Why Not Use Mean Kinetic Temperature in Cold Applications? Here’s why…

Paul Daniel, Vaisala

Senior GxP Regulatory Compliance Expert, Continuous Monitoring

Published: Dec 14, 2017

Life Science

Back for more on Mean Kinetic Temperature with a customer question:

Hi Paul,
I have reviewed Vaisala’s application note on mean kinetic temperature. But, I am still wondering what particular guidance states that mean kinetic temperature should only be used for room temperature environments and not for refrigerators or freezers? That is Rule #5 on your 6 rules for MKT.
But I’m still unclear to me exactly why not to use it in colder temperatures. Your assistance is greatly appreciated.

Thank you!
K

Paul Daniel wrote:

Dear K,
Thank you for checking out our training materials! We are happy to provide these to our customers and contacts in the industry at large. Here are the links to the webinar and the application note.

First off, you are not alone in your question, so below I have listed several questions we received in our last webinar on mean kinetic temperature. You will see that your question about cold chambers was asked then as well. And it is a very common question.
Second, to review our application note and the rules you mention, I will list all of them again here.

6 Mean Kinetic Temperature Usage Recommendations:

Mean kinetic temperature should not be used to compensate for temperature excursions in any application.
When using MKT, ensure you have an adequate number of samples (time/temperature). The more samples that are included in the equation, the more the calculation will represent the actual MKT value.
Mean kinetic temperature should not be used in areas where temperature is not well controlled.
Use mean kinetic temperature only if the storage temperature specified on the label of the product does not exceed 25˚C.
Mean kinetic temperature should not be used for products that require controlled low temperature.
Regardless of whether you use the mean kinetic temperature calculation or not, all temperature excursions should be investigated.

Finally, the answer to your question: Why Not Use MKT in Cold Chambers?

Because the best guidance we can use, and probably the only one to use in regulated environments, is the source of the idea that we can use mean kinetic temperature to evaluate temperature excursions: John Taylor.

Here is the name of the paper: The Pharmaceutical Journal (Vol 267), July 2001, J. Taylor. “Recommendations on the Control and Monitoring of Storage and Transportation Temperatures of Medicinal Products.”

I draw your attention to the lower left-hand corner of page 131 in the paper:

Strict conditions should be applied to the use of mean kinetic temperature. It is applicable only to storage of products under controlled room temperature conditions (I.E., those labeled “do not store above 25°C”). Mean kinetic temperature can be applied to the shipping of these products and it may provide a degree of assurance during prolonged transit periods. Mean kinetic temperature is not appropriate for products requiring controlled low temperature storage and shipping.

This paper basically brought the mean kinetic temperature calculation into its current use and describes how and why to use it. Simply put, the guy who first thought that MKT might be good for evaluating excursions also explicitly stated where not to use it, saying: “[MKT] is not appropriate for products requiring controlled low temperature storage and shipping.” It was a great innovation in its time, but we recommend sticking to the original guidance that accompanied the use of mean kinetic temperature.

That’s why we made recommendation 5 in our application note. We can also find that rates of change in chemical reactions and degradation pathways further support this position. Our caution against using the value for cold chamber applications is directly from the seminal paper on the topic, but also from the likelihood that you might have an uphill battle convincing an auditor that mean kinetic temperature for refrigerated chambers is a better solution for evaluating excursions than the other tools we at our disposal. Please contact me further if you have any questions!

Paul Daniel

Leave your comments below and scroll down to see the Q&A from our mean kinetic temperature...

Questions & Answers on Mean Kinetic Temperature

Is there a time frame or minimum number of samples required for calculating mean kinetic temperature?

That is up to your application. If you are looking at a short excursion, you’d want to capture that timeframe. With a short excursion analyzed inside a larger timeframe, you will then decrease the impact of the excursion. You need to pick the right time frame to evaluate the impact. If you are storing something g in a warehouse and you know that product is there for an average of two weeks, you’d want to look at a two week timeframe. Because if you looked at the duration of the excursion, your MKT value will be too high; if you look at a year of data, MKT will be too low.
In terms of number samples, I’d use as many samples as you can get because that increases the accuracy of the calculation.

If you have a temperature deviation in a warehouse – how long back in the temperature history should I go?

I wouldn’t go any longer than the amount of time that the product has actually been stored in the warehouse. That would be your long limit.

Why is the use of mean kinetic temperature limited to 25 – 30° Celsius? What if I have a chemical reaction that follows the Arrhenius equation at higher temperatures, isn’t MKT also ok within that higher range?

These recommendations come from the John Taylor paper that was published, and subsequently taken into use. We don’t have the opportunity now to ask him why 30° C. I think that if you can demonstrate, and you have the data from stability testing, to show that mean kinetic temperature works as a good way to evaluate your product quality at higher temperatures, go for it. But, without that data, we recommend following the accepted, and more conservative recommendations laid out in the Taylor article.

Is there a preferred way to calculate mean kinetic temperature for regulators?

First we need to clarify that using MKT is not a requirement, it’s a tool. So, you won’t find any information on how to make calculation in GxP regulations and guidance. However, if you are using the value at all in a regulated environment, you need to use a validated tool for calculation. The viewLinc enterprise software has the ability to calculate MKT, but you would need to validate the system for use in any GxP environment.

Averages seem like the easiest numbers to use as inputs for mean kinetic. Any reason not to use averages?

Yes, there are very good reasons not to and I would avoid their use in mean kinetic temperature calculations. Averages by their nature hide fluctuations in data. However, mean kinetic temperature data is intended to account for the impact of those fluctuations, so we don’t want to hide them.

Can we simplify the mean kinetic temperature calculation by using the min/max for the day?

If you recall our webinar, we had a slide showing some history …and this is a backward step. The only time I would use a min/max value for mean kinetic temperature is if I were using it for its original purpose in stability testing. For trying to assess product damage, min/max is not a good idea because they don’t represent reality.

Temperature change throughout a day isn’t linear. It introduces an assumption that 50% of the time was spent at maximum and 50% of the time was spent at minimum. With modern sensing systems that have sample rates you can set –1 minute, 5 minute, etc. – there is no reason to use a Min/Max for mean kinetic temperature.

Your rules make it sound like mean kinetic temperature is only good for controlled storage at room temperature. Is this not overly limiting the use of mean kinetic temperature?

Yes… the rules make it sound that way because that’s where we stand on the use of this calculation. The concern is that we not consider the calculation of MKT a silver bullet for temperatures-sensitive product storage. It’s just another tool for analyzing data. It’s useful in specific situations, that is: relatively rare excursions in well-controlled storage at room temperature. I would love it if MKT was a more general tool.

We need to remember that we have many tools at our disposal. The focus should always stay on maintaining our controlled environments in a stable state by more traditional means: mapping studies, monitoring, calibration, preventative maintenance. The mean kinetic temperature value is our last line of defense in special situations only.

Articles on GxP standards & regulations

Visit our repository of articles, white papers, application notes and eBooks...