Out-of-spec sensors, incubator hot/cold spots & 21 CFR Part 11 compliance

Edited transcript below

In this video interview, Vaisala Senior Regulatory Expert answers three questions from customers and industry contacts. If you have follow-up questions, please feel free to comment on this page or email us.

[00:11:01] Welcome to another Vaisala video blog on continuous monitoring and mapping systems. My name is Janice Bennett-Livingston. I'm the life science global marketing manager and I'm joined once again by Paul Daniel, our senior GxP regulatory expert. Let's jump right into questions. We have three good ones today.

[00:11:30] Question one: “When temperature sensors are out of operating range during routine monitoring, what checks should we perform?”

Paul: This can happen. Sensors are going to go out-of-spec.  Usually it can be chalked up to an obvious source: doors left open, power failures, things like that. But it's going to happen. Usually a good specification or policy will include not only the unit specifications, say 2 - 8°C for a refrigerator, but specs will also include the time and temperature range for acceptable excursions. That's the important part – the excursions. For instance, we might establish the refrigerator can be out-of-spec for a maximum of 30 minutes as long as the monitoring temperature probe stays between zero and 15°C. And an excursion of this sort can happen a maximum of once every 24 hours. So 30 minutes, once every 24 hours, between zero and 15 for that refrigerator.

But that's just an example. It means that if we have an out-of-spec condition that's within the excursion policy then there are no official quality actions we need to take as follow up. We still need to respond to the condition, which we probably will find out about because of an alarm from our monitoring system. But because the excursion is within our allowable limits, there are no quality actions we need to take.

Now, if the excursion limits are exceeded, then we need to do something because we went beyond our temperature and time limits. Now we need to do something official. The specific action we would take would be again determined by the monitoring procedure or policy, but it would likely involve some deviation or out-of-spec report so that we could evaluate the quality impact of the excursion on the products stored inside the affected chamber.

There should also be a periodic review of the performance of that chamber.  If for some reason we're seeing an unusual amount of excursions from a unit, then we want to investigate to determine the cause. It could be that preventive maintenance is required. Maybe there's a bad door latch, or maybe it's just poorly trained users who aren't using  closing the door correctly. It sounds funny, but it's happened before.

[00:13:48] Next question is on incubator monitoring:
“If we find two hot spots or cold spot locations in incubators, which one do we have to consider? Do we need to monitor both the locations during routine everyday monitoring?”

Paul: This question is one of my big pet peeves. The real answer here is this: You only need one monitoring probe for an incubator. And you don't stick in either the hot spot or the cold spot. You stick it on the wall, on the side, on the same side as the door handle. You place it far enough back so the door openings don't have a big impact on measurements. This is where almost every monitoring probe is located in almost every refrigerator, freezer, incubator, and stability chamber.

Honestly, I rarely see anything other than this because it's simple, repeatable, durable and it gives us what we need to monitor these chambers and determine if they've suffered catastrophic conditions. For instance, a power failure, a door left open or some other major failure.

Now, until the Good Distribution Practice revolution of the last eight or 10 years, no guidance even mentioned determining hot and cold spots. The guidances certainly didn't state to place a sensor in these locations for monitoring. The GDP revolution happened and our regulatory authorities needed to make sure people were trained. These were people working in warehouses and pharmacies who were not trained how to map and monitor, often beginners who were unsophisticated in validation processes.

Regulators had to make it simple and train these people, but they overemphasized the importance of the hot spot and the cold spot. Unfortunately, they shifted away from simply recommending that these hot and cold spot locations be used to inform our placement or our choices. They made the error of stating that monitoring sensors should be placed at a hot spot and the cold spot locations. You think about it, you'll realize it's actually ridiculous. Let me tell you why: say you have an incubator that's just a few cubic feet and you find the hot spot is in the very center and a cold spot in the corner. So now we need to put two probes in the unit and one of the probes is going to be mounted in the exact center of the usable space. This is exactly the spot where everybody wants to put their petri dish. So now this probe is going to have to be secured in place to make sure it doesn't get moved and we need to maybe put a little cage around it to make sure it doesn't get damaged. Now you've just used up the center of the space.

This is preposterous when you think about it, which is why no one does it. I've only seen one facility ever where they put two probes in each chamber to monitor the hot spots in the cold spots. And they complained about these same problems. So why is there this idea that we should do this? I think it comes from a misunderstood historical practice. Before the widespread use of computers to assist in temperature mapping, we had to calculate statistics manually, with a calculator, and with our eyes finding the highest values in a printout. That was the easiest way to figure out, with this pile of data, if you passed the test, was to figure out what the maximum temperature was and what the minimum temperature was; everything else was either lower or higher, or not as extreme. You just needed to know the maximum and minimum temperatures and those are the same thing as a hot spot or cold spot.

So the max/min temperatures were important analytically because with them, you could quickly determine if you passed the test. It's not that the hot spots and cold spots themselves mattered. They don't matter. It was just the easiest way to manually analyze the data. So I'm going to repeat this. Hot spots and cold spots don't matter if all your data is within spec and therefore within your acceptance criteria.

We did this mapping, this validation, to prove that the chamber can maintain the correct temperatures. So if we did that right, we can trust that work. Monitoring is in place to catch events that can't be predicted by the mapping study, I.E. doors left open, power failures, etc.. Monitoring is not done to check the hot spots and cold spots for tiny deviations. That's why we mapped. To wrap up a fairly long rant about a pet peeve of mine, I don't think you need to monitor hot spots and cold spots in an incubator. If you're dealing with a warehouse, totally different question, but definitely not for an incubator.

Video on wireless monitoring in CO₂ incubators

[00:18:28] Ok, thank you for that. To reward you for that extensive answer, I have another pet peeve question. This is a question we often receive:

“Is the viewLinc continuous monitoring system compliant with 21 CFR Part 11?”

Paul: Any time a customer asks a question like this, how viewLinc complies with 21 CFR Part 11, Annex 11, PIC/S or anything like that, I see it as an opportunity to educate. The most truthful answer to any question like this is: viewLinc doesn’t comply with 21 CFR Part 11 or any other regulation.  I know that sounds odd, so let me explain.

These regulations, 21 CFR Part 11 from the FDA and the GxP regulations from the European Medicines Agency, they apply to companies that make and sell pharmaceuticals, medical devices, cosmetics and other GxP-regulated products. Now Vaisala, we make software and measurement devices. So these regulations simply don't apply to us and by extension, they don't apply to our products like viewLinc. The GxP regulations apply to our customers; the people who are using our viewLinc system and our sensors to make drugs, medical devices and other pharmaceutical products.

Here’s an analogy: it’s like asking Mercedes Benz if their automobiles comply with traffic laws of the USA.  The USA has some national safety standards for car manufacturers, for instance minimum braking distance, seatbelts, airbags, etc. But traffic laws apply to the drivers of the cars, not the automakers, not to Mercedes, not to Chevrolet. Traffic laws are a different set of laws altogether. Now, Mercedes may be required by law to have a working brake pedal on the vehicle. It's up to the driver to obey the traffic laws and use that brake pedal to actually stop the car at the traffic light.

Sometimes customers fail to understand this distinction when they ask if the viewLinc system is compliant with drug manufacturing regulations. I just don't like to say, yes, our system complies with 21 CFR Part 11 because it's not exactly true. If I say that, then I'm saying we don't really understand this distinction between who 21 CFR Part 11 is intended to regulate.

What is the real question our customers are trying to ask? I think they are asking: “If I use viewLinc to monitor my drug products, will I be in compliance with the relevant regulations that govern my work?” Again the answer is that it depends on how you use it. Much like the analogy with the car, the driver must comply with the traffic laws by using the brakes, accelerator and steering wheel in a compliant manner.

My best answer to this question of viewLinc’s compliance is: “viewLinc is designed to help you achieve compliance to the regulations.

If you read any part of the FDA  Part 11, or the European regulations, or PIC/S, 90% of these regulations is about adequate written procedures and adequate training records to ensure people are trained on procedures. Regulators want to see documented proof that procedures were followed. It's all about procedures.

In some parts of the regulations, they call out some specifics like audit trails. But clearly, more than 90% of compliance with regulations is about how you use a system, whether you have the appropriate SOPs (standard operating procedures) in place, and whether or not you follow those SOPs.  So the best answer I have for you is that viewLinc is designed for compliance with GxP regulations. You won't find me saying that it's complying with 21 CFR Part 11, but I will happily tell you many times over that viewLinc is designed so that you achieve compliance with the system.

[00:22:40] We have a white paper on functions within the viewLinc software that help you comply with 21 CFR Part 11 and Annex 11. We have a similar paper on the vLog mapping system. Thank you everyone for watching. And we'll see you again—Stay well!