blog

# 5 Scientifically Sound Steps to Determining How to Monitor Temperature in your GxP Space

Published: Mar 21, 2013
Life Science

In this week’s blog, Paul Daniel does a hypothetical data analysis with the goal of determining the optimal method of monitoring a cold room.

#### Paul writes:

In our hypothetical situation, your mapping study would have identified the high-risk areas in the area you want to monitor, let’s say, a cold room. So, you now know where your hot and cold spots are. Next, we assume that the entire space, including the hot and cold spots, stayed within specifications during your study. Now we want to be able to monitor it effectively. Here are the ways I would analyze the data to figure that out:

1) The simplest and most obvious first step is to locate the hottest and coldest locations in the room spots. I’d put a monitor sensor at both locations. This is technically a “statistical analysis” because we are looking for maximums and minimums. However, this is expensive (there are two locations) and sometimes logistically difficult because that hot/cold spot might be in an inconvenient, difficult to control location.

2) I pick one spot that is most representative of the entire space. For example, say your cold room specifications are 2 – 8 °C. By analyzing the data, you establish that your representative spot is at 6.5°C when the hot spot hits 8°C. Likewise, you determine that the representative spot is at 3°C when the cold spot hits 2°C. So you create alarm limits for the representative location at 3 and 6.5°C. The downside here is that the most representative location might not be in a location where the temperature is easy to control. Plus, “helpful” proactive people might keep trying to correct your alarm limits of 3 – 6.5°C back to the 2 – 8°C that they “know” the unit is supposed to control at. This certainly depends on the level of quality sophistication at your site, the amount of GMP training provided to facilities employees, and the level of control you have over your monitoring system.

3 - a) Often people do the same analysis as #2, but apply it to a monitoring probe located at the same location as the control probe, or some other standard location, such as just inside the door.

3 - b) Usually though, it seems like most people just put the probe beside the door and monitor at the straight specification limits (2 – 8°C in this case). This seems strange to me because no analysis is done to adjust for the temperature variation in the space. However, the virtue of this approach is simplicity. For some groups, the advantages of simplicity outweigh the disadvantages.

4) If you want extra credit… Take into the account the accuracy error in your monitoring and mapping devices. The combined error could be up to, or even over 1°C, depending on the sensors you used for monitoring and mapping. I know, I know… That just makes it all more complex.

5 - a) One easy way to approach the whole situation is to make the warm spot off limits. If you have the space to spare, just cage it off and find the NEW warm spot, which will hopefully alarm less.

5 - b) OR, you could evaluate the data based on what you actually store in the cold room. If your product is always a paperboard box of 25 5cc vials of “Product X”, you could do a study on product temperature to justify raising the high alarm level to (for example) 10°C, by showing that product temperature remains below 8°C in those situations. Or perhaps you leave the alarm level at 8°C, but justify a delay on the alarm based on your data. However, the best way is always to have air temperature within limits, right.

If you have any questions or points to add on the above steps, please leave your comments below or contact me directly at [email protected]

## Paul Daniel

Senior Regulatory Compliance Expert
Email

Paul Daniel is the Senior Regulatory Compliance Expert at Vaisala. He has worked in the GMP-regulated industries for over 20 years helping manufacturers apply good manufacturing practices in a wide range of qualification projects. His specialties include mapping, monitoring, and computerized systems.

At Vaisala, Paul oversees and guides the validation program for the Vaisala viewLinc environmental monitoring system. He serves as a customer advocate to ensure the viewLinc environmental monitoring system matches the demanding requirements of life science and regulated applications.

Paul also shares his GMP experience through regular blog contributions, webinars, and seminars around the world. Paul’s expertise in the demanding GxP world is applicable to any industry where measurement is critical to product quality. Paul is a graduate of University of California, Berkeley, with a bachelor's degree in biology.