Lab Materials in the Age of Smart Laboratories: Why Quality Supplies Matter More Than Ever

As more and more testing moves to digital, instrument-generated results and the data generated from these instruments are being used more and more to make decisions, making quality-managed results even more important. This means investing in automation, moving off paper to digital, and implementing faster and faster instruments. None of these will be of benefit if the lab materials you use cannot stand up to the task.

Finding the most suitable lab material supplier is no longer a purchasing consideration but a cornerstone in any quality management system. Facilities must ensure that they use the correct laboratory consumables to produce the right result. Consumable materials, reference materials, reagents, glass, sample containers, etc., directly or indirectly affect the accuracy of all processes. In addition, sophisticated instruments do not perform well if the matched supply materials are not up to par.

Modern Laboratories Are Built Upon Consistency

A modern laboratory differs greatly from one even just 10 years ago. You may be thinking of the latest liquid handler, getting rid of paper, and driving systems integration.

But one fact has not changed. Consistency equates to solid science.

Analytical methods consist of many links, such as sampling, sample pretreatment, sample analysis, and data processing and analysis. Even if a small problem occurs in one link of the chain, it will bring a huge deviation to the laboratory test results. The consistent and stable standard of laboratory supplies directly affects whether the final laboratory testing standard is met.

Automation Increases Throughput — But Only If the Input Is Correct

You can’t talk about increased laboratory efficiency without talking about some form of automation. Liquid handlers, sample preparation, sample transport systems, and analytical processes that take ICP/MS and AA instruments off-line are a few examples of robotic systems that can be applied to decrease the time spent preparing and processing samples.

The benefits are numerous:

• Faster sample processing
• Fewer people handling samples
• Fewer transcription errors
• Fewer people handling samples
• More samples processed

But there are limits, too. Automation is the solution that most laboratories will choose, but robotic systems break if the inputs do not meet extraordinarily exacting standards. Pipette tips must be calibrated to their arm. Tubes must be shakable, breakable, and vial-able. Reagents should be unvarying across batches, as should calibration kits. They must be gold standards up until their expiration date.

If any of the above are even slightly off, the automated systems will spit out garbage data, and the whole process will have to be re-run manually, so the efficiencies of an automated process are gone.

It’s especially important in pharma research, where thousands of compounds might run in a week. Automation builds speed efficiencies. Unless every designated gear in the machine runs exactly as designed, those efficiencies are lost.

Accuracy Comes First, and There Hasn’t Been a Single Measurement Yet

A lot of people think about the precision of a laboratory as something that can be measured by a chromatograph, a spectrometer, or an analytical balance. This isn’t the full story at all.

Precision really starts much, much sooner than that.

Science that is not accurate begins at the point where the sample is being prepared.

Why is that? Because if there is a mistake with the volume, if an aliquot is too long in the wrong sterile strip tube, if the reagents have degraded, or if the reference standards have waned, nothing that an analytical instrument can do will help correct the error. The science is inaccurate because it wasn’t completed properly from the start. The science was flawed from the beginning. A machine can’t correct that. No amount of science-fiction technology can correct that flaw.

That’s why so much money is spent on laboratory supplies, including:

• Certified reference materials (CRM)
• High-purity chemicals and reagents
• Volumetric glassware
• Balances
• Sterile sample containers and laboratory-grade filtration materials, etc.

This is crucial when results need to be compared, especially over time, across different operators or instruments, across different laboratories, or in the case of regulatory examination, etc. Only confidence limits based on metrologically traceable reference materials, etc., have the authority or gravitas to survive that kind of critical evaluation.

Traceability Is Now Compulsory. Full Stop. The End.

Getting a result isn’t enough anymore. Now you have to prove how that experimental observation came into being.

By traceability, we mean a complete record of all actions and events that have occurred over the course of the experiment.

This is being done more and more with electronic data systems. Barcoding on samples or other assets speeds scrutiny. RFID tags attached to individual items sometimes provide mass information globally. Laboratory information management systems (LIMS) can log tests without the need to separately write down all that information by hand.

We just want there to be a record of everything relevant:

• The sample
• Where it came from
• How it was stored
• The reagents used
• How much was used
• How those reagents were stored
• How much sample was used
• The specific batch we’re counting from
• Calibration
• Who did it
• Quality control work
• Results

In case you miss something or simply enter the wrong code, box, etc., the quality investigation can show changes over batches more easily. An audit will go much more smoothly if confidence intervals have been established for counting, even between different laboratories.

Digital Labs Need Reliable Lab Materials

The lab of the future isn’t just more automated. It’s more connected. AI is helping interpret results. Predictive maintenance software can forecast equipment problems before they fail. Cloud computing lets labs in different locations analyse data simultaneously. All these innovations can deliver impressive efficiency gains.

But how much value do they deliver if you can’t trust the data in the first place? What if impurities from a poor-quality consumable contaminate the results? Or variations in reagents make data unreliable? Software, no matter how advanced, can only process data inputs. If there are problems at the raw material stage, then those digital processes automate the creation of poor-quality data.

Building Labs for the Future

The work carried out in scientific research, and the testing laboratories do for quality control, does not stand still. Automation, machine learning, and other highly technical processes are all subject to continuous evolution. In practice, we can expect labs to deal with bigger data sets and greater sample loads. There will also be a continued regulatory expectation that the data in question is of higher fidelity.

Even in these circumstances, the fundamentals hold true.

A reliable material is one that supports the application. One you can trust to deliver accuracy and data that can be relied upon, a workflow that is dependable, and a science-backed conclusion built on strong foundations. Automation certainly increases productivity. Advanced software models and algorithms help us identify patterns and discover meaningful outputs where there is high-dimensional data. Humans still need to be present to add context and additional information, but the models provide valuable insight into the laboratory environment.