Using Mapping Study Data to Select Monitoring Sensor Placement
To be honest, I don’t recommend using the mapping data to select the location of your monitoring sensors. The entire idea is counter-intuitive. If you just performed a mapping qualification to prove the ability of the unit to maintain the stable environmental conditions specified in the acceptance criteria, then you have proven that every spot in the unit is suitable for product storage. Since every spot mapped met the acceptance criteria, then all spots mapped are all equivalent and no spot represents a greater risk than any other spot.
Sure, we have always located hot and cold spots during mapping qualifications, but this was simply an artifact of the statistical analysis. The easiest way to prove that the mapping met the acceptance criteria was to find the hottest spot and the coldest spots. If those extreme spots were within specifications, then the entire unit was within specifications. Hot and cold spot identification is a result of a common sense approach to use simple max and min statistics to prove the acceptance criteria were met.
As regulators were creating Good Distribution Practices, someone had the mistaken idea that hot and cold spots were good locations for monitoring sensors. This mistake has made its way into guidance.
Let me unpack this notion. Monitoring sensors are not meant to monitor hot and cold spots. Monitoring sensors are properly used to detect catastrophic failures: power outages, failed cooling systems, and open doors, etc. Placing monitoring sensors at the hot and cold spots only means one thing: you don’t have faith in your temperature mapping and you don’t have faith in your unit. If you have so little faith in your unit that you need to monitor the extreme locations, the safest action is to replace the unit!
So, what’s happening in the industry? In over twenty years of this work, I have never seen an incubator, refrigerator, or freezer with more than one monitoring sensor. And almost every monitoring sensor I have ever seen in an incubator, refrigerator, or freezer was mounted on the side wall, at middle height, about 10 to 15cm from the front edge. The only realistic conclusion is that very few people are actually using the hot and cold spots to decide the location of the monitoring sensor in smaller controlled environments.
In larger environments, such as cold-rooms, walk-in freezers and warehouse, the expectation is to have more than one monitoring sensor. However, in these environments, sensors tend to be placed in a geometric pattern that give common sense coverage of the monitored area. I have seen cases in warehouses where monitoring sensors are placed in areas likely to experience fluctuations – such as near loading dock doors. It must be noted that warehouses are far more complex than refrigerators and incubators, with seasonal variations and changing use patterns throughout a given business day. This complexity justifies multiple sensors and some special placements.
It is important to remember that this idea of using hot spots and cold spots to locate monitoring sensors is NEW. The first place I could find it is in 2011 in GUI-0069 from Health Canada stating:
“Refrigerators and freezers used to store drugs should… have sensors for continuous monitoring and alarms located at the points representing the temperature worst case scenarios…”
This idea is new, along with Good Distribution practices.
Before 2011, the acronym GDP meant “Good Documentation Practices”. GDP (as Distribution) was created to expand the umbrella of GxP regulation beyond the manufacturers to include the distributors and pharmacies. Until this point, distributors and pharmacies did not really consider themselves to be under GxP. Even today, most of these entities lack formal validation groups to perform mappings, or the necessary Quality Systems to ensure that set-points are documented and controlled, and remapping occurs whenever significant changes are made to a unit. That is to say, these regulations were intended for groups that were mapping (but not validating and qualifying) environments that would not be controlled within a robust Quality Management System.
When viewing the issue of using mapping data to select monitoring sensor locations within the context of this history (in this case: the creation of a body of regulations for previously unregulated distributors) we can come to the conclusion that these instructions regarding hot spots and cold spots might not be intended to apply to GxP regulated Life Science manufacturers. These GxP companies had already been mapping and monitoring effectively without a specific GDP regulation.
In the end, each company needs to make its own decision about how to place monitoring sensors in controlled environments. This decision should be made using common sense and an appropriate assessment of the risks and the benefits. Personally, I would never recommend that an organization use two monitoring sensors in a small unit unless I was specifically instructed to do so in an audit finding.
I’ll leave you with this quote from the ISPE Good Practice Guide on Controlled Temperature Chamber Mapping and Monitoring:
“Once mapping is complete, the test data should be reviewed to determine the number of permanent monitoring sensors and where they should be located. For small units (less than 2 cubic meters) it may be considered acceptable to use one monitoring sensor located in the worst case location… For larger systems… more sensors may be justified…”
Remember that the ISPE produces some of the most risk-averse and complex guidance in the industry. At the end of the day, even if we are using a refrigerator to store valuable and sensitive pharmaceuticals, it is still just a refrigerator. As long as you have a working, accurate sensor in place with an alarm function, you can get on with your work.
Webinar: Audit-proof your Monitoring System
Check out this webinar to learn how to avoid these common mistakes in environmental monitoring:
- Validation documentation problems
- Outdated or non-existent training records
- No SOPs for environmental monitoring applications
- Past due calibration or missing calibration records
- Change Control procedures not followed
- Security issues – I.E. personnel access to software