Patient-like vs. commutable
Even though the term “commutability” was first defined in 1973, it was only 45 years later that the concept of commutability became an integral part of the Tietz textbook, the “bible“ in the field of Clinical Chemistry. This nicely illustrates that «commutability» lived in the shadows for a long time, often in favour of its more famous counterpart „stability“. But let’s start from the beginning...
Far from all biomarkers in patient samples are stable, which is unfortunate for manufacturers of quality control (QC) materials. For example, the half-life of the inflammatory marker Procalcitonin is about 24 hours in serum. And here is the dilemma: To what extent should a manufacturer of QC materials stabilize such labile biomarkers? Stabilization can result in a QC material that does not behave like a patient sample. But how does this in turn affect the daily work in your laboratory?
A QC material can be seen as a fire alarm. Obviously, you do not want your fire alarm to go off without reason, for example while cooking. However, it should alarm you in case of a real fire. This means, the fire alarm must be set to the right sensitivity. For QC materials used as daily controls (i.e. precision materials), the sensitivity switch is «to behave like a patient sample»: it must be sensitive enough to detect changes in the performance of the method. And it should be as sensitive as a patient sample.
If you use an artificially stabilized QC material as your daily control, it might have lost its sensitivity to detect fluctuations in your measurement system. Therefore, the measurement result of your daily control might look unsuspicious in your control card even if it should have given you a warning. A warning, expressed as a measurement result that differs significantly from the value you got the previous days or weeks. However, your patient sample measured afterwards is not artificially stabilized and, thus, much more sensitive. Therefore, your patient sample does respond to the fluctuations in the instrument and will in turn report a false value. In other words, a QC material that does not behave like a patient sample is a fire alarm that might not go off when it actually should. And the most unfortunate thing about it: you do not have a chance to know that something is wrong! In the worst case this can lead to an incorrect diagnosis or therapeutic treatment.
In 2015, SERO AS demonstrated that not all QC materials on the market behave like patient samples. This was shown by assessing the buffer capacity of QC materials compared to patient samples (Figure 1). One of the three tested quality control materials clearly did not behave like a patient sample. This means that this QC material from manufacturer B has lost its sensitivity for pH-fluctuations in the measurement system. And this is just one parameter to assess patient-likeness. You could go further and check for temperature sensitivity, protein and lipid composition to just mention some more.The above example highlights the importance of patient-like QC materials for one given measurement procedure in your daily control routine. In contrast, commutability of QC materials is demonstrated by looking at min. two measurement procedures. To be commutable, the relation among the obtained measurement results for a QC material has to closely agree to the relation seen in a representative patient sample for a stated quantity1. Therefore, commutable QC materials are required to make patient results comparable between different measurement procedures and across laboratories. Accordingly, the European legal framework asks for assessment of commutability when reference materials are intended to be used as trueness or accuracy controls in its draft of the revised ISO standard 175112.
However, it is important to understand that a commutable QC material does not necessarily need to be patient-like. This can be illustrated by a calibrator that for example consists of a highly purified analyte in a buffer, and therefore is not patient-like. This calibrator can still be commutable under specified measurement conditions, even though it is not useful as a daily control to check for fluctuations in your measurement system. Therefore, it is important for a manufacturer to establish the intended use of its QC materials.
Matrix-related factors
Turbidity
Abnormal viscosity
Presence of endogenous interfering substances
Use of procedures that result in physical changes, e.g. lyophilization
Analyte-related factors
Proteins of non-human origin
Isoenzyme patterns
Partially denatured protein
Glycosylation patterns
In practice, few multi-control QC materials are commutable for all their analytes. And very few QC materials are commutable for all measurement procedures on the market. This is because each step in the production of QC materials can affect the commutability of its analyte(s). Factors causing non-commutability can either be analyte- or matrix-related (Table 1). Therefore, it is impossible to predict commutability. This requires that commutability must be tested and documented.
References:
1 https://www.bipm.org/en/publications/guides/vim.html
2 https://www.iso.org/standard/69984.html
3 Clin Chem Lab Med. 2019 Jun 26;57(7):967-973. doi: 10.1515/cclm-2019-0154
WRITTEN BY
Johannes Bauer
Senior Project Leader