Questions

For most of laboratory history, how often you ran QC was not a question, it was a regulatory mandate. But the latest QC calculations enable laboratories to determine a completely customized QC frequency. Are laboratories ready?

What is QC Frequency?

Sten Westgard, MS
November 2023

In the earliest dawn of the laboratory age, how often to run QC was not a question: there wasn't any QC yet. Then came pooled patient samples, which evolved into commercial control materials. How often they should be run was codified by regulations: The CLIA Final Rule of 1988 established a minimum QC frequency for most tests of once a day, and for a few tests, once every 8 hours.

QC frequency, therefore, was wed to the clock. It was driven by the passage of time, detached from the performance of the method or the volume of patients tested or the interpretation of the results.

Even as new techniques emerged to help optimize QC, such as critical systematic error and analytical sigma-metrics, the frequency remained the same. It was possible to optimize the number of levels and rules, but the math and rationale for frequency was still out of reach.

So in the clock-dependent QC world, a laboratory running 10,000 samples a day would  run the same QC as a lab running 20 samples a day. Perhaps the high volume laboratory manager would run more, but by CLIA standards, running once a day was acceptable for both. And still is.

Parvin's MaxE(nuf) model, first published in 2008, finally provided a model for the timing of QC in relation to the volume of patients. It focused on minimizing the maximum risk to patient results. The switch from clock-based to patient-based QC frequency had occurred. And the analytical Sigma-metric enables the laboratory to determine how often to run QC. Higher Sigma metrics can potentially reduce their QC frequency to a regulatory minimum. Lower Sigma metrics might motivate a laboratory to increase the frequency of QC.

We're still digesting those implications.

The benefits of customizing QC frequency are challenging to implement. Since the laboratory world works on a multiconstituent control world, so it's hard to customize the QC frequency for single assays. A control that covers 40 assays will by necessity have only 1 QC frequency for those 40 assays. Some assays will be over-QC'd, others will be under-QC'd.

Further, only the higher volume laboratories are really in a position to take advantage of customized QC frequencies.  The highest volume laboratories, running thousands or tens of thousands of tests per day, can really enjoy the gains of efficiency of optimized QC. If the laboratory only runs 200 samples a day, on the other hand, there aren't as many changes that can be made. If the lab learns that the optimized QC frequency could be 1000 samples a day, the smaller lab will will lose out on that optimization. Typically, the regulatory and manufacturer recommendations require running QC every day. It's unlikely that regulations will acquire the flexibility to accept different QC frequencies for different tess. A simplistic floor for QC frequency, once a day, will probably remain in place.

Even if the QC frequency cannot be customized, there is still the ability to adjust the rules and levels for QC. A fixed QC frequency can be supplemented by reduced rules, wider limits, and reduced levels.