Medical Device Storage Temperature Tests and Requirements
- Thread starter Remus
- Start date
planB
Super Moderator
Re: Medical Device Storage Temperture
Remus,
there is no way around performing real-time aging tests when you have a product with limited shelf life, which can be driven either by the device itself, its packaging or both.
However, you can use accelerated aging study data for submissions and later marketing upon approval. Specifically, if you have a sterile product, ISO 11607-1:2006, Packaging for terminally sterilized medical devices -- Part 1: Requirements for materials, sterile barrier systems and packaging systems, is your friend, and tells you in sections 6.4.3 - 6.4.7 what to do. ASTM F1980 will then guide you through a standardized method to apply to your sterile barrier system.
When it comes to aging of the device itself, ISO 11607, sections 6.4.5 and 6.4.6 also apply:
HTH,
Gerhard
Remus,
there is no way around performing real-time aging tests when you have a product with limited shelf life, which can be driven either by the device itself, its packaging or both.
However, you can use accelerated aging study data for submissions and later marketing upon approval. Specifically, if you have a sterile product, ISO 11607-1:2006, Packaging for terminally sterilized medical devices -- Part 1: Requirements for materials, sterile barrier systems and packaging systems, is your friend, and tells you in sections 6.4.3 - 6.4.7 what to do. ASTM F1980 will then guide you through a standardized method to apply to your sterile barrier system.
When it comes to aging of the device itself, ISO 11607, sections 6.4.5 and 6.4.6 also apply:
HTH,
Gerhard
Re: Medical Device Storage Temperture
My question is diffrent.
How I support evidence for upper and lower limits of storage temperture. ASTM F1980 doesn't cover that. For instance I want to put -10C to 40C on my label, which test should I perform? It is hard to perform real time aging test -10C for 2 years.
My question is diffrent.
How I support evidence for upper and lower limits of storage temperture. ASTM F1980 doesn't cover that. For instance I want to put -10C to 40C on my label, which test should I perform? It is hard to perform real time aging test -10C for 2 years.
planB
Super Moderator
Re: Medical Device Storage Temperture
Remus,
without knowing the specifics of your device, it seems a bit unusual to label a storage temperature range spanning 50°C.
Food for thought for you:
HTH, Gerhard
Remus,
without knowing the specifics of your device, it seems a bit unusual to label a storage temperature range spanning 50°C.
Food for thought for you:
- Why not simply label "room temperature" for your device instead of such a large temperature range?: room temperature would be within your proposed range - you would be able to rather easily validate RT, and moreover, your users can easily follow such a temperature label
- Usually, aging process are driven by temperature (at least, that's the point of view accepted by regulators), which means that aging is accelerated by higher temperatures. That's why you usually only test your upper temperature limit - lower temperature slow down aging processes. This is also the idea behind ASTM F1980 and the Q10 theory laid out in there. Sure there are special devices around that should not be exposed to too low temperatures. So is your device prone to getting damaged below -10°C? Then you might think of limiting the transport temperature, but still set the storage temp. to RT.
HTH, Gerhard
Re: Medical Device Storage Temperture
I agree that in order to give meaningful answers more details about the nature of the device / application are required.
To address the temperature range issue - many devices are stored in a non-controlled environment, which in some places means temperatures as high as (and higher than) 40degC or as low as (or lower than) -10degC. It's not just transport, some warehouses are just not temperature controlled, and while there's a damping effect (the temperature inside is not the same as outside) their temperature range can exceed the range -10 to 40.
In that sense it's more true to reality to simulate aging through thermal cycling, because that's what the device is going to see over day/night variation (and seasonal changes). Some devices have failure mechanisms that resemble fatigue and are promoted by cyclic change. Thermal expansion and contraction can induce such cyclic stressing.
In general, accelerated aging and especially thermally-induced accelerated aging is a complicated topic that requires careful evaluation for each specific instance. The regulators have mostly chosen to follow a "one size fits all" approach, but I think that applying that approach mindlessly can sometimes lead to erroneous and even absurd results. I think regulators just couldn't handle the complexity and vast variation of devices in that regard, so they just chose the easy way out. The Q10 simplified model and the entire Arhenius concept carry some preliminary assumptions that are mostly ignored (or are unknown) by the majority of developers. Where those assumptions don't hold the use of the model becomes pseudo-science.
In my opinion each device being developed requires a thoughtful analysis to come up with a valid aging model and valid ways to prove it is actually valid. It's not a 5 minute exercise and unfortunately there's currently no published standard that I know of that deals with that mine field.
Anyone interested in more is welcome to contact me publicly or privately.
Regards,
Ronen.
I agree that in order to give meaningful answers more details about the nature of the device / application are required.
To address the temperature range issue - many devices are stored in a non-controlled environment, which in some places means temperatures as high as (and higher than) 40degC or as low as (or lower than) -10degC. It's not just transport, some warehouses are just not temperature controlled, and while there's a damping effect (the temperature inside is not the same as outside) their temperature range can exceed the range -10 to 40.
In that sense it's more true to reality to simulate aging through thermal cycling, because that's what the device is going to see over day/night variation (and seasonal changes). Some devices have failure mechanisms that resemble fatigue and are promoted by cyclic change. Thermal expansion and contraction can induce such cyclic stressing.
In general, accelerated aging and especially thermally-induced accelerated aging is a complicated topic that requires careful evaluation for each specific instance. The regulators have mostly chosen to follow a "one size fits all" approach, but I think that applying that approach mindlessly can sometimes lead to erroneous and even absurd results. I think regulators just couldn't handle the complexity and vast variation of devices in that regard, so they just chose the easy way out. The Q10 simplified model and the entire Arhenius concept carry some preliminary assumptions that are mostly ignored (or are unknown) by the majority of developers. Where those assumptions don't hold the use of the model becomes pseudo-science.
In my opinion each device being developed requires a thoughtful analysis to come up with a valid aging model and valid ways to prove it is actually valid. It's not a 5 minute exercise and unfortunately there's currently no published standard that I know of that deals with that mine field.
Anyone interested in more is welcome to contact me publicly or privately.
Regards,
Ronen.
Re: Medical Device Storage Temperture
Although I'm not implying either way whether this needs to be completed or not, really, it's not that hard.
It would involve a freezer, temperature control and monitoring, and 2 years worth of time. Turn it on, check that the temperature has held on a daily basis, and maybe a UPS in case of power loss. It wouldn't even take that many man hours.
Although I'm not implying either way whether this needs to be completed or not, really, it's not that hard.
It would involve a freezer, temperature control and monitoring, and 2 years worth of time. Turn it on, check that the temperature has held on a daily basis, and maybe a UPS in case of power loss. It wouldn't even take that many man hours.
Re: Medical Device Storage Temperture
While practically true, there are several problems with this approach. For example:
1. They will have to wait for 2 years before they have the final results. Suppose that 2 years shelf-life is a strict design input requirement, and after 22 months they discover that the device is failing - then what? Implement a change and go for another 2 years run?
2. The outcome will probably provide relatively little insight, as it will essentially be a binomial outcome - pass/fail. Sample size will have to be rather big to be statistically valid, and even bigger if they choose to test at interim points (eg every 3 months). If they only test at the end it will provide even less insight and might be too late for an efficient reaction.
3. Is it a representative scenario at all? Have real life transport and storage scenarios been analysed? Will the device actually spend up to 2 years at -10deg C, with minimal variation? Does that correlate with any foreseen failure mechanism in the device? Or maybe it's more likely that the device will experience thermal cycling (the characteristics of which are important), which might influence other failure mechanisms or similar ones in different ways / to a different extent?
It sometimes astonishes me how companies rush to implement large testing plans that actually provide little insight, instead of investing much less resources in serious thinking and orderly engineering / scientific analysis that can result in much less testing and much higher (real) confidence. By this I'm not implying to participants of this discussion, it's a general observation.
Ronen.
While practically true, there are several problems with this approach. For example:
1. They will have to wait for 2 years before they have the final results. Suppose that 2 years shelf-life is a strict design input requirement, and after 22 months they discover that the device is failing - then what? Implement a change and go for another 2 years run?
2. The outcome will probably provide relatively little insight, as it will essentially be a binomial outcome - pass/fail. Sample size will have to be rather big to be statistically valid, and even bigger if they choose to test at interim points (eg every 3 months). If they only test at the end it will provide even less insight and might be too late for an efficient reaction.
3. Is it a representative scenario at all? Have real life transport and storage scenarios been analysed? Will the device actually spend up to 2 years at -10deg C, with minimal variation? Does that correlate with any foreseen failure mechanism in the device? Or maybe it's more likely that the device will experience thermal cycling (the characteristics of which are important), which might influence other failure mechanisms or similar ones in different ways / to a different extent?
It sometimes astonishes me how companies rush to implement large testing plans that actually provide little insight, instead of investing much less resources in serious thinking and orderly engineering / scientific analysis that can result in much less testing and much higher (real) confidence. By this I'm not implying to participants of this discussion, it's a general observation.
Ronen.
Bonebuilder
Involved In Discussions
I'm just looking at the same subject for one of our current projects.
It's interesting to note the accelerated aging test protocol DuPont™ publish in their technical reference for their Tyvek 1073B and 1059B materials which is the go to material for lidding in so many medical devices. Quote:
Samples of DuPont™ Tyvek® 1073B and Tyvek®1059B were aged using the accelerated conditions listed below. Samples were rotated through the following cycle six times, which is equivalent to six years of aging.
After the accelerated aging, the samples were stored for five years at ambient temperature and humidity. These conditions were choosen before the publication of ASTM F1980-07. The results of these tests are shown in Table XII.
Do we think -4°F (-20°C) is sufficient?
Our colleague Remus wants to go down to -10° maybe he's right when you consider that Fairbanks Alaska averages -16.9°F (-27°C) in January with Grand Forks North Dakota coming second at -3.1°F (-19.5°C) Moscow gets down to 10°F (-12°C), Stockholm 23°F (-5°C) and Oslo 20°F (-7°C) so maybe we should be going lower?
Imagine your product sat on the tarmac of Moscow airport having been off loaded from its transport aircraft whilst the handler goes off for a cigarette!
Are they still allowed to have a cigarette these days?
I'd welcome any input to this debate.
It's interesting to note the accelerated aging test protocol DuPont™ publish in their technical reference for their Tyvek 1073B and 1059B materials which is the go to material for lidding in so many medical devices. Quote:
Samples of DuPont™ Tyvek® 1073B and Tyvek®1059B were aged using the accelerated conditions listed below. Samples were rotated through the following cycle six times, which is equivalent to six years of aging.
- Two weeks at 130°F (54°C) with a relative humidity equal to 17%
- Two days at -4°F (-20°C)
- Two weeks at 130°F (54°C) with a relative humidity between 70% and 80%
After the accelerated aging, the samples were stored for five years at ambient temperature and humidity. These conditions were choosen before the publication of ASTM F1980-07. The results of these tests are shown in Table XII.
Do we think -4°F (-20°C) is sufficient?
Our colleague Remus wants to go down to -10° maybe he's right when you consider that Fairbanks Alaska averages -16.9°F (-27°C) in January with Grand Forks North Dakota coming second at -3.1°F (-19.5°C) Moscow gets down to 10°F (-12°C), Stockholm 23°F (-5°C) and Oslo 20°F (-7°C) so maybe we should be going lower?
Imagine your product sat on the tarmac of Moscow airport having been off loaded from its transport aircraft whilst the handler goes off for a cigarette!
Are they still allowed to have a cigarette these days?
I'd welcome any input to this debate.
Last edited:
I'm sorry but I think canned solutions are meaningless here.
DuPont have no choice but to select some generic protocol, hoping it will be relevant for a majority of cases - their products are used in too many different applications and it's not practical to cover the entire range.
Any device manufacturer must consider their own specifics - device design and construction (including packaging), characteristic failure mechanisms and actual shipping, storage and use conditions.
DuPont have no choice but to select some generic protocol, hoping it will be relevant for a majority of cases - their products are used in too many different applications and it's not practical to cover the entire range.
Any device manufacturer must consider their own specifics - device design and construction (including packaging), characteristic failure mechanisms and actual shipping, storage and use conditions.
marmotte
Diapason Consulting
Dear all,
Like often here, I'm piggybacking an existing thread
.
So just to clarify on what have been said above - please correct me if I am wrong :
Reason I'm asking:
Like OP my product (IIa, non sterile) is labelled with a range for temp and humidity. We have been historically claiming a five years shelf life. The documented evidence, which postdates the claim, is a test report we put together after comparing retention samples and new products with an in-house protocol.
I need now to change supplier for the material of a non-critical part of the product ; new supplier can provide BioC etc... + letter stating they have validated their 2 years shelf life.
I would like to avoid having to decrease the labelled shelf life claim for 2 reasons:
The new supplier cannot provide 5 years old samples which I could compare to new parts. But if they can provide say 3 years old samples, kept in uncontrolled conditions, would it be acceptable to submit these samples to a "2years scope accelerated aging" then compare with new samples and claim 5 years?
Is it acceptable to follow the ASTM recommendations with using 25C "storage temp" even though the label says -10 to 40?
Does somebody around have any recommendation regarding the best aging protocol (cycles?) for polymer foams and fabric laminates?
I have spent a few hours today reading several threads here and elsewhere about this topic. Altogether, it seems to me that regulators have been in the past couple of years raising up the steam on this subject - very often completely overlooking the "commensurate to the risk" approach. I have seen countless stories where FDA or NBs were asking for documented test reports after refusing written rationale in spite of an obvious very low risk
Like often here, I'm piggybacking an existing thread
.So just to clarify on what have been said above - please correct me if I am wrong :
- ASTM F1980 is a "one size fit all" accepted by the regulators - but it is not on the harmonized list?
- There is no other standard (harmonized or not) besides of ISO11607 widely used by the medical device world to specify an accelerated aging protocol.
- Both these standards are dedicated to sterile products yet are used -by lack of an alternative - for non-sterile produy\cts as well.
- None of these standards, and none known standard, does provide a guidance for the parameters to use to specify accelerated aging when the labelled storage conditions are given as a range
Reason I'm asking:
Like OP my product (IIa, non sterile) is labelled with a range for temp and humidity. We have been historically claiming a five years shelf life. The documented evidence, which postdates the claim, is a test report we put together after comparing retention samples and new products with an in-house protocol.
I need now to change supplier for the material of a non-critical part of the product ; new supplier can provide BioC etc... + letter stating they have validated their 2 years shelf life.
I would like to avoid having to decrease the labelled shelf life claim for 2 reasons:
- Inventory / supply management wise - it is much easier for us to manage a longer shelf life
- A change of label would trigger notification/review by our National regulatory body (who is also our NB) and FW to our regisster agents Without this change of shelf life - the change can be considered non significant (same material, just different supplier)
The new supplier cannot provide 5 years old samples which I could compare to new parts. But if they can provide say 3 years old samples, kept in uncontrolled conditions, would it be acceptable to submit these samples to a "2years scope accelerated aging" then compare with new samples and claim 5 years?
Is it acceptable to follow the ASTM recommendations with using 25C "storage temp" even though the label says -10 to 40?
Does somebody around have any recommendation regarding the best aging protocol (cycles?) for polymer foams and fabric laminates?
I have spent a few hours today reading several threads here and elsewhere about this topic. Altogether, it seems to me that regulators have been in the past couple of years raising up the steam on this subject - very often completely overlooking the "commensurate to the risk" approach. I have seen countless stories where FDA or NBs were asking for documented test reports after refusing written rationale in spite of an obvious very low risk