Why laboratories need to thoroughly choose a LIMS
The elusive nature of laboratory quality, likening it to a mythical unicorn. It emphasizes the relative nature of quality and the need to tailor indicators to specific lab contexts. Highlighting the role of Laboratory Information Management Systems (LIMS), it discusses how LIMS can help identify and address errors, ultimately optimizing quality management. The article concludes with steps for customizing LIMS to enhance quality, emphasizing the importance of active involvement, setting criteria, and implementing trigger warnings for error detection.
Everyone talks about quality. Everyone works hard for quality.
However, what exactly constitutes laboratory quality is unclear.
It is like some mythological beast. Like a unicorn. There is constant discussion about it, much is sacrificed in its name, it is endlessly strived for. There are known characteristics and metrics. But no one has ever seen the elusive Quality itself. You only know life is bad without it and good when its around.
You can get the right and perfectly accurate result for a given test. But if you fail to deliver it in time, that result is of poor quality.
You can do a thorough and timely test, but deliver it to the consumer in an incomprehensible way. That has nothing to do with quality.
Similarly, you can deliver a perfect result, but that way is unfamiliar to the clinician, and they make an error while interpreting it. Is this quality or not?
Or maybe the clinician interpreted everything correctly, and the analysis was extremely useful for the patient. But in the process, the lab lost two test tubes. OK for the doctor, but bad for the lab. Quality does not mean reducing revenue.
In the pursuit of Quality, lab people try to capture it piece by piece in the hopes of making a beautiful whole. So how can we get our hands on this evasive Quality?
Clever lab people, working in large groups over many years, created guidelines for quality indicators in the lab. The term Key Quality Indicators emerged. Here you can find about fifty measurements for twenty-six indicators, all phases of the princess included¹. The article outlines all the possible errors, complete with their corresponding units of measurement.
If you ask AI what it thinks about quality, it gives you all the main aspects you need to pursue: accuracy, precision, turnaround time, sample integrity, equipment maintenance and calibration, staff competency, customer satisfaction, error rates, compliance with regulatory standards, and continuous quality improvement.
Well, this is certainly not the wrong answer. But if every lab achieved perfection in each and every aspect of Quality, the choice between them would be moot. Moreover, this perfect lab does not exist. This Shangri-Lab, if you will, remains just out of reach beyond the clouds. Back here in reality, the truth is some labs will be stronger in some aspects of quality as opposed to others. So then, how to choose between labs? You can only look at their results for some characteristics and conclude that for characteristic X, Laboratory A is better than Laboratory B. And for characteristic Y, on the contrary, Lab B is better than A. Which is more important?
If you try to compare labs to award one of them a blue ribbon, you will walk around with that ribbon until it fades.
It turns out that quality is relative. The Unicorn is not the same for everyone.
Take fast results for instance. Of all the metrics, it is one of the most important indicators. After all, a lab must meet deadlines.
For a hospital lab, that is in fact the case. You need to receive a result quickly and treat the patient.
But what if you deal with an outpatient? They will come to the doctor only after three weeks. The sample itself has a shelf life of, say, three days. Here time isn’t as much of a factor.
We don’t always need the same perfection. There are situations in which precision and speed are precious. When the price is a human life, there is absolutely no room for errors. But there are also other situations and other errors which are much less crucial. For example, when you measure glycated hemoglobin and mistake 2 for 3, it is not as terrible as if you’d mistaken 6,4 for 6,7 or vice versa. The latter leads to clinical errors, and the patient will suffer. So, you have different “red lines” on different occasions and measurement sites.
Another example. You can compare laboratories by the number of hemolyzed samples. But if the first lab is hospital-based, the second works with dialysis patients, and the third with outpatients, the amount of hemolysis will vary. More severely afflicted patients are more likely to develop hemolysis, so it’s harder to avoid. It may be that the lab that works with the very severe categories of patients has more hemolysis but copes better than the other two.
Contextual differences between labs are numerous. There’s no way to come up with one-size-fits-all quality indicators. You can tie them to the context of each laboratory. But for each lab, the value of the concrete indicator will be different, and the set of indicators will also be different. So, how could you possibly compare them?
Okay, but are the metrics the same for the same TAT? Again, no. You can score by the average: “There is not a single sample that takes longer than the required time”. Or: “On average, we are producing results 6 hours ahead of schedule”. Or maybe we should divide all samples into centiles: 98% meet the deadline. Which way is the best way to measure?
Each time, you find that quality depends on a specific task.
If you compare two laboratories for practical purposes, you can look at what is important to you and what is not. For example, if it is crucial for you, you look at whether the lab advises clients when test results are inconsistent with the larger clinical picture. If you don’t need that, you don’t look at it.
So, if laboratory A does what you need better, you go to A.
It turns out that it is not the lab that defines quality. The customer, the patient, the regulator, the professional community, the competitors, and the lab all have different, competing points of view.
All this generates numerous sets of characteristics each with various measurements. If only there were a way to organize a World Cup of Quality with every lab competing in order to crown a definitive Champion of Quality, but alas… In reality it’s hard to compare several laboratories on one specific indicator because that metric likely has a different meaning for each of its respective parties. Rather than zeroing in on one subjective metric, try and keep the larger context of your needs in mind.
So it turns out that our Unicorn does not look the same to everyone. What it should look like depends on the specific goal of whoever is trying to measure it.
Let’s look from inside the lab.
You aren’t quite sure what this mythic Quality is, but harnessing it is your first priority. You consider your patients and do not want to compete with other labs on price alone.
To understand the Quality, you look at it from the inside out. If you look at the Key Indicators in the article quoted above, you will see they are in fact error indicators:
“Errors in…”
“Incorrect…”
“Inappropriate…”
Amazing that in describing Quality we actually articulate what hurts it. It’s like the story attributing Einstein’s argument to an atheist professor. The professor said there is no God because He could not create evil. Then the student asked him if darkness existed. The professor said it existed. The student objected that darkness was simply the absence of light. Then he asked the professor if evil existed. Yes, the professor said. The student objected that there was no evil in itself but only the absence of God in one’s heart.
The same is true with quality. It is in fact the absence of errors. It would be nice to find all these errors, which degrade your precious Quality. You see them and compare yourself with other labs.
“Sure, I make 1000000 errors a day. But others make much more, so I’m the champion!”
But measuring errors is so frustrating!
It always seems to be someone else’s error, another mythic trait! And this someone does not want others to see their mistake. Even if they are conscientious, they don’t want their coworkers to blame them.
Not only that, but we’re talking about mistakes made by highly trained people. Errors call their professional credentials into question. That means the likelihood of owning up to a mistake decreases even more.
Moreover, no one is proud of errors because someone’s life may be behind the test tube, which makes people even more emotionally burdened.
We don’t like to report errors, or even look for them.
Again, it’s hard to tell the preventive errors from the non-preventive ones. You can do your best and still have hemolysis when you are working with seriously ill people. So when you tally mistakes you might not be determining your lab’s quality as much as you are measuring the severity of your patients’ condition.
But let’s say you counted the number of errors in our lab, and they were all preventive. Does that make your work better? Not at all! You’re becoming like a codependent AA group. You get together and confess to each other: “Today we botched 10% of the tests, while everyone else maxed out at 2%. We do badly, but we admit our mistakes, and tomorrow we’ll come here again with renewed vigor!”
That’s not enough. If you have identified errors, you must take all measures to correct them.
That’s where the Unicorn comes in. Who? LIMS, of course. The all powerful.
It’s going to help us with two things. Firstly, it will log errors without special effort.
When error logging becomes separated from the individual it has a way of simplifying things. Thanks to the data in LIMS, you can understand who made the mistake. Not in order to scold that person of course, but to know whether we have set the correct processes and tools and trained the employee well. It might be that they need extra help.
Life becomes a little brighter. You don’t have to chastise or scold anyone. Moreover, we can immediately recognize the source of the problem.
For example, many modern biochemistry analyzers detect hemolysis in the assay. LIMS will quickly show you if this hemolysis is due to a particular phlebotomist or delivery route. You immediately calmly offer people a solution and take action. As you continue to record the results, you’ll see if you’ve achieved the desired outcome.
The worst thing is if you’ve figured out how to change your process, but your results have not improved!
For this you need a system that allows you to track all stages of a sample run and all sources of problems at many different stages.
LIMS is a Quality catching machine. If you put the wrong barcode on a sample and you don’t have the same number of samples in the order, LIMS will sound the siren, and you’ll have no chance to overlook that. LIMS solves some of the errors and allows us to fix them quickly.
Do you remember the different costs of an error at different stages of testing? That is why we need triggers that will help us to tell right from wrong in a timely manner.
Some errors are visible on current monitors, especially those related to hardware utilization.
Most modern LIMS have an equipment monitor where you immediately see where you forgot to put the reagent and what is not working. The alarm allows you to correct the situation instantly. We didn’t even notice the error because the reaction was so fast.
Some will say, “If you can correct an error, it is not an error.” That’s not true in the lab, however. An error is still an error, even if we can correct it.
We care about the number of situations we have to handle outside the standard process. The very fact of deviating from the standard process degrades quality and uses up a lot of labor. We need to get results, but the cost matters.
Sometimes an error doesn’t harm the patient. But it does affect the business. Take billing errors, for example. They indirectly affect how much money you spend on quality assurance. It is valuable that LIMS can eliminate the consequences of irregular situations faster. That saves not only time and money but also your reputation.
1. The quality specialist should be involved in the process daily.
2. Set your own acceptance criteria for the desired parameters. You can set parameters based on a wide variety of goals that you often need at the same time:
LIMS manages everything from flows and patient data to your efforts and results related to quality, its criteria, and indicators. But the key is you must enter all the necessary data in order to get a truly useful result. Doing so will unlock the full potential of LIMS: it will measure the efficiency of your system within your lab’s context, finding shortcomings and allowing you to improve your Quality.
LIMS should be such that you can work with the data promptly.
You cannot turn the information system into a black hole, with information lying dead and beyond use. Instead of simply dumping quality management rules into the LIMS, it must be able to read and react to these rules, instantly pointing out errors to you. A LIMS should answer the following questions:
There should be indicators, including pre-and post-analytic stages, that signal all wrong situations. Then you apply them directly to improve your contextual quality.
5. LIMS should tell us not only what, but also when. Trigger warnings are a necessary part of your quality system. That very feature prevents you from making errors in the most sensitive moments- when they can lead to the most serious circumstances.
If you manage all that, you will see that the Quality Unicorn is not a mythical beast forever beyond your reach. It is your beloved pet and trusty companion who will follow you on every adventure.
https://about.vivica.us | info@lifedl.net
© 2024 Life Data Lab, LLC.
Vivica and the Vivica logo are trademarks of Life Data Lab, LLC.
Life Data Lab, LLC is an FDA-registered device manufacturer.
Vivica™ is an FDA-listed, class I laboratory information management system.
