Vaisala Senior GxP Regulatory Expert Paul Daniel often receives questions about validation. Most questions are common to others in GxP-regulated industry, and fairly easy to answer with materials that are already published in previous blogs and articles. But, occasionally we receive a question that is entirely new.

In this blog, Paul answers a question he's never received before. To ensure his answer is reasonable, he contacted a fellow validation professional with decades of experience in life science validation. Sometimes, two experts are better than one - especially if the question is exceptional. 

With that in mind, we invite you to the webinar "Validation: Ask me Anything." Paul is joined by Josh Keenan by Axis Consulting. Together, they will attempt to answer all questions about validation for GxP compliance. Learn more and register...

M. wrote:

Good day Paul,
 
I was recently watching one of your webinars and I have a question related to the effect of seasonal changes on temperature-controlled spaces. In our company we have conducted Summer and Winter temperature mapping studies in warehouses according to the WHO model guidance for temperature mapping of storage areas for time- and temperature-sensitive pharmaceutical products. From these temperature mapping activities, we have found that we can best control the room temperature when using a temperature set point of 72 °F (22.2°C) during summer conditions and 73 °F (22.8°C) during winter conditions. Normally, our temperature-sensitive storage locations are controlled to one validated temperature set point, so I wanted to see if you had any thoughts on whether it is appropriate to have two set points, one for winter and one for summer conditions. If so, should we proceduralize a switch between the two set points in the shoulder months we make the switch, April and October. I appreciate any insight you can provide on this matter.
 
Thanks,
M

Paul wrote: 

Thanks for reaching out to me! I think I can answer, but I’d also like to check in with a colleague with whom I worked on editing the latest revision of the ISPE’s Good Practice Guide for Controlled Temperature Chambers. Hopefully the both of us can more fully answer your question. 
 
For my part, in an ideal world, there would always be a single, year-round set point. As you’ve apparently found, the real world is more complex. I’m wondering if you were able to determine if a set point of 72°F in the winter was too cold?  Or, if a set point of 73°F in the summer was too hot?  I’m guessing you got no traction in this regard, or we wouldn’t be talking.  Can you find a compromise at 72.5?  I mean, 1 degree Fahrenheit, is a half degree Celsius.  That’s really such a minor difference, it’s probably less than the instrument error for the control sensors.  
 
Did you have your HVAC specialists see if they could adjust the HVAC system so that you could operate with a single set point?  You have probably done this, but just checking in…
 
I haven’t experienced a situation where a warehouse was operated with seasonal changes in set point.   But that doesn’t mean it’s not feasible. I have however, seen a company use incubators in this way.  Not with seasonal changes to set points, but with frequent changes in set point based on product specifications.  To support the change in set points, they mapped each unit at each set point. They sent their concerns to Vaisala when it came to calibrating their temperature probes.  As you would expect, it required an unusually wide range to accommodate their set points.
 
But back to your warehouse; in my opinion, proceduralizing the switch-over to a seasonal set point would be the way to go.   My instinct says to link this switch to some other event that is already scheduled, such as PM or a calibration.  Maybe you could do it as part of a change control, to make sure it is properly tracked? 
 
My instinct to proceduralize the switch stems from my experience with warehouse mapping. Changing set points seasonally is unusual and unusual things tend to provoke questions from auditors. A procedure in place is your best preparation for questions during an audit.  

Let’s check in with my colleague and fellow editor of the ISPE good practice guide…D. 

D wrote: 

HI Paul and M.,

Thanks for including me in this discussion. First, Paul and I agree that there typically is no need to  change a set point  in a warehouse and he brought up several good points. In my experience, I have not had to change chamber settings for winter or summer applications. But for full disclosure, I have only managed a few chambers that were located outside and were used to store GMP products and so did not need to be adjusted.

Here are a few things that I would consider: 

•    What is the test acceptance criteria? Per USP it could be 15°C to 30°C or if it’s for a drug stability situation it could be ±2°C. The reason I ask is: does one degree set point change keep it from going out of specification as per test acceptance criteria, or are you just noticing that the trend looks different and want to compensate to maximize the allowable temperature alarm limits? 

•    Where are the probes located? If they are too close to the wall or ceiling, this could cause a difference in readings. The thermal break location is going to move in the wall and if the sensors are too close, it could be impacting the reading.

As mentioned, 1 degree of temperature is small. If we consider the measurement uncertainty of the sensors being used, does the total readable temperature still come out to 1°C?

This is an interesting question, so I called a friend of mine who owns and operates a chamber manufacturing company and presented this problem to him. He said that he has not experienced this issue. He builds his chambers to meet GMP and other regulatory requirements and his thoughts were the same as mine; it could be a number of issues.

1.    Sensor placement
2.    Air supply source
3.    Air velocity
4.    Chamber load

All this being said, if you must change the set point per the seasons, my concern matches Paul’s:  from an audit standpoint, any activity that is uncommon in GxP-related applications can get tricky. This is what I recommend: 

a.    Determine when you will temperature map and at what settings. Some in the auditing community feel that if you change the set point, or any other setting for that matter, you must remap the chamber. This might depend on how much of a change will be allowed. Some say that changing the set point one degree won’t make that much of a difference. This would depend on how big a span your test acceptance limits are. 

I worked at a major pharmaceutical manufacturer for almost 30 years and was the primary audit contact for Development Stability, Reference Standards and Laboratory equipment. If I had to change the condition of a chamber, I would map it before the change so I could “bracket” testing from start of that condition to finish for that condition and then change the set point or any other adjustments I would need to make. Then I would map it again. If the 1st test failed it would require me to submit a deviation to prove the samples were safe. If the second test failed…. Actually, it couldn’t fail because you are allowed to do whatever you want to the chamber before starting the test. I would even collect data while testing the chamber.

b.    If you want to go back and forth between set points without mapping each time, you might be able to get by with mapping the chamber at the current temperature, and then waiting until the season changed and then map again. My problem with this is, how do you know how cold or hot it’s going to get and how would that affect the temp in the chamber?

c.    Continuous mapping: I haven’t tried this before, but if you are mapping continuously and all sensors were alarmed, I think you could justify any change to settings as you would always know if you within limits. Vaisala has a webinar on continuous mapping where Paul explains how this works.

In my experience, auditors focus mostly on whether you are using the chamber according to the chamber manufacturers’ recommendations and that products and samples are stored in the required temperatures. Be sure that all sensors are calibrated and documented.

Thank you again Paul for including me on this – it is a tough question!

Best regards,
D.

M. replied: 

Guys, I really appreciate your detailed responses.  To be a bit more specific, we are doing qualification-level temperature mapping in our large warehouses for controlled storage 20-25°C rather than intermediate-use chambers.  

We do allow excursions per the USP guidelines, but in general, we keep all sensors (every data point) between 20-25°C during the week-long summer and winter mapping studies.  We allow for explainable excursions (e.g., a dock door opened), or other short excursions of less than an hour.

The problem we have encountered is that some of our storage spaces have a few sensors that consistently dip below 20°C in winter or above 25°C in the summer. Historically our facilities group has simply moved the thermostat seasonally so that we run a little cooler in the summer and hotter in the winter (within the range) to avoid having hot/cold spots that fluctuate out of range. 

We have an older warehouse and air handling equipment. As you suggest, we have been requiring them to lock down the set points in order to consider it qualified and not require a remapping  for three years.  When they change the set point, I have been requiring a new mapping study.   

We try to avoid the magical 25°C at any data point to align with  USP <659> Packaging and storage requirements and we need to do a mean kinetic temperature calculation, which doesn’t have any clear set of instructions or acceptance criteria.    

After this discussion with you both, I am going to continue to push for single set points and improved airflow or insulation to mitigate the seasonal fluctuations.  

Thanks again!
M.