |
| Instrument identification | Asset ID, model, serial number, and calibration report |
| OOT finding | Failed function, range, test point, error, tolerance, uncertainty, and decision rule, if applicable |
| Review scope | Applications, equipment use, and time period reviewed |
| Supporting evidence | Usage records, calibration history, intermediate checks, and process records |
| Technical evaluation | Comparison with application requirements, possible measurement effects, and relevant uncertainty |
| Limitations | Missing information, untested conditions, or assumptions |
| Conclusion | Identified impact, no identified impact, or insufficient evidence |
| Disposition | Actions involving affected work and the instrument |
| Review and approval | Responsible personnel, approvals, and completion records |
The goal is not to create more paperwork. It is to keep enough information for another qualified person to understand what was reviewed and why the conclusion was reached.
What If the Evidence Is Incomplete?
Consider the VM-017 example again.
Suppose the customer cannot find complete usage records or cannot determine whether the instrument’s past error was consistent with the assumptions used in the evaluation.
The calculations alone would not be enough to support a conclusion of no identified impact.
The organization may need to review other records, obtain independent evidence, expand the investigation, or apply temporary controls.
If important information is still missing, the investigation record should explain what is unknown and how it affects the conclusion.
Not finding evidence of an impact is not the same as proving that no impact occurred.
- Determine Appropriate Actions and Close the Investigation
An OOT investigation can lead to several outcomes.
No Identified Impact
The available evidence supports the conclusion that the OOT condition did not affect the previous measurement decisions reviewed.
The conclusion should identify the work covered and any limits of the evaluation.
Identified Impact
The evidence shows that one or more previous measurements or decisions were affected.
The organization must decide what actions are needed under its quality system. These may include actions involving products, processes, customers, or regulatory requirements.
Insufficient Evidence
The available information does not support a reliable conclusion about previous measurements.
More investigation, temporary controls, or other actions may be needed based on the possible risk.
ISO 9001:2015 requires organizations to determine whether the validity of previous measurement results was adversely affected when measuring equipment is found unfit for its intended purpose, and to take appropriate action as necessary. This requirement remains in the newly released ISO 9001:2026.
An OOT calibration result does not, by itself, establish that the instrument was unfit for its intended purpose. That determination requires evaluating the calibration findings against the requirements of the application in which the instrument was used.
Not every inconclusive investigation indicates a problem with the measurement management system. However, when missing records or inadequate controls prevent the organization from evaluating previous measurements, the organization should investigate why that information was unavailable.
This may reveal a separate problem with equipment tracking, recordkeeping, or other measurement controls. The organization should evaluate whether corrective action is needed to address the cause and prevent the problem from happening again.
Separate Instrument Disposition From Previous-Use Evaluation
Returning the instrument to service and evaluating previous measurements are two separate activities.
An instrument may be adjusted, repaired, and successfully calibrated without resolving whether previous measurements were affected.
Likewise, an investigation may find no identified impact on previous work even though the instrument still needs adjustment, repair, or calibration before returning to service.
The customer is responsible for deciding what actions are needed for affected products and processes under its quality system.
The calibration provider can explain the calibration results and provide supporting information. However, the customer must determine how the instrument was used and decide what actions are needed for previous work.
Closing the impact investigation also does not necessarily close a separate root-cause investigation or corrective action. Those activities should be handled under the organization’s quality system.
Conclusion
An out-of-tolerance calibration result is a reason to evaluate previous measurements. It does not automatically mean that previous work was affected.
An effective investigation identifies the affected measurement function or range, determines how the instrument was used, and evaluates whether previous decisions could have changed.
It also documents the evidence supporting the conclusion.
The goal is not simply to determine whether the instrument passed or failed calibration. It is to understand what the OOT finding could mean for the measurements and decisions that depended on that instrument.
When the evidence supports a conclusion, the investigation should explain why. When important information is missing, the record should identify those limits and any actions needed.
This approach helps organizations make informed decisions and maintain clear records of how possible measurement impacts were evaluated.
Frequently Asked Questions
Does an out-of-tolerance calibration result mean previously accepted products are nonconforming?
No. An OOT result identifies a failure to meet an instrument’s applicable calibration acceptance requirements. Whether previously accepted products were affected depends on the instrument’s actual use, the application’s requirements, and evidence about its performance during that use.
How far back should an OOT investigation go?
The review period should be supported by evidence, including previous calibration results, relevant intermediate checks, and documented events. If that evidence does not justify a shorter period, the review may need to extend to the last acceptable calibration or verification covering the affected function and range. A previous passing result does not establish exactly when the OOT condition began.
Can the as-found calibration error be applied to previous readings?
Only when evidence supports its use for the measurements and period being evaluated. The error found during calibration does not prove that the same error existed earlier or at other measurement values. Applying it to previous readings can help explore a possible impact, but the calculation alone does not establish what actually occurred.
Does a passing as-left calibration result close the OOT investigation?
No. Passing as-left results establish the instrument’s performance under the reported calibration conditions at the end of service. They do not determine whether its earlier condition affected previous measurements or acceptance decisions.
What if instrument usage records are missing?
Review other available evidence, such as inspection records, production logs, operator information, and independent measurements. Missing records may require a broader review or additional controls. If the evidence remains insufficient, document what is unknown, how that limits the conclusion, and what actions are needed under the organization’s quality system.
Learn More About Evaluating Out-of-Tolerance Results
For a deeper explanation of calibration evidence, historical measurement performance, and previous-use evaluations, download SIMCO’s eBook, Beyond the OOT Result: Could Previous Measurements Have Been Affected?