Ensuring the integrity of stability studies is the primary driver for nearly all logistical activities of a stability program. People, programs, and paper hold the links together in the stability chain of custody—a process as critical to study integrity as chamber validation and monitoring, security logs, data verification and documentation of exceptions. But also, a process sometimes as tenuous as links of twine.
The first of many stability-specific links in the chain is when samples arrive at the stability storage location where samples will probably reside in a staging area prior to placement in an environmentally controlled chamber. Staging might be a few hours, days, or weeks, but subsequent links involve the samples in the chamber for months and years with numerous door openings, interval pulls, cleanings, temperature excursions and such, all events to be accounted in the life of the samples that are ultimately pulled, delivered, held in the lab, tested, and reported. Even the final link—sample destruction—must be accounted for to prove material did not end up back on the shelf.
The concept of sample chain of custody is identical to that of data integrity, in fact the renowned acronym ALCOA for data integrity can be applied directly to chain of custody. Here’s what that looks like:
- Attributable: Each stage of sample handling must be attributable to the prior stage and to the process by which the sample arrived at each stage.
- Legible: Documentation of the transfer between stages (and any changes while at each stage) must be readable. If documented in a logbook then obviously it must be legible, but legibility in the chain of custody realm also means it must be easily retrievable from an electronic system, if applicable.
- Contemporaneous: If you wait till the next day to document the transfer of a sample to a new segment in the stability chain it’s like you’ve put an open carabiner in place of a chain link–you can no longer vouch 100% for the integrity of that sample. I recommend the logbook go into the room or at least outside the door for immediate documentation (the same goes for a laptop or some hand-held inventory system device).
- Original: Each person involved in a sample transfer should document their part. Minimally the delivery and receiving persons document their actions directly in the system or the logbook. Any time a transaction requires the recollection of a third party (“I saw them drop it off earlier”) the chain is broken.
- Accurate: You don’t necessarily need minutes and seconds, but at least the date of each transfer. Just as important, though, is the accuracy of place, which stability elevates to a critical factor. We must know if the drop off location was environmentally controlled or at least monitored throughout the time the sample was located there. We’ll discuss more below about what needs to be accurately recorded.
ALCOA in sample integrity applies to the entire chain of custody, but for brevity, from here forward this article will refer primarily to the specific link between manufacturing/packaging and the stability program.
This brings us to the perennial questions of paper logs—are they necessary? For that matter, when logging samples, how much information is needed? Does some documented verification of the entries have to occur? Is the integrity of a study ruined if a sample is not logged contemporaneously or accurately?
The patient is of primary concern, of course, but realistically, our focus is on what an auditor would expect to find. The same goes for logbooks. Without rehashing ALCOA again, on a practical level an auditor will need to see at least the following things:
- Sample ID (if available at this stage). This ID should have all critical information associated to it in your LIMS system, but if no sample ID is assigned, the following will be needed: product ID, batch number, and quantity.
- Differentiation between delivery and receipt with signature and date for each.
- Locations – although this should be important, most auditors will not place a premium on sample delivery/receipt to a specific location. This is just my experience, and it might be because they didn’t detect the stability impact, or it might be that they assumed the delivery location was the same as the location of the sample log. Based on that assumption, while reviewing logbooks auditors have, In my experience, often looked up and asked if that particular location was monitored and controlled.
- Legibility and neatness. A messy logbook begs to be paged through slowly.
That’s all you need to document at each link of sample custody. I’ve intentionally left out a verification line and as noted earlier, delivery time gets more specific than necessary (at least for deliveries to a controlled and monitored location at long term storage conditions). Local procedures may beg for more information, but caution on the side of less information will pay off in the end. First, it will be less tedious and time consuming to log samples, which means it’s more likely to get done. Second, more information means more clutter, more inadvertent missed entries, and more risk of errors for auditors to find. They will be more impressed with neatness than with volume of information if they see that each link of the chain is secure.
But why no verification line? I propose that you combine verification with the receipt process. Place the onus on the delivery person to document sample ID or product information, quantity, and delivery location. In doing so the sample handling procedure can dictate that the role of sample receiver’s signature is to verify the sample delivery information. Voila, built in verification. This is magical since one of the most frequent “gotcha’s” I’ve seen from logbook reviews is missed verifications. Almost definitely, a future step in the stability process will require someone to pull the “Receipt Date” from the logbook which is another built in review that ensures documented receipt and verification.
With that process in place, you can move most of your sample transfers to a paperless inventory system—ideally one housed in your stability application or LIMS. This has the added advantage of an audit trail for additional integrity. Ideally the electronic logbook process would link the chain all the way from the manufacturing/packaging process through stability and on to recording of test results—pie in the sky, maybe, but certainly a nice pie. Without a continuous digital chain of custody like that, though, you will most likely require some form of paper logbooks, unfortunately.
So, what happens when you come in on Monday morning to find a sample outside your office door with a note saying, “Here’s your 2023 Annual Lot. Thanks, Frank.”? Well, Frank gets an earful, that’s for sure and some exception documentation will be needed—assuming Frank failed to follow a procedure. But what about the stability study? I can dream up just two basic options:
- The cleanest solution would be to have Frank bring over a fresh sample and take his old sample away. Yes, have him take it away so you don’t have to manage and document destruction in your shop. In this case, Frank’s area can own the exception document, too.
- Have Frank come back over and fill out the logbook with a note that he actually delivered it on Saturday, but inadvertently forgot to document it that day. The exception documentation is more complicated and will probably need to be owned by the stability group to explain how you know:
- This really is the sample Frank dropped off.
- The sample really was outside your door undisturbed and not tampered with for two days.
- The material was not exposed to environmental extremes while in the hallway.
When dealing with the myriad other impacts to sample integrity, such as shipping and customs delays or misplaced samples, one form or another of these two options will probably come into play.
In conclusion, there is always room for improvement in the stability sample chain of custody. Here are three ideas to consider:
- Revise roles of delivery and receiving personnel to enable built in review. For example, have one person pull samples from chambers and another deliver samples. For smaller programs, one pulls, and another receives. In this way the second person can automatically verify the first person’s task.
- Require that data used for study documentation be pulled from the sample logbook or the relevant inventory system. This ensures that the study creator performs an additional check on custody documentation.
- Critically review the data required in logbooks with an eye for removing information not linked to the sample custody transition. Usually, information can be eliminated that can be found in a LIMS or inventory system.
Tracking of material sent to stability from the manufacturing/packaging floor is as critical as samples sent for testing. Ultimately chain of custody for stability samples has more links than samples for release and in-process testing, and our challenge is keeping those links held together by documented handshakes between people, logbooks, inventory systems and LIMS.