The development of a functional Quality Control (QC) strategy is one of the most important roles of clinical lab managers everywhere. Regular QC helps ensure that test results are accurate and reliable for guiding patient care. But what happens when the QC procedures themselves are subject to error?

This is a key concern of the Working Group on the development of Patient-Based Real-Time Quality Control (PBRTQC) of the Committee on Analytical Quality [1], part of the International Federation of Clinical Chemistry and Laboratory Medicine, which aims to promote international standards and best practices in clinical labs. One of the objectives of the group is to assist laboratories in implementing PBRTQC to complement conventional QC protocols.

With conventional QC, labs typically run synthetic controls, perhaps at the start and end of a shift, or beginning and end of the day. With PBRTQC, labs use virtually every patient result generated throughout the day to identify bias, relying on some patient population parameter such as the mean or median of results across all samples tested to highlight outliers.

This patient-based QC approach has been used for decades in hematology labs because there was no alternative viable Quality Control material for red cell parameters. Now, experts believe it will be adopted for clinical chemistry labs as well. To understand how and why this may happen, the Lab Insights team spoke with Dr Tony Badrick, CEO of The Royal College of Pathologists of Australasia and one of the leaders behind the PBRTQC push.

What’s wrong with conventional QC?

Badrick believes that as labs increase their test throughput, conventional QC approaches become less appealing.

“At one stage we used to run a lot of QC samples, sometimes between every eight or nine patient samples,” he says. “Now we’re running more and more samples with fewer QC samples between them.” This reflects the greater stability of modern analysers, but also creates a problem if bias develops in an assay during the period between QC samples where hundreds of patient samples may be affected. Yet running more QC samples throughout the day could add significant cost to a workflow.

Another problem is that conventional QC samples are synthetic, may deteriorate during the day, and are handled differently from real patient samples. That makes these QC sample results less comparable to patient samples and non-commutable, raising the possibility that a problem with the analyzer could escape notice until it’s too late.

“It’s becoming more apparent that conventional QC has problems, and that those problems are significant,” Badrick adds.

How does patient-based QC work?

The concept behind PBRTQC is fairly simple: by averaging “normal” samples throughout the day, operators can see immediately if values begin to move outside an expected range, just like conventional QC. “You take something you’re measuring routinely, and you use that to try and identify when something has changed with the measurement system,” Badrick explains.

Let’s look at sodium as an example. With a PBRTQC approach, the lab’s software would generate a rolling average for the measured sodium levels from every patient, constantly updating that average as more samples are run. If that rolling average begins to deviate outside the expected range, it would automatically be flagged as a possible quality control issue and the analyzer could be tested and recalibrated if necessary. The reason why the average sodium level of the patient population may change would be if the patient population changes or the measurement system changes, which is bias.

PBRTQC works best for tests run frequently in a lab, so there are enough samples to support the continued generation of rolling averages throughout the day. It is not appropriate for some tests such as tumour markers and other tests that are conducted infrequently or where there are a lot of abnormal results.

PBRTQC also requires a fairly stable patient population, which is not realistic for all labs. A tertiary hospital lab that tests inpatients all morning and outpatients all afternoon, for instance, might have to establish different acceptable range limits based on these two very different patient populations.

For standard community-based testing, though, patient-based QC can be highly effective, says Badrick, noting that LabCorp has been using this technique [2] for years. It’s also cost-effective, since there is no need to buy lots of synthetic control samples. Most labs that adopt the approach run a hybrid model, using conventional QC in the morning and evening and patient-based QC throughout the day.

How to get started with patient-based QC

The thought of shifting to a PBRTQC method may sound daunting, but Badrick believes that most labs will be able to implement this approach over time as more and more manufacturers build patient-based QC protocols directly into the analyzers they sell. “In the ideal situation, the analyzer should flag when things go out of control and let the user know when it’s time to recalibrate the instrument,” he says.

In the meantime, Badrick and his teammates at the Committee on Analytical Quality are doing their best to ensure that the clinical laboratory community has the resources to adopt patient-based QC even before it is built into instruments. They have published several papers and tools on validation, software, simulation and more to help labs get started with best practices. They also recommend setting up some pilot sites and report the results of those efforts in the future.

“It’s not an impossible task,” Badrick says. “It’s a change in mindset as much as anything else.”

Resources

Check out these publications to learn more about how to implement patient-based QC:

Implementation of Patient-Based Real-Time Quality Control 
Critical Reviews in Clinical Laboratory Services, 2020

Patient-Based Real-Time Quality Control: Review and Recommendations
Clinical Chemistry, 2019

Recommendations for Laboratory Informatics Specifications Needed for the Application of Patient-Based Real-Time Quality Control 
Clinica Chimica Acta, 2019

References

[1] Committee on Analytical Quality, International Federation of Clinical Chemistry and Laboratory Medicine

[2] Fleming, K.J., Katayev, A., (2015). Changing the paradigm of laboratory quality control through implementation of real-time test results monitoring: For patients by patients. Clinical Biochemistry. 48(7-8), pp.508-513