Maas: IFCC reference methods for pH, gases and electrolytes in blood: reference materials 253

Eur. J. Clin. Chem. Clin. Biochem.Vol. 29, 1991, pp. 253-261© 1991 Walter de Gruyter & Co.

Berlin · New York

IFCC Reference Methodsfor Measurement of pH, Gases and Electrolytes in Blood:Reference Materials1)

By A. H. J. Maas

Faculty of Chemical Technology, Technical University Eindhoven, The Netherlands

(Received January 22, 1991)

Summary: The Scientific Division Committee on pH, Blood Gases and Electrolytes of the InternationalFederation of Clinical Chemistry (IFCC) pröduced recommendations to attempt to make the results of pH,blood gas and electrolyte analysis from different clinical chemistry laboratories internationally compatible.

The aim of this lecture is to discuss the essential aspects of

1. the IFCC approved (1986) reference method for pH measurement in blood,

2. the IFCC approved (l 988) reference method for tonometry of blood,

3. the provisionally proposed recommendations on the expression of results obtained with ion-selectiveelectrodes for measuriag sodium, potassium and calcium in serum, plasma or blood and

4. the reference method for the determination of ionized calcium in serum, plasma or blood.

Also reference materials for quality control of pH, blood gas and electrolyte measurements are reviewed.Failures of several types of currently available quality control materials are discussed.

' ion-selective electrodes for the measurements of theFor the assessment of the state of the acid-base bal- electrolytes gives problems because the results may beance and gas transport, sophisticated Instrumentation different from those obtained by flame photometry.is available by which routine measurements of pH often inad ate accuracy and precision lead to in.PC02 *ndp02 can be made simultaneously m a small ^ t care Errors can be ̂ To maResamp e of blood Wrth the development of , -se ec- reeuta medicall useful thefe is & need to establish

üve electrodes, wh,ch enable measurement of the elec- ^ requirements of measurements.trolytes sodmm, potassium and ionized calcium di^rectly in blood, these analysers now include blood The International Federation of Clinical Chemistrygas/electrolytes analysers. (IFCC)/Scientific Division Committee on pH, Blood

A , , , . , Gases and Electrolytes, previously named the ExpertAlthough th1S mstrumentation is weU manufactured p&nel Qn and Blood Gas^ ̂ ^ work -n 19??the possibility of makmg great errprs is not excluded Qn ̂ documents deali ^ the terms of refer.e. g. a deviating residual hquid junction potential mthe pH measurement. In addition the introduction of

. . . . l . definitions and terminology,') Based on a lecture given at the Symposium "Reference Meth- f A

ods in Clinical Chemistry - Objectives, Trends, Problems" 2- reterence methods andof the Congress Biochemische Analytik 90, München, May 8, r .3. reierence materials,

Eur. J. Clin. Chem. Clin. Biochem. / Vol. 29, 1991 / No. 4

254 Symposium: Reference Methods in Clinical Chemistry

to atlempt to make the results of pH and blood gasanalysis from different clinical chemistry laboratoriesinternationally compatible.

In order for the measurement process of a quantityto be meaningful, a hierarchical System äs dia-grammed in figure l has been developed by scientistsof the National Institute of Standards and Technology(NIST), previously the National Bureau of Standards(NBS) of the USA (1). This System indicates thecoupling of the analytical methods (definitive method,reference method, field or routine method) and pri-mary and secondary reference materials. The functionof each of the components is to transfer accuracy tothe level immediately below it and to provide trace-ability to the level immediately above it, therebyassuring compatibility in the overall measuring sys-tem.

I want to show you how this System may be appliedto the establishment of the reference method for pHmeasurement-in blood (2) and thereafter of the ref-erence method for tonometry of blood (3).

1.1 pH

pH is defined äs the negative logarithm to base 10 ofthe relative molal activity of hydrogen ions:

pH = - Ig aH+

The definitive method for pH measurement in diluteaqueous Solutions is based on measuring the electro-motive force of a cell with a hydrogen electrode anda silver/silver chloride electrode without a liquid-liquid junction i. e. without transference. This cell isoften called the Harned cell. , r

The definitive method is employed by reference lab-oratories, e. g. the NIST, previously the NBS in USA,which has assigned pH(S) values to a series ofprimaryaqueous calibration Solutions in the pH ränge from 3to 11 with an estimated uncertainty in pH of ± 0.005:the so called NBS-buffers.

Reference methods for pH measurement are based onthe use of a cell with a hydrogen gas electrode and areference electrode with a concentrated KC1 liquid-liquid junction. Primary reference materials are usedfor calibration. These reference methods may be usedto establish pH values of secondary calibratioil solu-tions.

The hydrogen gas electrode is unsüitable for pH meas-urements in biological fluids due to protein contam-ination of the platinum surface, where the presentIFCC reference method for pH measurement in bloodshould be based on the glass electrode.

The IFCC reference method for pH measurement inblood is based on the use of a cell consisting of a H"1"ion-selective glass electrode and a reference electrode

> r

o Field method developmentand evaluation

e Preparation of routineworking reference materials

o Routine internal andexternal quality assurance

IIDefinitvemethods

A1 Prinl refer\mate

>

"io Reference method

L. development and1 >, evaluationiary \ence r "~irials/

rIV

Referencemethods

/Seco"~i refer

\mat€

1

ndary\ence j>rials/

fVI

Fieldmethods

secondary reference materials

e Critical quality assuranceapplications

«

Fig. 1. Schematic Präsentation of the relationship among the technical components of an "idealized". accuracy-based chemicalmeasurement System. ;

Eur, J. Clin. Chern. Clin. Biocheni. / Vol. 29,1991 / No. 4

Maas: IFCC reference methods for pH, gases and electrolytes in blood: reference materials 255

with a saturated KC1 solution äs salt bridge accordingthe scheme:

"Ref. Electrode" "Glass Eleclrode"

R, l KC1 (sald.) || sol. X l G l inner ref. sol. l R2 (cell b)

Electrodes R! and R2 are connected to the pH-meter.The whole System is thermostatted at 37 °C.

Two primary calibration Solutions of the NBS seriesare needed for calibration: phosphate buffer ofpH = 7.392 (37 °C) is used for setting the zero pointof the Instrument and the phosphate bufferpH = 6.839 (37 °C) for the slope adjustment. In figure2 a schematic presentation of the pH measuring sys-tem is presented. The glass electrode ought to be agood approximation of the hydrogen electrode. Thecapillary type is chosen for the reference method ofpH measurement in blood. To obtain the true plasmapH a small bridge of plasma should be interposedbetween the blood and the concentrated KC1 to avoidthe effect of the blood cells minimizing the residualliquid/liquid junction potential.

The pH measuring System of the BMS2-Mk2 fromRadiometer (Copenhagen, Denmark) is an adequateSystem which fullfils the requirements of the presentedreference method. The pH values of blood samplesdetermined with this System can be directly comparedwith pH values obtained with the definitive method.

For further details reference should be made to thedocument for pH measurement which was approvedin 1986 by the IFCC (2).

l .2 Tonometry

The reference method for tonometry of blood makespossible the preparation of blood samples with knownpartial pressure of gases (CO2 and O2).

Definition

Tonometry designates the procedure of equilibratinga liquid sample with a gas of known compositionunder controlled conditions in order to establish aknown partial pressure (tension) of a gas in the liquid.The Instrument employed is called a tonometer.

MeasuringInstrument0

Referenceelectrode-

R l(Hg/HglCl1)·

1 —W

5mm·.· 10mm-*-»

~~~^

'Liquid/ liquidReservoir Junction

-Removable las selectrode

-Solution X

(Ag/AgCl)

4Mjjn*rreferencesolutionr

-Capillary

l l

Thermostated at 37·C (310,15 K)

Fig. 2. Schematic presentation of the pH measuring System.

i Eur. J. Clin. Chem. Clin. Biochem. / Vol. 29,1991 / No. 4

Methods of tonometry

Tonometry may be performed in two different ways.

a. Using a closed System, a gas and a liquid arebrought into contact; after equilibrium the result-ing composition of the gas phase is measured. Thismethod is laborious and time-consuming and can-not easily be adopted for routine work.

b. Using an open System, a gas mixture of knowncomposition flows through the System generatingidentical partial pressures of gases in the fluidsample and the gas mixture. In contrast with theclosed System this method eliminates gas analysisafter tonometry and has been chosen for the pres-ent reference method.

The requirements of Instrumentation and equipmenlare a gas mixture, a humidifier and a tonometer, thelatter two being thermostatted at 37 °C.

The primary gas mixtures of carbon dioxide, oxygenand nitrogen are gas mixtures prepared gravimetri-cally according to the definitive method from theeonstituent gases. The constituent gases should be ofsufficiently high purity and not contaminated withone of the other constituents of final mixture. Therelative inaccuracy of the composition of the gasmixture i. e. CO2 and O2 substance fractions, shouldbe less than ± 0.3% of the stated value.

256 Symposium: Rcference Methods in Clinical Chemistry

For certification of gas mixtures, a service is providedby national Institutes and by some industries.

Also gas mixtures prepared from pure gases withcalibrated gas mixing pumps meet the requirements.

In the hwnidifier the gas is preheated at 37 ± 0.1 °C,then saturated with water vapour by bubblingthrough water at the same temperature and the actualbarometric pressure, to prevent evaporation or con-densation of water in the sample. The gas is admittedto the equilibration chamber of the tonometer withoutcooling or alteration of pressure.

A film type tonometer which should provide a mini-mum volume of 3 ml of tonometered blood has beenchosen for the reference method.

In practice, a Laue tonometer (4) from the firmEschweiler (Kiel, Germany) is, according to our ex-perience, an adequate System. This tonometer is pref-erable to the IL-tonometer (Instrumentation Labo-ratory SpA, Milan, Italy) which has a temperaturegradient of 0.3 °C between the sample and the water-bath (5). In addition this System needs much moregas mixture (5).For further details reference should be made to thein 1988 approved IFCC document on tonometry (3).

The tonometered blood samples are secondary refer-ence materials which should be employed for qualitycontrol when measuring the partial pressure of carbondioxide and oxygen in blood. This allows both theaccuracy and the precision of the instrumental meth-ods to be evaluated. The method may also be appliedfor determining acid-base and oxygen transport pa-rameters e. g. pso.

2. Electrolytes

There are several positive features of ion-selective elec-trodes that lead to the conclusion that they haveimportant roles to play in clinical chemistry.

1. Ιοη-selective electrodes sense the electrochemicalactivity of the ion species in plasma water ratherthan its substance concentration in whole plasma.The activity in plasma is probably the more im-portant from a biochemical or physiological pointof view.

2. Ιοη-selective electrodes enable a direct measure-ment of electrolyte levels to be carried out in smallsamples of undiluted serum, plasma, whole blood,and urine which is a great advantage in paediatricand intensive care where rapid analyses are essen-tial.

3. Ιοη-selective electrodes enable us to monitor theions under investigation continuously with extra-corporeal devices or with intravascular catheter tipdevices in a patient in vivo, for example during thecourse of an Operation.

Consequently, measuring Systems with ion-selectiveelectrodes for sodium, potassium, ionized calcium andrecently also for lithium became routinely available,but not without problems in their clinical chemicalapplication (6, 7).

Ιοη-selective electrodes for sodium and potassiumions may be used on two different ways, namely inundiluted or in diluted samples. These detenninationshave been called repectively direct and indirect poteri-tiometry, but this terminology is incorrect; in bothcases the measurement is performed directly in diffef-ent samples.

When the sample has been highly diluted the totalconcentration expressed in mmol/1 serum or plasmashould be measured at the end. The ionic strength ofthe dilution will be equal to that of the calibrator.This type of measurement with ion-selective electrodesusually gives no problems, in contrast to measure-ments in undiluted samples.

The problems of measurement in undiluted specimenswhich are arising with the introduction of ioii-selectiveelectrodes for sodium, potassium and also calciumions in the clinical l borabories are:

a. What should be reported?

History was largely determined that ions such ssodium are usually measured on prediluted specimensusing flame photometry. In principle the flame pho-tometer measures the total substance concentrationof the ion and the ion-selective electrode measures theactivity of the free ion in the water phase. Hencedifferences between results have been obtained byseveral authors. It is therefore necessary to decidewhether we want to report activity or substanceconcentration in either the total volume of plasma(rhol/1) or only in the water phase of plasma (mol/1),the latter being equivalent to molality (mol/kg).

b. Differences in measuring technique

Different Instruments on the market for the measure-ment of sodium, potassium and ionized calcium byion-selective electrodes on diluted specimens give dif-ferent results, indicating a problem of measurementtechnique i.e. calibration Solutions, selectivity andsensitivity of the electrodes, measuring thne, liquidjunction potential and temperaiure.

Eur. J. Clin. Chem. Clin. Biochem. / Vol. 29,1991 / No. 4

Maas: IFCC reference methods for pH, gases and electrolytes in blood: reference materials 257

It will be clear that there is a need to standardize themeasuring technique, to develop a reference methodfor ionized calcium and to prepare reference material.

To considerthe above-mentionedproblems, a EuropeanWorking Group on lon-Selective Electrodes has beenset up by the International Federation of ClinicalChemi$try, consisting of Clinical Chemists and man-ufacturers with a common interest in exchanging ideasand Information on ion-selective electrodes and tryingto solve the problems by the increasing use of thisnew technology. Their goal is to gather data concern-ing the clinical chemical application of ion-selectiveelectrodes and to obtain consensus concerning defi-nitions, terminology, sampling, calibrating of Instru-mentation, quality control and reference intervals.

The group started in 1982 and has held meetings inOslo (1983), Oxford (1984), Helsinki (1985), Graz

l (1986), Danvers (1987), Stresa (1988) and Monterey! (1990), in which European, American, Australian and

Japanese invited scientists both from profession andindustry participated.

2.1 Sodium and potassium

For many years, results of determinations of sodiumand potassium ions in physiological fluids have beenexpressed in terms of the substance concentration oftotal sodium and potassium ion (mmol/1). Hence theuse both of substance conceijtrations of sodium andpotassium and of their reference intervals is firmlyestablished in clinical Interpretation and practice. Fur-thermore, it can be envisaged that analytical Systemswhich measure substance concentration, such äs flamephotometers, will continue to be used alongside ion-selective electrode determinations in undiluted plasmafor the forseeable future. In consequence, the conven-tion proposal represents a pragniatic compromisewhich attempts to facilitate the introduction of ion-selective determinations of sodium and potassium ionconcentrations in whole blood or diluted plasma intoclinical practice while minimising the risk of clinicalmismterpretation.

A convention for sodium and potassium is proposedwhereby, for routine clinical purposes:

A. Results of ion-selective electrode determinationsöf sodium and potassium in whole blood andundiluted plasma should be reported in terms ofconcentration (mmol/1).

B. Results of measurements on Standard normal spec-imens should conform exactly with those obtainedby flame photometry on the same speciinens.

Eur. J. Clin. Chem. Clin. Biochem. / Vol. 29,1991 / No. 4

C. Standard plasma specimens are herein defined äshaving mass concentration of plasma water of 0.93kg/l, plasma bicarbonate concentration of 24mmol/1, plasma pH of 7.40, and concentrations ofalbumin, total protein, cholesterol and triacylgly-cerol within the reference ränge for healthy sub-jects.

For practical purposes, it is satisfactory if the valuesof the flame photometer and the ion-selective elec-trode for normal plasma are equivalent. This may beachieved in several ways (6, 7):

1. With pooled normal plasma samples.

2. With some spiked samples of which the concentra-tion of sodium or potassium has been increasedby adding known amounts of sodium Chloride andpotassium and decreased by partial Ultrafiltration.

3. With a correlation study between ion-selective elec-trode and flame atomic emission spectrometry of100 — 200 sera samples with electrolyte concentra-tion covering the clinical ränge and fulfilling thenormal conditions.

When this convention is used, results reported by ion-selective electrodes in undiluted normal plasma areequivalent to the substance concentration of the ion.However, in samples with normal plasma water massconcentration or with abnormal concentrations ofcomplexes of the ion, results are numerically differentfrom true substance concentration.

Table l shows calculated examples of the differencesbetween the substance concentration obtained byflame photometry and the results obtained with ion-selective electrode when the ionic strength or the massconcentration of water varies. When the appropriateconversion factor (1.27 mol/1) is applied, the ion-selective electrodes ireading, "clNa", is the same äs theresult obtained by flame photometry for normalplasma. With the same factor, small discrepancies areobserved when the ionic strength of the plasma varies,because the activity coefficient of Na* is dependenton the ionic strength. However, the discrepancies arevery small, even with extreme pathological variationsof ionic strength, and the discrepancies do not haveany practical consequences. Large differences ariseonly when the water concentration of the plasmachanges, for example, because of severe hyper- orhypoproteinaemia or severe hyperlipaemia. In the ex-ample where the mass concentration of water hasdecreased by 15% to 0.80 kg/l, the differences betweenthe flame photometer and the ion-selective electroderesults is 20 mmol/1.

258 Symposium: Reference Methods in Clinical Ghemistry

Tab. 1. Calculated cxamples.

Spccimen FlameC'tNa«mmol/1

lon-selectiveelectrodeöNa+ * 10"3

"c IN.,*mmol/I

DifferenceCtNa — "c"lmmol/1

Normal plasmayNa+ = 0.747

= 0.933 kg/l/ = O.l60mol/kg

HypernatraemiaTNa, = 0.737

= 0.933 kg/l/ =0.190 mol/kg

HyponalraemiayNa+ = 0.760

= 0.939 kg/l/ =0.130 mol/kg

HyperlipaemiayNa+ = 0.747

= 0.800 kg/l/ =0.160 mol/kg

140

170

110

120

110.2

132.0

110.2

140

167.7

111.9

140

0.0

-h 2.3

- 1.9

-20.0

* Calculated concentrations obtained by multiplying ion-selective electrode öNa+ values 1.27.

Physiologically and biochemically, results obtained byion-selective electrodes are to be preferred becausethey more accurately reflect the pathophysiologicalStatus of these ions in plasma water and are thus morerelevant clinically than those reported by flame pho-tometer. Actually, therefore, this is a problem of theflame photometer and not of the ion-selective elec-trode. It is recommended that results obtained withseverely hyperlipaemic sera always be accompaniedby an explanatory remark, i.e. that the flame pho-tometer result is spuriously low ("pseudohyponatrae-mia"), but that the ion-selective electrode result is notaffected.

2.2 Calcium

The relationship between ionized calcium and totalcalcium is complex and variable for each sample.Ionized calcium reflects better the homeostasis ofcalcium than total calcium, due the dependence oftotal calcium on proteins and nutrition. Total calciummay underestimate hypercalcaemia and overestimatehypercalcaemia. Therefore ionized calcium is clini-cally the relevant value.

Convention of reporting ionized calcium

In principle, ionized calcium measurements could bereported äs concentration (mmol/1 plasma water), mol-

ality (mmol/kg plasma water) or äs activity. To avoida proliferation of units, the convention is herebyadopted of reporting ionized calcium measurementsäs concentration expressed äs moles per liter plasmawaier and not per liter plasma. Therefore the measuredcalcium ion activity should be converted to substänceconcentration of calcium ion by a factor which isprimarily dependent on the reciprocal molal activitycoefficient of calcium ion of normal plasma.

For practice the calcium ion-selective electrode shouldbe calibrated in terms of concentration. The compo-sition of the calibration Solutions is chosen such thatthe activity coefficient of the calcium ion is assumedto be identical both in calibration Solutions and nor-mal plasma i. e. by convention 7 == 0.160 mol/kg.

The pfoposed recommendations for a referencemethod of ionized calcium are analogous to pH. Specrifications and properties of calcium-selective and ref-erence electrodes, the salt bridge, the cell geometryand millivoltmeter äs well äs primary and secundarycalibration Solutions (7= 0.160 rnol/kg) and the pro-cedure of measurement are provisionally described(8).A prototype measuring cell System for the referencemethod for the measurement of ionized calcium inserum, plasma and whole blood is under developmentby Covington & Maas, and will be tested by an inter-national groüp öf scientists. v;

Eur. J. Clih. Chem. Clin. Biochem, / Vol. 29,1991 /No. 4

Maas: IFCC reference methods for pH, gases and electrolytes in blood: reference malerials 259

Preanalytical factors of influence

The concentration of ionized calcium in blood, plasmaor serum may be influenced by pH changes of thesample, calcium binding by heparin and dilution bythe anticoagulant solution. Therefore, and in view ofthe increasing interest in the determination of ionizedcalcium in the clinic, recommendations for optimalconditions for the collection and processing of bloodspecimens for both routine measurements and forestablishing reference ranges are also in preparation.

For users of ion-selective electrodes who wish to ex-press results of ion measurements in terms of activity,values of activity coefficients will be recommended inthe near future, thereby providing calculated activityvalues of sodium, potassium and calcium in a numberof specified aqueous Solutions which are suitable forInstrument calibration.

3. Quality Control

3.1 pH and blood gases

Generally, a quality control material is a stable andhomogeneous material which can be amply supplied.One or more properties of it are determined by ref-erence methods. In blood gas and pH analysis, acid-base and oxygen buffering characteristics are mostiipportant. Thus good quality control material for pHand blood gases prepared from blood should havenormal acid-base properties, oxygen buffer capacityand oxygen haemoglobin (Hb) equilibrium curve. Itshould also have a low and stable fraction of hae-miglobin (Hi). It should not clog tubing or polluteelectrode membranes.

The ideal quality control material for blood gas andpH measurements is the sample itself with referencevalues obtained by using the reference methods forpH and tonometry.

The next best material for quality control purposesin blood gas chemistry is fresh, whole human blood,freed of platelets and leukocytes to reduce the rate ofoxygen consumption and carboii diöxide productionin the sample. This is accomplished by centrifugingthe blood, removing the platelet and leukocyte layerby aspiration and resuspending the erythrocytes inthe plasma. The reconstituted blood sample is thenequilibrated in a tonometer to establish a known/7CO2and pO2. The pH of this tonometered blood must bemeasured separately, using the IFCC approved (1986)method for pH measurement in blood.

This material is designated a primäry quality controlmaterial.

Storage of this material in a gas-tight syringe in icedwater is possible for up to 2 hours (5, 9).

However, the disadvantages of blood are its pooravailability, its instability due to metabolism and thefact that it may constitute a biohazard to the user.The inconvenience and limitations of blood use ledto Substitution with various serum and aqueous prep-arations äs control matrices.

At present, quality control materials are available withmatrices that are different from blood. The productsare divided into several groups according to compo-sition (2):

1. Aqueous electrolyte Solutions.

2. Aqueous electrolyte solution containing ethyleneglycol.

3. Aqueous electrolyte solution containing protein.

4. Aqueous electrolyte solution containing fluorocar-bon Suspension.

5. Aqueous electrolyte solution containing stabilizedred blood cells.

6. Frozen haemolyzed red blood cell solution.

7. Aqueous electrolyte solution containing humanHb.

The aqueous materials have advantages such äs along shelf-life and their availability in ampoules whichare ready for use. However their disadvantage is pooroxygen buffering, although the fluorocarbon materialhas an oxygen buffering capacity which is, at a single/? 2, equivalent to that of blood. Moreover, certainProblems of pH measurement are not reproduced byaqueous materials without protein.

Stabilized blood preparations have a short shelflifeand a poor and changing oxygen buffering capacity,due to methaemoglobin formation. The haemoglobinsolution has the disadvantages that values for pCO2and oxyhaemoglobin concentration tend to be de-creased, while values for pH, pO2 and methaemoglo-bin concentration are increased. Oxygen buffering,however, is close to that of blood, although poorer inthe physiological pO2 ränge of 7 —20 kPa. Hence, thedifferences between the control matrix and the pa-tient's blood specimens introduced a new problem:non-identical behaviour.

When the assigned pH, pCO2 and pO2 values of theseproducts have been determined using the referencernethod for pH in blood and the present referencemethod for tonometry of blood respectively, thesematerials are designated secondary quality control ma-terials. If the assigned pH, pCO2 and pO2 values of

Eur. J. Clin. Chem. Clin. Biochem. / Vol. 29,1991 / No. 4

260 Symposium: Reference Methods in Clinical Chemistry

ihese products are established by routine measure^·ments on the blood gas Instruments themselves, theseproducts should be designated tertiary quality controlmaterials with expected values. These tertiary controlsdo not allow evaluation of the accuracy of the Systemalthough the precision may be properly assessed.

An ideal solution for quality control of blood gasdeterminations should contain haemoglobin with anormal O2 affinity, a normal ctHb and a low :\ΓΗ»comparable to that of fresh whole blood. Also thehaemoglobin (Hb) solution must be free of erythro-cyte stromata and other precipitable matter duringstorage at 4 °C and use at 37 °C, and it should beeasy to prepare. If, in addition to the pH and gasvalues, the ctHb and the fractions of all Hb-derivativesare known, then such a Hb solution may be suitablefor quality control of Hb-meters and CO-oximeters.

Sprokholt (9 — 13) and Maas have described a stablestromafree haemoglobin solution which overcomes thedisadvantages of whole blood and is sufficiently sim-ilar to blood to neglect the matrix differences. Stroma-free haemoglobin solution may be buffered in Orderto obtain the desired acid-base properties. Stroma-free haemoglobin solution contains functional Hbwith the same oxygen buffering capacity s in normalblood; this is achieved by using modulators that alterthe oxygen affinity. A suitable solution for pH andblood gas control may be prepared by tonometry witha gas mixture of known CO2 and O2 content, and byreading the pH from a pH/log pCO2 curve previouslydetermined with the reference method for pH. Sinceprotein, Hb, is present, the pH measurement can beproperly checked.

The stability of stroma-free haemoglobin sotutionmay be improved greatly by adding NADH, whichacts s a Substrate of the methaemoglobin-reductaseSystem and keeps the met-Hb (Hi) concentration low.

3.2 Electrolytes

Quality control of ion-selective electrodes has beenperformed generally with protein-free Solutions. How-ever the influence of protein both on the ion-selectiveelectrode itself and on the diffusion potential at theboundary of the salt bridge, s well s the small mV-interval for the whole pathophysiological r nge of theNa+- and Ca2+-ion-selective electrode, fequire the useof protein-containing control material.

The scheme of the liquid junction in figure 3 showsthat a protein-free solution is capable of monitoringonly the diffusion potential of the calibrator, and thata solution containing protein monitors that of theplasma or serum sample. This has been demonstrated

BoundaryU

3

2

Io3-1

-2

-3-4

Concentrated.KCI

"* ConcentratedKCI

ΤΓΤΤ

• "· *·' '

$\

:·.*.·*

Sample

·<ι

Sample

Jcont

(

Liquid junction potentialPlasma; Serum

, Liquid junction potentialCalibrator

^Protein-based electrolytecontrols meet plasma, serum

=sr*^ Protein-free electrolytecontrols meet calibrators

Fig. 3. Schematic presentation of the pirotein^effect on the liq-uid junction.

experimentally, and the control results obtained withan analyser for ionized calcium are presented in figure4 (14). Two protein-free calcium Solutions (l mmol/1and 2 mmol/1) were measured daily and showed co -stant values. Two ser m-based controls measured si-multaneously, however, showed a sharp rise in value.Patient or other protein-containing s mples wo ldhave shown a similar rise. The similarity of the cali-brator and the protein-free control solution explainswhy this liquid junction effect was not observed usingthe protein-free control solution. This disturbance ofthe liquid junction occujrred because the serum matrixwas different. This favours the use of protein-con-taining Solutions in the quality control process ofionized calcium analysers using ion-selective elec-trodes.

2.0

1.9

1.81.7

1.6

1,5

1,4

1.3

1.2

1.1

1.0

0.9

- o Q

-B-B-B H H

* .._ Λ

^J" *__ . ±

ι ι ι ι ι ι ι ι

.: .·_xv 1 1 1 1 1 1_

,t[d]Fig. 4. The concentration of calcium ions (cc^+) measured on

an ionized calcium analyser with protein-free Solutions(A and B) and serum-based controls (F and H).

Eur. J. Clin. Chem·. Clin. Bioeherri. / Vol. 29,1991 / Np. 4

Maas: I FCC reference methods for pH, gases and electrolytes in blood: reference malerials 261

Standard protein-containing reference materials forcontrol ofsodium and potassium ion-selective electrodesare under development in the USA by NIST (15). InJapan, Daiichi Pure Chemicals Co., Ltd, in coopera-tion with the Chemicals Inspection & Testing Institute(CITI), has been manufacturing a freeze-dried serumfor control of sodium and potassium ion-selectiveelectrode measurements (16). In the Netherlands, sta-

ble bovine albumin-containing controls for pH, bloodgases and electrolytes are already commercially avail-able from EURO-TROL B. V. (Wageningen, TheNetherlands). In Europe, plans are being made by theBureau of Reference of the Community at Br sselsto standardize the ionized calcium determination inblood plasma by developing both a reference methodand protein-containing control material.

References1. Durst, R. A., Siggaard-Andersen, O. & Maas, A. H. J.

(1982) Reference and quality control materials for pH andblood gases. In: Proceedings of the 5th meeting of the IFCCExpert Panel on pH and Blood Gases held at Copenhagen,June 16-18. 1981 (Siggaard-Andersen, O., ed.) pp. 89-100.

2. Maas, A. H. J., Weisberg, H. F., Burnett, R. W., M ller-Plathe, O., Wimberley, R D., Zijlstra, W. G., Durst, R. A.& Siggaard-Andersen, O. Approved IFCC Methods. Ref-erence method (1986) for pH measurement in blood. J.Clin. Chem. Clin. Biochem. 1987,25,281 -289. Clin. Chim.Acta 1987,165, 97-109. Labmedica 1986/1987, 3, 33-37.Biochim. Clin. 1988, 72, 241-249.

3. Burnett, R. W., Covington, A. K., Maas, A. H. J., M ller-Plathe, O., Weisberg, H. F., Wimberley, R D., Zijlstra, W.G., Siggaard-Andersen, O. & Durst, R. A. IFCC Method(1988) for tonometry of blood: Reference materials forpCO2 and pO2. J. Clin. Chem. Clin. Biochem. 1989, 27,403-408. Ann. Biol. Clin. 1989, 47, 373-376. Biochim.C n. 1989, 73, 945-949. J.I.F.C.C. 1989, 7, 78-81. J.Biomed. Lab. Sei. 1989, 2, 185-192.

4. Laue, D. (1951) Ein neues Toiiometer zur raschen quili-brierung von Blut mit verschiedenen Gasdrucken. Pfl gersArchiv 254, 142-143.

5. Sprokholt, R. & Maas, A. H. J. (1981) Some experienceswith commercially available tonometers. In: Proceedings ofthe 6th meeting of the IFCC Expert Panel on pH and BloodGases held at Groningen, August 23—25, 1981 (Oeseburg,B. & Zijlstra, W. G., eds.) pp. 28-49.

6. Maas, A. H. J., Siggaard-Andersen, O., Weisberg, H. F. &Zijlstra, W. G. (1985) Ion-selective electrodes for sodiumand potassium: A new problem of what is measured andwhat should be reported. C n. Chem. 37, 482—485.

7. Maas, A. H. J. (1988) Proposed recommendations on ion-selective electrode deterrninatipns of the substance concen-tration of sodium, potassium and ionized calcium in serum,plasma or whole blood. In: Clinical Chemisiry, AnOverview.Proceedings of the 13'th International Congress of ClinicalChemisiry, and the 7th European Congress of Clinical Chem-istry, held at the Hague, the Netnerlands, June 28—July 3,1987 (den Boer, N. C., van der Heiden, C, Leijnse, B. &Souverijn, J. H. M., eds.) pp. 277-294.

8. Boink, A. B. T. J., Buckley, B. M., Chri^tiansen, T. F.,Covington, A. K., Maas, A. H. J., M ller-Plathe, O., Sachs,Ch. & Siggaard-Andersen, O. (1987) Reference method forthe determination of the substance cpncentration of ionizedcalcium in serum, plasma or whole blood. In: Methodologyand clinical applications of ion-selective electrodes. Vol. 8.Proceedings of the 8th meeting of the European WorkingGroup on Ιοη-Selective Electrodes held at Hotel RestaurantPfeifer, Graz, Austria October 2—4, 1986 (Maas, A. H. J.,Buckley, B., Marsoner, H., Saris, N.-E. L. & Sprokholt,R., eds.)pp. 39-62.

9. Sprokholt, R. (1987) Quality control in blood gas chemis-try. Thesis, Utrecht, The Netherlands.

10. Sprokholt, R., van Ooik, S., van den Camp, R. A. M.,Bouma, B. N., Zijlstra, W. G. & Maas, A. H. J. (1987)Quality control material containing hemoglobin for bloodgas and pH measurement: Preparation of stroma-freehemoglobin solution. Scand. J. Clin. Lab. Invest. 47, Suppl.188, 69-82.

11. Sprokholt, R., van Ooik, S., van den Camp, R. A. M.,Bouma, B. N., Zijlstra, W. G. & Maas, A. H. J. (1987)Quality control material containing hemoglobin for bloodgas and pH measurement: Improvement of the stability ofstroma-free hemoglobin solution. Scand. J. Clin. Lab. In-vest. 47, Suppl. 188, 83-92.

12. Sprokholt, R., van Ooik, S., van den Camp, R. A. M.,Bouma, B. N., Zijlstra, W. G. & Maas, A. H. J. (1987)Preparing a quality control material for blood gases, pHand hemoglobin using stroma-free hemoglobin solution.Scand. J. Clin. Lab. Invest. 47, Suppl. 188, 93-99.

13. Sprokholt, R., van Ooik, S., van den Camp, R. A. M.,Bouma, B. N., Zijlstra, W. G. & Maas, A. H. J. (1987)Evaluation of a quality control material containing hemo-globin for blood gas and pH measurement. Scand. J. Clin.Lab. Invest. 47, Suppl. 188, 101-112.

14. Sprokholt, R., Boink, A. B. T. J. & Maas, A. H. J. (1988)Quality control of measurements of ionized calcium withion-selective electrodes. In: The 6th International Sympos-ium on Quality Control-Osaka 1987. Excerpta Medica, Ltd.Tokyo, pp. 61—71.

15. Koch, W. F. & Paule, R. C. (1989) Development of aStandard reference material for ISE measurement ofsodiumand potassium. In: Methodology and Clinical Application ofΙοη-Selective Electrodes. Vol. 10. Proceedings of the Wthmeeting of the European Working Group on lon-SelectiveElectrodes, held at Stresa, Italy, September 21 — 24, 1988(Maas, A. H. J., Buckley, B. M., Manzoni, A., Moran, R.F., Siggaardr Andersen, O. & Sprokholt, R., eds.) pp. 99-108.

16. Kuwa, K. & Yasuda, K. (1988) Method Evaluation toElectrolyte Measurement with ISE and Present Status inJapan. In: Methodology and Clinical Application of lon-Selective Electrodes. Vol. 9. Proceedings of an InternationalSymposium on the measurement of blood electrolytes orga-nized by Electrolyte/Blood Gas Division of AACC and theIFCC European Working Group on Ιοη-Selective Electrodes,heldat Danvers M A (USA), September 23-25, 1987 (Bur-ritt, M. F., Cormier, A. D., Maas, A. H. J., Moran, R. F.& O'Connell, K. M., eds.) pp. 67-81.

Prof. Dr. A. H. J. MaasZwaardemakerlaan 45NL-3571 ZB Utrecht

Eur. J. Clin. Chem. Clin. Biochem. / Vol. 29,1991 / No. 4