Quick facts
- NIST Handbook 44 is a model code for commercial weighing and measuring devices, adopted by the National Council on Weights and Measures and published annually by NIST. It is not a Canadian regulation and it contains no code for a steam sterilizer.
- What the handbook does supply is a definition of traceability in practice: Appendix A lists seven essential elements, from realisation of SI units to accredited technical competence.
- It also supplies an acceptance arithmetic. A field standard used without correction must have a combined error and uncertainty smaller than one-third of the applicable device tolerance.
- A certificate that cannot be tied to one asset, one procedure, one reference standard and one stated uncertainty is not evidence; it is paper with a date on it.
Most calibration arguments in a clinic are not about whether the instrument was adjusted, but about whether the certificate proves it. This page reads NIST Handbook 44 for the two things it genuinely contributes, the traceability elements and the one-third acceptance rule, then converts them into a field-by-field test for accepting a calibration certificate. It is written for the clinic manager or quality lead who has a folder of certificates and no way to tell which ones would survive a question.
The commercial side of the same problem is familiar: a 24-hour self-contained biological indicator, 50-pack arrives with a certificate of analysis, a sterilizer arrives with a validation report, and an incubator arrives with nothing at all. Each of those documents is doing a different job, and only one of them is a calibration certificate. The indicator tape, 3-roll set illustrates the point in reverse: it is a process indicator whose performance is stated in a standard, not a measuring instrument whose error must be certified.
What is NIST Handbook 44, and what does it cover?
Handbook 44 is titled Specifications, Tolerances, and Other Technical Requirements for Weighing and Measuring Devices. The 2026 edition states in its introduction that the requirements "comprise all of those adopted by the National Council on Weights and Measures, Inc. (NCWM)", and records that the body renamed itself from the National Conference on Weights and Measures to the National Council on Weights and Measures with effect from 13 August 2024. NIST publishes it from the Office of Weights and Measures, the 2026 edition was approved by the NIST Editorial Review Board on 4 December 2025 and supersedes the 2025 edition, and it carries the identifier NIST HB 44-2026 with DOI 10.6028/NIST.HB.44-2026.
Two scoping facts matter more than the identifier. The first is that the handbook is a working document for weights and measures officials and for the manufacturers, installers and service agencies that deal with them, which the introduction says in as many words. The second is that its codes address commercial weighing and measuring equipment. Neither describes a clinical sterilizer, a biological indicator incubator or a chamber thermometer.
| Question | Answer for Handbook 44 | Consequence for a Canadian clinic |
|---|---|---|
| What equipment does it govern? | Commercial weighing and measuring devices: scales, liquid-measuring devices, vehicle-tank meters, fueling systems and comparable equipment | Covers none of the sterilizer instruments a clinic calibrates |
| Who adopts it? | The National Council on Weights and Measures, then state and local jurisdictions in the United States | Has no legal force in Canada on its own |
| Can it apply outside commerce? | The introduction allows applicable language to be cited as a standard in noncommercial applications, including where it is referenced as part of a quality system or in a multi-party contract | It binds a clinic only when the clinic's own quality system or a contract points at it |
| What does it supply that is transferable? | The traceability elements in Appendix A and the one-third rule for test standards | Directly usable as acceptance criteria for any certificate you are handed |
That last row is the reason this handbook is worth reading outside a weights and measures office. The technical content of Appendix A is not about retail scales at all; it is about what has to be true before a measurement is allowed to support an accuracy decision. The sterilization compliance hub keeps the documents that do govern the sterilization side of the same question, starting with the validation requirements in ISO 17665:2024 as they read for Canadian clinic practice.
What does Handbook 44 say traceability requires?
Appendix A, section 3.1.1, is headed "Essential Elements of Traceability" and states its purpose in one sentence: to ensure that field test standards and test methods provide for measurements traceable to the International System of Units through NIST or other national metrology institutes, they must satisfy the essential elements, which it attributes to NIST IR 6969, Good Measurement Practice for Ensuring Metrological Traceability. The elements are listed as realisation of SI units, an unbroken chain of comparisons, a documented calibration program, documented measurement uncertainty, a documented measurement procedure, accredited technical competence, and measurement assurance.
Read as a checklist rather than a preamble, those seven items are the only audit trail a certificate has. The definition underneath them is the one every accreditation body uses, from the International Vocabulary of Metrology: metrological traceability is the "property of a measurement result whereby the result can be related to a reference through a documented unbroken chain of calibrations, each contributing to the measurement uncertainty". A certificate is the document that carries one link of that chain, which is why a certificate naming no reference standard cannot contribute a link to anything.
The same appendix supplies a supporting condition in section 3.1: testing apparatus is adequate only when it is properly designed for its intended use, retains its characteristics under normal use, is available in denominations appropriate to the determination, "and when it is accurately calibrated". A clinic that buys a reference thermometer and never calibrates it has satisfied four of those five conditions.
What acceptance criterion does Handbook 44 set?
Section 3.2 of Appendix A states the arithmetic plainly. Where a standard is used without correction, "its combined error and uncertainty must be less than one-third of the applicable device tolerance". Where a correction is applied, "the uncertainty associated with the corrected value must be less than one-third of the applicable device tolerance". The stated reason is that the device under test should get as nearly as practicable the benefit of its own tolerance.
Two practical readings follow. First, a certificate has to state enough about the reference standard for you to check the one-third relationship; a certificate that says only "calibrated against a traceable standard" gives you nothing to divide. Second, Appendix A section 2.3 warns against adjusting equipment to sit at the edge of tolerance rather than as close to zero error as practicable, which is worth quoting to a service provider who returns a device at 95 percent of its allowable error and calls the job done.
The uncertainty half of that criterion only works if the certificate states an expanded uncertainty with its coverage factor. The NIST guidance on expressing uncertainty defines expanded uncertainty as the combined standard uncertainty multiplied by a coverage factor k, notes that k is typically in the range 2 to 3, and records that k = 2 corresponds to an interval of approximately 95 percent confidence when the normal distribution applies. A certificate quoting a bare "±0.5 °C" with no k and no confidence level has stated a number without stating what it means.
Which fields must a calibration certificate carry?
The table below converts the seven elements, the one-third rule and the uncertainty convention into the fields a reviewer should actually look for. The right-hand columns are the useful ones: the failure mode is what happens when the field is missing, and the remedy is what to ask for before filing the certificate.
| Field | Which requirement it serves | What goes wrong without it | What to request |
|---|---|---|---|
| Unambiguous identification of the item: make, model, serial number, asset tag | Ties the result to one physical device | The certificate cannot be attributed to the instrument in your room, so the chain has no starting point | Re-identification and reissue, and an asset tag on every calibrated device |
| The reference standard used, with its own certificate number and calibration due date | Unbroken chain of comparisons | The chain stops at the laboratory door and cannot be checked | The reference certificate identifier and its due date on the face of the report |
| The measurement procedure or method identifier | Documented measurement procedure | Results are not reproducible, and a second laboratory cannot be compared | The procedure reference, or a statement of the method used |
| As-found and as-left results, in SI units | Supports the accuracy decision in Appendix A | You cannot show whether the device had drifted or was adjusted back | Both values on the certificate, not just the final one |
| Expanded uncertainty with the coverage factor stated | Documented measurement uncertainty | A bare tolerance figure cannot be compared with the one-third rule | U with k, and the confidence level it corresponds to |
| Environmental conditions where they affect the result | Repeatability of the measurement | Results taken at different temperatures are silently mixed | The conditions recorded, or a statement that they were held within a stated range |
| A conformity statement, if the laboratory makes one | Turns numbers into a pass or fail claim | A "pass" with no cited tolerance is not a finding | The tolerance and the clause it comes from |
| Issuing laboratory identity and accreditation scope | Accredited technical competence | You cannot tell whether competence for that parameter was ever assessed | The accreditation body and the scope covering the parameter |
| Calibration date and issue date | Measurement assurance over time | Devices drop out of the recall cycle unnoticed | Both dates, plus the interval your programme assumes |
Nine rows is not a bureaucratic wish list. Each row maps onto one of the seven elements in Appendix A, which means a laboratory that works to ISO/IEC 17025:2017 is already producing most of them: that standard devotes clause 7.8 to reporting of results and clause 7.8.4 specifically to calibration certificates, and an accreditation assessment checks the reports against those clauses.
Why is a vendor certificate not the same as an accredited calibration?
Three documents that arrive in the same envelope are routinely confused. A certificate of analysis describes a manufactured lot and its test results. A validation or installation report describes a process and the evidence that it works. A calibration certificate describes one measurement performed on one instrument, with a stated uncertainty, by a laboratory whose competence for that parameter has been assessed.
Only the third one supports a traceability claim, and only when the laboratory's accreditation scope covers the parameter. A supplier-issued sheet with a sticker and a date is not a calibration certificate, however official the letterhead looks; a serial-number-specific report from an accredited laboratory is, even if it arrives as a plain PDF. The distinction matters when a hospital group, an accreditation surveyor or an insurance file asks who calibrated the reference instrument your sterilization records depend on.
How long should certificates be kept, and who asks for them?
Keep the certificate for as long as the device is in service plus the retention period that applies to the records it supports, and keep the reference standard's certificate alongside it, because the two documents are only meaningful as a pair. In a dental practice the anchoring record requirement is the sterilization log, which the RCDSO standard requires to be maintained for at least ten years from the date of the last entry; a calibration certificate that explains why a log entry looks the way it does should live at least as long as the log.
The people who ask are predictable. A College inspector asks about the records the practice's own manual promises. An accreditation surveyor asks for the evidence behind a quality-system claim. A hospital group's quality department asks for the certificates behind the monitoring results they receive. In each case the question is the same shape: can you show a chain from the result on your log to a reference standard, with each link documented.
Where does Handbook 44 not apply in a Canadian clinic?
It does not apply as law. Trade measurement in Canada is a federal matter administered by Measurement Canada under the Weights and Measures Act, and Handbook 44 is a United States model code adopted by United States jurisdictions. Nothing in a provincial sterilization standard, and nothing in the federal medical device framework, imports it.
It does apply when you choose to make it apply. The introduction's section on using the handbook says applicable language may be cited as a standard in noncommercial applications, for example when the handbook is referenced as part of a quality system or in a multi-party contract. That sentence describes exactly how a voluntary code acquires force: somebody writes it into a document that has authority over the work. If your quality manual cites Appendix A section 3.2 as the acceptance criterion for field standards, then it is your criterion, and a reviewer can hold you to it.
Keeping those two facts apart is the whole discipline. The handbook is useful precisely because it is technical and unglamorous, and it is dangerous only when a clinic cites it as though a Canadian regulator had adopted it.
Download the printable sterilization log sheets, or use the log generator on the same page to build a record that names the instrument behind each result; institutional buyers can open a wholesale account for monitoring stock, and a specific record question can go to the compliance desk.
Sources
- NIST Handbook 44-2026, Specifications, Tolerances, and Other Technical Requirements for Weighing and Measuring Devices (PDF)
- NIST Office of Weights and Measures, Handbook 44 current edition
- National Council on Weights and Measures, Handbook 44 2026 edition
- NIST Technical Note 1297, Guidelines for Evaluating and Expressing the Uncertainty of NIST Measurement Results
- JCGM 200:2012, International Vocabulary of Metrology (PDF)
- Joint Committee for Guides in Metrology
- ISO/IEC 17025:2017, General requirements for the competence of testing and calibration laboratories
- ISO 17665:2024, Sterilization of health care products, moist heat
- NIST Physical Measurement Laboratory, Office of Weights and Measures
- Public Health Ontario, provincial advisory committee document on reprocessing of medical equipment and devices (PDF)
- Royal College of Dental Surgeons of Ontario, Infection Prevention and Control standard
Related reading
- Calibration traceability to NIST: what sterilizer and monitoring records must show
- How SCC and CALA accreditation affect sterilization and contamination control in Canadian labs
- How to write a sterilization log that passes an inspection
Frequently Asked Questions
Does NIST Handbook 44 apply to a dental sterilizer?
No. The handbook's codes address commercial weighing and measuring devices, and it contains no code for steam sterilizers or for biological indicator incubators. Its traceability elements and its one-third rule are the parts a clinic can legitimately borrow.
Is a calibration certificate legally required in Canada?
Not as a general licence condition, but the records a clinic must keep do require it in practice. A College standard that obliges a practice to keep sterilization records makes the calibration behind those records a question a reviewer can ask, and the answer is the certificate.
What does "traceable to NIST" mean on a certificate?
It should mean the result is linked through a documented, unbroken chain of calibrations to a national reference, with each link contributing to the uncertainty. A logo or a phrase alone does not establish that, and the reference standard's certificate identifier is what lets you check it.
What uncertainty statement should a clinic insist on?
An expanded uncertainty with the coverage factor stated, so that the number can be compared with a tolerance. The NIST guidance on expressing uncertainty describes the expanded uncertainty as the combined standard uncertainty multiplied by a coverage factor typically between 2 and 3.
How long should calibration certificates be kept?
At least as long as the records they support. Where a sterilization log is retained for ten years from the last entry, the certificates behind that log should be kept for the same period, together with the certificates for the reference instruments used to produce them.
What should we do with a certificate that names no serial number?
Treat it as unusable and ask for a reissue. Without an identifier the report cannot be tied to the instrument in the room, so it contributes no link to the chain, and a reviewer will notice that the folder contains a document rather than evidence.
CliniEco Medical supplies sterilization monitoring consumables in Canada and distributes under Health Canada MDEL #35334.
0 comments