Chain of Custody / 483

Chain of Custody is the documentation of your samples lifecycle history. Most people don’t realize that Chain of Custody starts very early in the manufacturing process and continue through the testing phase. All stages of a sample’s life cycle have some sort of chain of custody or an inventory process involved. These processes are all critical and accuracy is essential. They are a cGMP process and need to be documented at the time work being done.

Related FDA483/Warning Letter

The dispensing record of the [REDACTED] exhibit batch [REDACTED] shows [REDACTED] raw material lot [REDACTED] to be weighed by [REDACTED] 08/31/05 and checked by [REDACTED] one day before, 08/30/05.
www.fda.gov

Beginning in the manufacturing process, the API weigh records documents how much of the API was used in the batch. These quantities can be calculated to a theoretical yield quantity. The batch records themselves document the entire manufacturing process, from beginning to end. In it, there are records of the in-process testing and those results, stability sampling, reject and the final yield. The final yield is the quantity entered into the For Sale inventory.

The sampling of a batch for stability samples needs to be collected across the manufacturing process and not from one general pull. The pulling process should be a random pattern with samplings in a Beginning/Middle/End of batch configuration. This collection pattern has the potential to identify quality issues that may be missed if collected from only one mass sampling. Samples should be labeled as to their stage of manufacturing during the pulling process and documented during the chain of custody process of moving the samples from production to stability.

This collection pattern can bring questions from auditors for the difference or uniformity of test results for the different stages. The tricky part is how you defend the pattern. Do you pull samples for each section and perform triple testing and increase your sample and testing costs? Or do you set a specific pull pattern from the different stage that involves each level and increases your stability pull time? Or do you take the time to mix the incoming samples randomly and pull samples and test? The first option has the potential for questions on the robustness of your testing program, if there is variability in the sample results. The second option leave questions on the process used to determine what samples are pulled at what interval. The third option, the easiest of all, has its drawbacks, as it does not allow an investigation of process failure, as you do not know where in the manufacturing the sample came from.

The first two options require the supporting documentation of sample deduction from each group. They increase the complexity and increase the documentation requirements.

Your stability program was found questionable. It was noted that the records of an employee documenting samples pull out dates, for different stability studies of your drug products, when in fact the employee was not physically onsite on the day of the event. Furthermore, the employees accepting the samples to be analyzed cannot confirm the accuracy of the documented date.

When it comes to documentation of the Chain of Custody for a sample, there are different ways to record it. Below is a basic pattern for recording the related data.

chain-of-custody.jpg

Each of the above stages is a critical part of your samples lifecycle. This documentation helps in exception investigations, because it documents the time period between pull and receipt at the lab before testing. The documentation ties together the entire movement process.

• The Pull Date should be the actual date the samples are removed from the storage chamber. This date should not be confused with the scheduled pull date, which is the calculated date from the stability initiation that testing is supposed to be started. Due to workload, weekends or holidays this date may be different from the scheduled pull date.
• The Shipping Date may be different than the pull date depending on the sample handling policy. This date identifies the date the samples are removed from the chamber and moved to the labs. When shipping samples to an outside lab, this documentation should include the following: date, tracking number, Temperature Monitoring Device (TMD) serial number used. The history copy of the TMD data should be filed and retained per the records retention policy. This data is a critical piece of information when investigating an out of spec result. The TMD data should support the movement of samples and include a summary of the shipment. It should include a reference to the sample, TMD serial number, the start, stop of data recording and duration. A print out of extremes seen and receipt. Outages need to be accessed and a discussion of impact to the sample needs to be documented.
• The Date Received in the lab needs to be documented. This date verifies when the samples moved from the chamber area to a less regimented storage area in the lab. The labs need to be proactive and record the receipt on the day of arrival, and not at a later time where there may be a discrepancy. Auditors are looking at this requirement more closely and calculating compliance to records requirements. This date is used in the testing turnaround to calculate the compliance to the testing policy.

When documenting all the processes above, it can be done manually or electronically. Each step should have a second check in the respective areas to verify the quantities, calculations and dates. Each stability study should have an inventory section and the deductions need to be accurate and verifiable against the number of samples physically in the chambers. In the electronic records system, the process should be a validated and link the test and sample.

Related FDA483/Warning Letter

A Laboratory technician in charge of pulling samples from the stability chambers including the vault used for [REDACTED] noted backdating on different stability protocols and verbally confirmed by stating: “it is hard to get two peoples to open the vault “will fill in the date if left blank, at next time point, based on what the pull date is scheduled for.”
www.fda.gov

An overlooked requirement in the chain of custody process is the Inventory justification process. There should be a discussion point in the Stability SOP that describes the inventory justification process. When and how often

Leave a Reply

Your email address will not be published. Required fields are marked *