The things labs test and calibrate don’t magically appear out of nowhere. Somewhere along the line, there is a sample taken.
Many labs aren’t in charge of sampling. That’s something left to clients, clinical staff and production line staff. Just because it’s somebody else’s responsibility does not mean laboratories can ignore it.
In fact, even though you answer “Not applicable” in your Assessment Information Document for NATA, the lab is not absolved from answering questions on the topic of sampling.
Let’s start with what ISO 17025 has to say.
The ISO 17025 Framework
While it does not provide specific sampling procedures, it does require that laboratories have a documented sampling plan and follow recognised sampling methods appropriate for their specific testing and calibration activities.
ISO 17025 requires laboratories to establish a sampling plan that defines how samples will be collected, identified, and handled. The aspect of sample handling often falls within the spectre of labs because samples that have not been handled correctly might mean that they are rejected once they reach the lab. That means, all labs essentially hve an active role in sampling even if the actual collection of the sample is not performed by laboratory staff.
The plan should be based on recognized and validated methods. There are often standards covering sampling. A quick check of a standards writing body such Standards Australia or a sector specific association shouold reveal these types of methods.
Depending on why the testing or calibration is being performed, there could be some merit in confirming with clients what sampling strategy has been used. If the items are from a production process, for instance, what is the statistical basis underpinning the sampling plan?
Sure, you can have that little disclaimer of “The results relate only to the samples tested”. But is that really what the people paying for the lab’s service are expecting? Be bold and show your clients you care about these kind of things!
Laboratories must use appropriate and validated sampling procedures. These procedures should consider factors such as the characteristics of the material being sampled, the purpose of the analysis, and any relevant regulations or standards.
This is where labs can really assist clients, especially those who come to the lab withut any significant scientific background. If you’re in an environmental testing lab, how many times have you received a water sample in a Coke bottle with instructions to test the water for pesticides from the farmer worried about spray drift into his dam?
Sampling is frequently used because gathering data on every member of a target population or every product produced by an organisation is often impossible, impractical, or too expensive. Sampling lets you draw conclusions or make inferences about the population or product lot from which the sample is drawn. But there is an art to ensuring the samples adequately represent the whole population.
Before anything happens to collect the sample, the population itself should be reasonably well defined. What are we trying to measure? It’s no good testing an ear swab for microorganisms when the patient is suffering from an infected leg! Nor is it appropriate to calibrate over two points at one end of an item’s whole working range if that fails to demonstrate the performance of the equipment.
The samples collected must be representative of the material being tested. This means that the sampling process should minimize bias and ensure that the samples accurately reflect the properties of the larger population from which they are taken.
Many organisations use statistics to sensibly approach the task of sampling. And there will be some variation and uncertainty associated with the act of sampling. We have a course on MU in Sampling that we run periodically. If you’d like to understand more about this topic, drop us a message to join the wait list.
Chain of Custody
Laboratories are required to establish a chain of custody process for samples. It need not be an official chain of custody with tamper-proof seals and signatures at every stage. What this process ensures is that the samples are properly tracked from the time of collection to the time of analysis to maintain their integrity and prevent contamination or tampering.
Once the sample hits the lab, the chain of custody and identification process is more within the control of the lab. Laboratories must establish procedures to ensure that samples are uniquely identified as a component of this chain of custody process.
Each sample should be assigned a unique identifier, and this identifier should be recorded and maintained throughout the entire testing or calibration process.
Typically we give some sort of sample number when there are lots of samples or use a serial number for pieces of equipment. The system for sample identification does not need to be super complex with secret codes using Julian calendar dates. Keep it simple so you can easily find samples and their associated records.
All sampling and sample management activities must be well-documented. This includes recording details such as the sampling location, date and time of sampling, person responsible for sampling, by whom and when samples were received in the laboratory, and any relevant environmental conditions.
Validation and Verification
There’s no getting away from the topic of validation and verification. Any sampling methods used should be validated and verified to ensure their suitability for the specific testing or calibration activities. You can refresh your memory on Validation and Verification with our recent article, Unveiling the importance of method validation.
Personnel involved in sampling must be trained and competent in the sampling methods they use. Training records should be maintained. Likewise if there is something difficult with sample management processes in the lab, training will be important.
Laboratories must implement quality control measures for sampling, including the use of appropriate equipment, calibration of sampling tools, and the use of suitable containers for sample storage.
QC processes for sample management within the lab should also be established. That could range from double checking sample information against a test request form, through to monitoring temperatures of fridges. The trick is to think about the critical points in the process and put in checks at those points
Storage and Preservation
Samples should be properly stored and preserved to prevent degradation or contamination until they are analysed. Storage conditions should be in accordance with recognised standards or procedures.
Laboratories must have procedures for proper sample handling to prevent contamination, degradation, or alteration. This includes guidelines for the appropriate storage conditions, container types, and handling protocols specific to the nature of the samples being analysed.
Facilities for sample storage should provide adequate protection against environmental factors that could affect sample integrity, such as temperature, humidity, and security.
While we’re on the topic, it’s important to think carefully and broadly about the integrity of the sample and preserving that integrity as a whole. This is not just whether the sample has been frozen and in the correct bottle. It might also cover how representative the sub-sample used for testing is of a lake, a production batch, or a person. The considerations about having a representative sample discussed above are just as important at this stage as in the prior stages.
Procedures for the safe and secure transportation of samples from the sampling site to the laboratory should be established. If you’re expecting highway robbers, then extra security might be in order. If not, traditional means of protection should be OK.
Laboratories are required to define and document their policies regarding the retention of samples. This includes determining how long samples will be stored after analysis, as well as procedures for the disposal of samples once retention periods expire.
In determining the sample retention period, think about how long the sample maintains its integrity with respect to the analyses and how long the customer might mandate for sample retention. Often there is a disconnect between these two periods and you should have a discussion with the client if that’s the case.
Laboratories should have contingency plans in case of sample storage or handling incidents that could compromise sample integrity. These plans should include actions to be taken in case of emergencies, such as equipment failures or power outages. To develop your contingency plan, start by looking at the risks in the sampling and sample management process. We’ve written previously on Risk Management and our tips might help you with this.
That’s a lot for a lab to deal with!
The reality is that most laboratories get limited information on samples or items for testing or calibration. We’re encouraged to give unique sample numbers but this does not mean lab people should fail to appreciate what a sample represents.
It’s all too easy for lab work to be isolated from the outside world. And the pressure of production of numbers makes it more difficult to think beyond the walls of the lab.
Before undertaking the expensive testing or calibration activity your lab has been commissioned to complete, if something doesn’t add up between all the things you know the client is trying to achieve and the nature of the sample received, now is the time to speak up! Your client might just have made a mistake and will thank you for catching it.
If you feel a little overwhelmed with how to wrangle samples within your lab, then get in touch. We can cut through the fog to give you clarity to ensure both you and your customers are delighted with the service you deliver to them.