Remapping frequency is actually a question about risk. As a rule of thumb, I start with a period of every 3 years, then adjust up or down based on risk and product value. If the unit is a refrigerator with a good history storing relatively cheap stuff like lab reagents, then I would move the frequency out to 5 years. If the unit is a stability chamber with tight specifications that performs critical tests that could delay production, I would move the frequency to once a year. I know some risk averse companies that remap everything every year. Whatever approach you choose, be sure to document your rationale and remapping frequency in an approved policy, procedure, or specification.
I personally recommend against this as a general practice. Measuring the air temperature is the best practice; it represents a worst-case condition because product will experience less temperature fluctuation than the surrounding air. Usually, folks only ask this question when their environment is failing to pass validation with sensors in air. Always try to perform temperature mapping with your sensors in the air first. You can map with sensors in some sort of buffer material. However, you must clearly document the material and volume used and why you considered this acceptable. Use as small a buffer as possible, because this amount of material will now become part of your procedural controls for the mapped unit. You will need to control that unit to ensure it is not storing anything with a volume smaller than what you are using to buffer the sensor.
There is no clear regulatory guideline on an exact sampling frequency to use. I have never performed a mapping with a frequency longer than 15 minutes as an upper limit. A frequency of one minute would generate too much data and provide little added benefit. For these reasons, I would recommend a sampling frequency of 5 to 10 minutes. This provides a balance between collecting too much data while still providing a high enough data resolution to capture the thermal behavior of the environment. Please note that these sample frequencies are based on the assumption that environment is being mapped for at least 48 hours.
I had a question following the mapping webinar in regards to GMP warehouse mapping. We mapped our warehouse following summer / winter mapping profile for original qualification and I was wondering is there a guidance / requirement to continue with summer / winter for re-qualifications?
Currently, we have a re-qualification, mainly empty chamber for Controlled Temperature Unit (CTU), procedure that we execute every three years since the original qualification. The warehouse was not originally captured in this procedure, but we did perform a single temperature mapping (not tied to summer / winter) in order to confirm that the space was still under acceptable control. Again, just asking from your experience: is this something that is detailed, as continued summer / winter mapping should occur for re-qualification studies?
First, full marks for having mapped your warehouse at both seasonal extremes. While this is what we are supposed to do, in my experience, your attention to mapping your warehouse correctly puts you at the top of the class.
To hear that after such a comprehensive validation effort you then re-mapped is impressive.
Regarding the regulations and guidance, I think you can interpret the current documents to stipulate performing a full seasonal mapping for requalification. I think you could make a risk-based argument to do a simpler mapping on the requalification (knowing it might lead to a full seasonal remapping). But the key point is: if you want to take the risk-based approach, you must document that this is your objective and create acceptance criteria for the single re-mapping so that you can use that data as a rationale for either performing, or not performing, a full seasonal remapping. As long as it is written down, is justified by data, and you follow it, you have shown the necessary diligence.
If I have only one EMS sensor in each piece of equipment, how do I know what is the best location to place it?
Some would say that this is why we map the chamber; to select a representative location for monitoring. However, in practice, most facilities just attach the sensor to the wall of the unit, on the same side as the door handle, maybe 20 to 30 cm away from the door.
Now, I have a lot more to say on this topic of using mapping study data to determine monitoring sensor location. So, in our next blog, I will write more on this topic, sharing what I learned in validation work, and in seeing many facilities and applications over the past two decades. Stay tuned!