NIST thermometer

[lef5501]

Just wondering how often everyone is getting their NIST thermometer recertified? I can't tell that ours has been done since it was bought long ago. This is the thermometer that we use to check all our other thermometers in-house. Some people are thinking that the certified NIST is good forever, however I feel that is should be recertified at some point.

[Franklyn]

We just send it out, once per year, to one of the companies that can issue calibration certificates. Sorry, but I don't recall the name. It is not very expensive and pretty quick-easy-painless to do.

[PamD]

You might want to check with your biomedical department. Our biomed dept has a company that recertifies their test equipment annually. When the company comes to the hospital to do the biomed stuff I have them do our NIST thermometer, a couple of IR thermometers, and our tach.

[SRMC BB]

We got caught on an inspection once about not having our NIST thermometer certified....So, now its goes out annually to Barnstead international (Which is now Thermo Fisher Scientific) for a 3 point re-cal. They issue a certificate and when it's returned I check all the thermometers in the lab against it.

Hope this helps!!!

[JohnD]

Just wondering how often everyone is getting their NIST thermometer recertified? I can't tell that ours has been done since it was bought long ago. This is the thermometer that we use to check all our other thermometers in-house. Some people are thinking that the certified NIST is good forever, however I feel that is should be recertified at some point.

There are a number of vendors out there that can provide the service. The turn around time is important to most customers. We have been doing a lot in the last 20+ years.

Keep in mind that a glass thermometer will change over time due to the strains in the glass and other factors.

Regards,

John Dozier

[Ellen Zeigler]

These need to be done on an annual basis. The only group that has had issue with us (in my former life with a blood center) has been the FDA. We were unable to convince the remainder of the clinical laboratories of this as a matter of good laboratory practice, so we can't borrow throughout the lab. Some facilities buy certified thermometers for the ranges that they use and discard when the calibration "expires". They state that they have found this less expensive than sending out an NIST on an annual basis. The same would apply to stopwatches and timers...

[JohnD]

These need to be done on an annual basis. The only group that has had issue with us (in my former life with a blood center) has been the FDA. We were unable to convince the remainder of the clinical laboratories of this as a matter of good laboratory practice, so we can't borrow throughout the lab. Some facilities buy certified thermometers for the ranges that they use and discard when the calibration "expires". They state that they have found this less expensive than sending out an NIST on an annual basis. The same would apply to stopwatches and timers...

The issue with discarding devices after the calibration expires is that you end up with a unknown whether that device was still in calibration at the time it expired. Even if it was compared to a new device there is a level on confidence (95%) can't be made unless the new device is 4 time more accurate in accordance with ANSI Z540. It is important to continue to have traceability from one device that would be discarded to the replacement device.

[Joanne P. Scannell]

Again, a lot of talk and extrapolation and we get all befuddled.

Back to reading the actual requirement(s) ...

TRM.31500 Is an appropriate thermometric standard device of known accuracy available? (NIST-certified or guaranteed by manufacturer to meet NIST standards.)

NOTE: Thermometers should be present on all temperature‑controlled instruments and environments and checked daily. Thermometric standard devices should be recalibrated or recertified prior to the date of expiration of the guarantee of calibration.

I could not find any reference in the FDA CFR requiring the use of an NIST certified thermometer. If someone can find that reference, please post it!

We purchase certified thermometers and use them as our 'standard' to check all other thermometers that are either not formally certified or have past their certification deadline. This is done Quarterly. It's simple ... line them all up in a waterbath (or incubator) just make sure at least one thermometer is indated certified and use that one as the 'correct temperature' to which others are allowed to vary whatever your tolerance is.

btw: Anyone hear that story about why mom cuts the ends of the ham before she cooks it?

[JohnD]

Again, a lot of talk and extrapolation and we get all befuddled.

Back to reading the actual requirement(s) ...

I could not find any reference in the FDA CFR requiring the use of an NIST certified thermometer. If someone can find that reference, please post it!

We purchase certified thermometers and use them as our 'standard' to check all other thermometers that are either not formally certified or have past their certification deadline. This is done Quarterly. It's simple ... line them all up in a waterbath (or incubator) just make sure at least one thermometer is indated certified and use that one as the 'correct temperature' to which others are allowed to vary whatever your tolerance is.

btw: Anyone hear that story about why mom cuts the ends of the ham before she cooks it?

Please see this post regarding the requirements of calibrating thermometers.

http://www.bloodbanktalk.com/forum/showthread.php?t=2978&page=2

When calibrating Liquid in glass thermometers a procedure such as NIST monograph 150 should be observed. The working standard should be 4 times more accurate then the unit under test in order to have a confidence factor of 95%. When calibrating using a liquid bath the standard and the UUT tips should be at the same depth to account for the temperature gradients within the bath fluid. In the case with 305mm full immersion thermometers either you must use a bath deep enough or calculate a stem correction to obtain the correct reading. Most cal labs use a fixed mounted scope with a reticle to read the glass thermometers to prevent any parallax and obtain a higher resolution of the reading.

[Joanne P. Scannell]

Are you a Blood Banker? Your posting looks and reads like you are a representative who pulled something out of the manufacturing regulations.

We don't do the actual calibrations out here in the field ... we just check our thermometers against a thermometer that has been calibrated, and certified as such, by a manufacturer.

[Alternatives]

Again, a lot of talk and extrapolation and we get all befuddled.

Back to reading the actual requirement(s) ...

I could not find any reference in the FDA CFR requiring the use of an NIST certified thermometer. If someone can find that reference, please post it!

We purchase certified thermometers and use them as our 'standard' to check all other thermometers that are either not formally certified or have past their certification deadline. This is done Quarterly. It's simple ... line them all up in a waterbath (or incubator) just make sure at least one thermometer is indated certified and use that one as the 'correct temperature' to which others are allowed to vary whatever your tolerance is.

btw: Anyone hear that story about why mom cuts the ends of the ham before she cooks it?

Hello JP,

Here an extract of a warning letter from FDA to an hospital based in Oklahoma in 2008, following an inspection:

"Failure to calibrate equipment used in the compatibility testing, storage, and distribution of blood and blood components on a regularly scheduled basis as prescribed in the SOP Manual to assure that it performs in the manner for which it was designed [21 CFR 606.60(a)].

Your firm has not calibrated the thermometer used to calibrate the

blood

bank

thermometers against a National Institute of Standards and Technology (

NIST

) traceable thermometer since December 29, 1992. In addition, the

blood

bank

thermometers that are used to monitor the refrigerators used for storing

blood

components and reagents, and the thermometers used in monitoring the temperatures during compatibility testing, are not calibrated on an established written schedule. "

(...) " In addition, there is no documentation that the thermometer used to calibrate the

blood

bank

thermometers has been calibrated against a thermometer traceable to a

NIST

standard at written intervals."

Hope it will help,

Benoît

Edited June 5, 2009 by Alternatives

[Joanne P. Scannell]

Let's look at 606.60 a) ....

§ 606.60 Equipment.

(a) Equipment used in the collection,

processing, compatibility testing, storage

and distribution of blood and blood

components shall be maintained in a

clean and orderly manner and located

so as to facilitate cleaning and maintenance.

The equipment shall be observed,

standardized and calibrated on

a regularly scheduled basis as prescribed

in the Standard Operating Procedures

Manual and shall perform in

the manner for which it was designed

so as to assure compliance with the official

requirements prescribed in this

chapter for blood and blood products.

Again, I cannot find using an NIST as a requirement ...

This inspection report you posted sounds like this BB hasn't been checking their thermometers at all which is indeed a violation of CFR 606.60. Doesn't matter how, you just have to check them periodically according to how you wrote your SOP. Yes, the FDA Inspector will determine if what you've written is acceptable ... this inspector is obviously stuck on the NIST method. There are other ways to assure thermometers are acceptable ... one is to use indated certified thermometers.

[clmergen]

"Indated certified thermometers" freshly purchased usually come with a certificate that is NIST traceable. At least the ones I purchase do.

As for the thermometers I do get calibrated, it is done using standards tracable to NIST and I get a document that says that. These thermometers are then used to validate that the rest of my other thermometers are reading correctly. I do all of this once per year and this is clearly documented.

[JohnD]

If a company choices to use its own mythology to calibrate any devices.They should have documented uncertainties of the processes and be able to prove that the methods are reproducible iif they are doing calibrations outside the scope of any governing bodies ie NIST and ANSI defined requirements. If not then what is being done isn't traceable back to the NIST standard.

Correction. If the bath has a stability of 2c and the standard is accurate to 1c then the root sum squared should be applied to have a 95% confidence in the standard reading.

This is reason ANSI Z540 suggest that the standard should be in all practical purpose 4 times better than what is being calibrated which is a k factor of 2. Otherwise the uncertainty should be documented on the calibration report.

[Lcsmrz]

JohnD:

Please stay in touch with this board and educate us!

Most blood bankers have never heard of Metrology and think a certificate of accuracy and NIST traceability from the manufacturer means the thermometer is accurate to one-third of the scale division -- that's what we were taught years ago. Uncertainty and uncertainty budgets are meaningless topics, and we can't figure out how two brand new NIST traceable thermometers could differ by 2 C inside the same temperature-controlled device, life a refrigerator.

My life was changed years ago after eating lunch with a Metrologist ...

[JohnD]

Some manufactures of liquid in glass thermometers only determine the zero point during manufacturing and expand the scale from that point. It has been our experience when purchasing thermometers for our customers a small percentage won't pass a multi point calibration because they were only checked at one point and expanding the scale from that point which leads to errors if capillary and the length of the glass aren't always the same. Our customers define the temperature points that they commonly use and if they don't we calibrate the full range and always include a 0c point.