Calibrations vs. Quality control (QC)

There is confusion in regarding the terms "calibration" and "quality control", or QC (or IQC- Internal QC). Many POCT operators talk about calibrating their POCT analyser or process when they are actually analysing QC samples. 

 

The processes of calibration and quality control are two sides of the same coin: they complement each other. The calibration establishes an initial point of measurement or data point in a reaction; the QC checks that the calibration is correct. Together, they determine the reliability of the method, i.e., the accuracy and precision of the method. 

 

The processes described on this page apply to all laboratory testing, not just POCT.

 

Operating principles

Modern laboratory and POCT analysers work on a variety of operating principles- spectrophotometry, electrochemistry, potentiometry, amperometry, chemo- and electro-luminescence, reflectometry, etc. 

 

Regardless of their operating principles, all measuring processes must be calibrated before use. The simplest explanation of calibration deals with traditional spectrophotometric principles.

 

Types of calibrators

There are also different types of calibrators- some POCT analysers, e.g., Glucocard meters, or Roche CoaguChek meters use a factory-set chip, replaced in the analyser each time a new box of strips is opened. Lot numbers of calibrator and test strips must be identical so as to eliminate any error caused by variation between lot numbers.

 

Some analysers like the i-STAT analyser use a cartridge with a built-in calibrating solution- after the patient's blood is loaded into the cartridge, the cartridge is inserted into the analyser. The analyser mechanism causes a small reservoir of calibrator fluid to move across the electrodes in the cartridge. After the electrodes are calibrated, the patient's blood sample is shifted into position and analysed. 

 

On the other hand, bench top blood gas analysers require a separate liquid or gas calibrator to be analysed in a process analogous to that of a patient sample. 

 

Liquid calibrators can be made from whole blood, serum or plasma or aqueous solutions, depending on the assay. 

 

If you are interested in more information about any of the above topics, please contact the POCT Coordinator.

 

Calibration

A calibration determines the initial value(s) in a reaction. These values are equivalent to the known concentrations in the calibrating solutions. A minimum of 2 calibrators is required to establish a calibration factor, which is often referred to as "k".

 

Calibrations are required when the k value is either unknown or has shifted in value from a previous calibration. 

 

There are many reasons why this might have happened. Here are some of them: 

  • a new analyser

  • a new test

  • a slow change in analytical conditions. Sometimes referred to as "drift". There can be various reasons for drift. It is usually detected when the QC fails.

  • new hardware, e.g., a replacement electrode

  • new reagents

  • new calibrator 

  • out of range QC

  • any other change in analytical conditions

We can either make calibrators in the lab (very carefully!), or we can purchase them from a manufacturer, the usual procedure these days. (See the sidebar for more information). 

 

The exact values of concentration of each calibrator- there may be more than one analyte you are calibrating, say in a blood gas analyser- are keyed into the analyser and stored there. The lot number and expiry date is usually also stored. All these values will not be altered until a change of calibrator with different values, lot number and expiry date is introduced.

 

The calibrator is then analysed in a manner identical to that of a patient sample but the values derived from each reaction are stored in the analyser in special files relating to calibration.

 

The analyser then calculates the k or calibration factor for each test and stores the value in its memory. 

 

How it works

Of course, this procedure can be done manually on the bench with a sheet of graph paper and a calculator. It used to be done this way in laboratories all the time. However, these days we have analysers.

 

With most biochemical analyses, e.g., glucose if the absorbance is plotted against concentration, you would get a straight line on a graph, i.e., a linear reaction:

 

    

 

The angle or slope of the line gives you the k value, which remains constant for the whole of the line length.

 

k is calculated using Beer-Lambert's Law. (see the sidebar). This is a fundamental physical law which states that:

 

Given a linear spectrophotometric reaction (see above), that:

 

Absorbance is always proportional to Concentration.  

 

Once the calibration factor is included in the calculation,

 

Absorbance multiplied by k equals Concentration. 

 

Therefore, when a QC sample of known value, or a patient's sample with an unknown value is analysed, its energy value or Absorbance in the reaction is multiplied by k to give the Concentration.

 

Here is another way of looking at it. The graph on the left details the calibration procedure. The graph on the right details QC or patient sample analysis. Note the direction of the red arrows on each graph:

 

CALIBRATION Q C or PATIENT

 

1. Value (Concentration) of calibrator is known. 1. The Concentration of the QC or patient sample is not known.
2. Reaction rate (Absorbance) of Calibrator is measured. 2. Reaction rate (Absorbance) of QC or patient sample is measured.
3. The calibration factor, k, is calculated 

k = Concentration divided by Absorbance

 3. The Concentration is calculated.

Concentration = Absorbance multiplied by k

 

Notes:

  1. For Beer-Lambert's Law to hold true, i.e., if "absorbance is proportional to concentration", then a number of conditions must be included in the calculation- the choice of reagents, the light measuring capabilities of the analyser, the length of the measuring chamber or cuvette, the wavelength chosen to measure the reaction and so on.

  1. k is represented by the slope of the line on the graph. So long as the reaction stays linear, i.e., the line stays straight, k will have an identical value for any particular reaction.

  1. When the reaction is not linear, things really get tricky and rather interesting! But that's another story... :-)

Quality Control (QC)

 

Once we have calibrated our analyser, we still need to know if it has been calibrated correctly. Thus the QC results check on the calibration. They should tie in together, each proving the correctness of the other.

 

Click on this link for a far more detailed discussion on Quality Control. 

 

NOTE: The QC analysis checks the operator's ability to perform the analysis and whether the correct result will be forthcoming on the patient sample. 

 

What it does not check for is the quality of the patient's sample. The quality of the sample, especially if a capillary whole blood sample, is fundamental to the generation of the correct result.

 

Points to remember:

  • Calibrations must be performed if analysis conditions alter in any way, e.g., a change of consumables, hardware or reagents or if the QC fails.

  • No analyser should ever be used without a current calibration.

  • No analyser should ever be used without a recent correct QC result. 

  • QC samples should be analysed regularly- at least once per day.

  • All changes in procedure, reagents, lot numbers, etc; maintenance and troubleshooting problems; and QC results must be documented.

  • All error codes, incorrect QC or procedural lapses MUST BE DOCUMENTED. Use the Error Log associated with each analyser- each Error Log is to be found in its associated Operators Manual .
What calibrates the calibrator?

A calibrator's value is determined by the "Reference Method", the recognised most accurate and precise methodology for the test in question.

 

Reference Methods utilise high tech equipment like mass spectrometers, atomic absorption (AA) spectrophotometers and Gas Chromatograph (GC) or High Performance Liquid Chromatographs (HPLC). The purest substances are used to obtain the most accurate assays. 

 

Calibrator solutions, supplied by manufacturers, are made from these purest of substances. They are analysed repeatedly on most common laboratory analysers. The results are statistically analysed to give the Average or Mean Value for each constituent on each analyser type and often for different test kits, whose test principle (and therefore sensitivity) may vary. These values are then keyed into the appropriate laboratory analyser and a calibration is performed for each analyte. 

 

You may ask why all lab tests aren't analysed directly by Reference Method. Cost and efficiency come into it. Often processes using mass specs, AA or HPLC are elaborate, expensive and time consuming. Even so, some lab tests are measured using these methodologies- drug assays or vitamins. Generally though, methods are chosen that are simple and easy to use, but which still give reliable accurate results.

 

The routine methods used must give values close to the reference method in order to be accurate. However, different methodologies do have different characteristics. That is why it is important that each lab establishes their own range of values for their QC material.

 

It is also important for each laboratory to participate in External Quality Assurance programs, so that they can compare their results to other labs to ensure that all results are identical within an acceptable degree.

A bit more detail...

 

The commonest type of laboratory analyser is called a Spectrophotometer. In its simplest form, it is an analyser in which a beam of monochromatic (single coloured) light, adjustable to any particular light wavelength, can be passed through a coloured solution derived from the analyte in question. An example is urea which gives a blue colour using the commonest method. This solution can be either the calibrator or patient/QC samples, depending on whether the test is being calibrated or samples are being analysed.

 

Usually complementary coloured light is used, e.g., a red beam would be passed through a green solution. 

 

A percentage of the light, called the Absorbance, is assimilated by the coloured solution. The Absorbance is directly proportional to the Concentration of the substance in solution.

 

The emerging light, the Transmittance, is inversely proportional to the Concentration. It is measured by a detector, converted to a digital signal, processed electronically and the expression is finally multiplied by the Calibration Factor. The result can then be displayed on a screen or can be printed out.

 

(In most calculations, even though the transmittance (t) is actually measured, the Absorbance (a) has more relevance as this is directly related to the Concentration (c).

 

a 1/t

c 1/t

a c 

 

Concentration

The value of the substance under investigation. Various units of measurement are g/L, mmol/L, umol/L, IU/L, nmol/mL and so on.

Beer-Lambert's Law

A° C or  A°. k = C

where: 

A° is Absorbance

is the symbol for "proportional to"

C  is Concentration

.   is the symbol for "multiplied by"

k   is konstant or calibration factor