Serum, plasma or blood? Some confusions about the NPU terminology - Dynamic NPU Manual

Concentrations of substances in whole blood compared to plasma

A concentration is the amount of a substance in a solution divided by the volume of that solution. “Blood glucose” could therefore be thought of as the amount of glucose in a sample of blood divided by the volume of blood. Glucose is dissolved in the water phase of the blood, and because water is not evenly distributed in blood, proteins and lipids displace water, and erythrocytes in the blood contain a lot of protein in the form of haemoglobin. The water content per volume of erythrocytes is therefore less (0.71 kg/L) than the water content in blood plasma (0.93 kg/L) or in  hemolyzed whole blood (0.84 kg/L). The concentration of glucose is therefore higher in plasma than in whole blood, which is higher than in the erythrocytes. When referring to blood glucose it must therefore be stated which blood glucose concentration is meant (Glucose in Blood, Plasma, or Erythrocyte).

The NPU system provide codes for many quantities with both Plasma and Blood as system quantities where measurements of concentrations are expressed for both types of matrices.

The difference between concentration values in plasma and whole blood can be described by “the mean plasma/whole blood concentration ratio”. Some examples are given in the table below. For substances like glucose and ethanol which are evenly distributed in the water phases of plasma and whole blood the P/B ratio is 1.11 – 1.2. Other substances differ in P/B ratio because of exclusion of larger molecules from the intracellular space of erythrocytes and other cellular elements of the blood, or due to specific binding to proteins with different intra and extracellular concentrations.

NPU code Quantity P/B ratio
NPU23869 P—Ethanol; mass c. = ? g/L  
NPU54578 B—Ethanol; mass c. = ? g/L 1.16 (1)
NPU53100 P—Ethyl glucuronide; subst.c. = ? µmol/L  
NPU53856 B—Ethyl glucuronide; subst.c. = ? µmol/L 1.7 (2), 1.3 (3)
NPU01879 P—Diazepam; subst.c. = ? µmol/L  
NPU53813 B—Diazepam; subst.c. = ? µmol/L 1.79 (4)
NPU21531 P(vB)—Glucose; subst.c. = ? mmol/L  
NPU04093 B(vB)—Glucose; subst.c. = ? mmol/L 1.11 (5)

For electrolytes (Na+, K+, Ca2+, Mg2+) the differences between intra- and extracellular concentrations are large, due to active transport of the ions over the cell membrane. It is normally the extracellular concentration, or plasma concentration, that is intended to be measured regardless of specimen type.

Ion selective electrodes are sensitive to the ion activity of the water phase of the plasma which is in contact with the electrode membrane. As the water content of plasma is 0.93 kg/L the activity is 7 % higher than in plasma, but the devices are normally calibrated to show the concentration in plasma. Free calcium ion can thus be measured through the membrane in contact with the water phase of heparinized whole blood, heparinized plasma or serum. Regardless of the specimen type the measured values are converted to reflect the in vivo ionized calcium activity in plasma (6). The challenge is that the concentration of heparin should be kept as low as possible as heparin binds calcium, and excess heparin induces false low calcium activities.

The intracellular ion concentrations differ by several orders of magnitude to the extracellular concentration. Hemolysis will therefor affect the plasma concentration of these ions, for example, if the potassium concentration is false. In principle, the calcium concentration in plasma is falsely decreased in hemolyzed samples (7), although the final effect on the results reported also depend on the degree of interference from haemoglobin with the measurement procedure in use.

Measurements of substance concentrations in a hemolyzed full blood sample are sometimes performed, for example glucose, C-reactive protein, and (most probably?) creatinine. Except for glucose (see table above) the concentration values are not reported but converted to the expected value for that of the plasma concentration.

The measurement of the ion activity in the fully hemolyzed whole blood sample can be done but has no biologic meaning and is therefore not requested for clinical purposes.

The NPU code for “Blood—Calcium ion (free)” is probably not requested.

NPU60437 Blood—Calcium ion(free); substance concentration = ? millimole per litre

The reason for the creation of the code could be to express that measurement is performed in a sample of heparinized whole blood with an ion-selective electrode, and not in a sample of plasma or serum. In the NPU terminology there is unfortunately no way to specify a specimen type, unless the specimen is part of the definition of the quantity intended to be measured in vivo. For the calcium ion activity and other electrolytes there is no difference in what is intended to be measured, whether the sample is serum, plasma or whole blood.

Conclusion: The specimen type should be reported independent of the NPU code.

Does an ion selective electrode measure the same quantity as a chemical measurement?

It has been argued that the measurement principle should be part of the definition of what is measured in undiluted samples with ion specific electrodes. As long as the intended quantity to be measured is the same (that is the ion concentration in the plasma phase) the measurement principle should not be specified according to the basic principles.

The reason behind the discussion is that chemical measurements area measure of the substance concentration in the whole volume of the solution (see above). The ion selective electrodes measure the chemical activity of the ion in the water phase of the sample and converts the value to that expected for plasma assuming a normal lipid content of the sample and thus a water content of 0.93 kg/L. In cases of lipemia the lipids displace water and reduce the normal water content to a lower value than 0.93. A chemical measurement measures the substance concentration per whole volume of the solution including the lipid phase. The chemical measurement produces in such cases a lower value than the ion selective electrode.

Conclusion: It could be argued that chemical measurements and ion selective electrodes refer to different quantities, and that the measurement principle therefore should be reported for electrolyte determinations. Against this view holds that the difference has clinical significance only in cases of severe lipemia. It seems therefore preferable to report measurement principle and method independent from the NPU codes also for measurement of ions.

Do we need NPU codes that distinguish between serum and plasma?

When blood is clotted in a tube outside the body, a clear liquid is produced over the clot. Serum has since long been used as a convenient specimen type for measurement, but serum is not found in vivo anywhere in the body. Serum is used as specimen, because its convenient (and cheap) and because measurements are thought to reflect that of the blood plasma concentrations in vivo. Serum is thus a specimen type, like heparinized plasma or EDTA plasma, which all reflects the metrological system: blood plasma. For principal reasons, the NPU codes do not contain information about specimen type, which is part of the sampling process outside the body. However, after discussions with users, national codes have been defined in Sweden (and Norway?) in two cases with “serum” as a specification to the system:

  1. For concentration of potassium. Slightly higher values observed in serum compared to plasma which is thought to be due to leakage of intracellular potassium into the serum during the coagulation process. It could be discussed whether the distinction is necessary for potassium. The difference in concentration between heparinized plasma and serum is usually small, and the reason for a difference is not clear for cases with high platelet and leukocyte numbers.
  2. For concentration of “total protein”. Lower values occur in serum because serum does not contain fibrinogen which is consumed during the coagulation process when serum is formed.

Conclusion: For the component: potassium, the distinction between serum and heparinized plasma might not be necessary and the distinction between serum and heparinized plasma can instead be made as different specimen types. For “total protein” the intended quantity to be measured clearly differs between serum and heparinized plasma. A more suitable formal description of the measurand would be “Protein excluding fibrinogen” with Plasma as the formal system.

Sometimes it might be complicated to distinguish the metrological system used.

Base excess is a derived and theoretical quantity, more used in Scandinavia and Europe than in the US (8,9). The challenging question is what the metrological system is. Is “actual base excess” a quantity that is expressed per volume of blood plasma or per volume of whole blood?

The base excess is an estimated variable for either a base excess of the extracellular fluid of the body (corresponding to mean Hb=50 g/L), sometimes termed standard base excess or SBE or cBase(ecf). Base excess can also be expressed for whole blood (Hb=actual Hb or approximately 150 g/L), which is sometimes termed actual base excess, ABE or cBase(B).

In Sweden a local SWE-code with ”Blood” as the system for “actual base excess” has been recommended, instead of the codes NPU12518—21 which have Plasma as system.

Conclusion: The system Plasma as used for the concepts defined by NPU12518—21 is probably incorrect. It should therefore be further discussed if new NPU codes should be created with Blood as the system for the concept sometimes called “actual base excess”. As this might create confusion, an alternative might be to correct the system term from Plasma to Blood for NPU12518—21 completely.

Blood as specimen type

C-reactive protein (CRP) is an example of a substance that can be measured in a whole blood volume, but the results are always expressed as the corresponding plasma concentrations and are expected to be compatible with results achieved in specimens of plasma and serum. The term “blood CRP” can therefore not be misinterpreted to denote the concentration of CRP in whole blood volume. The same holds for “blood sodium” or other substances whose concentrations never are expressed per whole blood volume. The concentration of glucose on the other hand can either be measured in and expressed per volume of whole blood or be expressed per volume of plasma. The expression “blood glucose” is therefore ambiguous. To avoid mistakes it has, in many countries, been agreed to always express glucose concentration as the concentration in blood plasma, and accept the general phrase “blood glucose” to cover also the concentration of glucose in plasma.

Generally, there are no unique NPU codes for specific specimen types. What is meant by the various system specifications for glucose, e.g. “P(vB)—Glucose; subst.c.”, is that the glucose concentration varies within the different body compartments such as the venous and arterial systems. The specifications do not concern the specimen type, just the anatomical site, when this is a part of the definition of the measurand.

20230929/20231003/Gunnar Nordin


(1) Jones AW. Alkoholtest på sjukhus inte helt lätt att använda för rättsligt bruk. Omräkning av etanolhalt i plasma eller serum till promillehalt i blod. Läkartidningen. 2008;105(6):367-8.

(2) Hoiseth G, Morini L, Polettini A, Christophersen AS, Johnsen L, Karinen R, et al. Serum/whole blood concentration ratio for ethylglucuronide and ethyl sulfate. J Anal Toxicol. 2009;33(4):208-11.

(3) Neumann J, Beck O, Helander A, Dahmen N, Böttcher M. Sensitive determination of ethyl glucuronide in serum and whole blood: detection time after alcohol exposure compared with urine. J Lab Med. 2020;44:211-19.

(4) Jones AW, Larsson H. Distribution of diazepam and nordiazepam between plasma and whole blood and the influence of hematocrit. Ther Drug Monit. 2004;26(4):380-5.

(5) D’Orazio P, Burnett RW, Fogh-Andersen N, Jacobs E, Kuwa K, Kulpmann WR, et al. Approved IFCC recommendation on reporting results for blood glucose: International Federation of Clinical Chemistry and Laboratory Medicine Scientific Division, Working Group on Selective Electrodes and Point-of-Care Testing (IFCC-SD-WG-SEPOCT). Clin Chem Lab Med. 2006;44(12):1486-90.

(6) Boink AB, Buckley BM, Christiansen TF, Covington AK, Maas AH, Muller-Plathe O, et al. International Federation of Clinical Chemistry (IFCC), scientific division: IFCC recommendation on sampling transport and storage for the determination of the concentration of ionized calcium in whole blood, plasma and serum. Journal of Automatic Chemistry. 1991;13(5):235-9.

(7) Robinson JL, Seiden-Long I, de Koning L. Identification and implementation of hemolysis interference thresholds in serum ionized calcium measurement. Clin Biochem. 2020;78:66-7.

(8) Kofstad J. Base excess: a historical review – has the calculation of base excess been more standardised the last 20 years. Clinica Chimica Acta. 2001;307:193-5.

(9) Berend K. Diagnostic Use of Base Excess in Acid-Base Disorders. N Engl J Med. 2018;378(15):1419-28.