CLIN. CHEM. 35/3, 400-404 (1989)

400 CLINICAL CHEMISTRY, Vol. 35, No. 3, 1989

Differencesbetween Capillaryand Venous Blood-AlcoholConcentrationsas a Functionof Timeafter Drinking,with Emphasison SamplingVariationsin Leftvs RightArmA. W. Jones,’ K-AJOnsson,2 and L Jorfeldt3

Twelve healthy men drank 0.8 g of ethanol per kilogramofbody weight during 30 mm after an overnight (10 h) fast. Atnine exactly timed intervals (30-390 mm after the start ofdrinking), blood was sampled through indwelling catheters incubital veins on the left and right arms. Immediately thereaf-ter, capillary blood was sampled from fingertips on the leftand right hands. The blood ethanol concentration (BAC) wasdetermined by headspace gas chromatography. The SD foralcohol determinations in venous blood, including the left vsright arm sampling variation, was 30 mg/L (range 8.3-83 mg/L), whereas for capillary blood the SD was 35 mg/L (range11-60 mg/L). This difference much exceeded the purelyanalytical errors: SD = 2.67 mg/L for venous blood and 14.2mg/L for fingertip blood. During the first 60 mm after thesubjects started to drink, capillary BAC exceeded venousBAC, the mean difference at 30 mm being 136 mg/L (range36-216 mg/L). In the postabsorptive state later than 60 mmafter drinking, venous BAG exceeded capillary BAC [meandifference 58 mgIL (range 0.0-170 mg/L)], the values forvenous and capillary BAC crossing 37 mm (range 6-77 mm)after the end of drinking. Apparently, the source of bloodanalyzed, venous or capillary, must be considered in clinicalpharmacokinetic studies of ethanol.

Most clinical biochemistry tests involve the use of wholeblood, plasma, or serum as the biological matrix. For somesubstances, such as glucose and lactate, the results dependin part on whether venous blood or capillary blood isanalyzed (1, 2). The concentration of ethanol in blood israrely reported with reference to the source of blood ana-lyzed, i.e., artery, capillary, or vein. Ethanol distributes intothe total body water without binding to plasma proteins andis cleared from the body chiefly by metabolism in the liver(3). In routine analytical work, scant attention has hithertobeen given to possible differences in concentration of ethanolin different vascular compartments, or to the significancethis might have in medicolegal practice. The precision ofblood-ethanol analysis is usually well defined, being derivedfrom the variance of replicate determinations, with aliquotstaken from the same pool of blood in one single collectiontube submitted to the laboratory for analysis (4).

Automated methods of forensic blood-ethanol analysishave coefficients of variation (CV) of <1%, so instead ofattempting to reduce analytical errors of the method evenflu-ther, investigators should turn their attention to theimportance of sampling variations (5). Statutory limits forblood-alcohol concentration (BAC) in motorists are rigorous-ly enforced, regardless of the way in which the forensic bloodspecimens were obtained (6). The components of variationintroduced during sampling are generally considered to benegligible or zero. The widespread use of per se statutory

‘Department of Alcohol Toxicology, National Laboratory of Fo-rensic Chemistry, and Departments of 2Clinical Pharmacology,3Climcal Physiology, and2 Internal Medicine, University Hospital,581 85 Linkoping, Sweden.

ReceivedOctober 28, 1988; accepted December 21, 1988.

BAC limits as evidence of impairment means that smallincrements in BAC make the difference between punish-ment or acquittal in borderline cases (6). The significance ofvariations in blood source and sampling site, in relation tothe concentration of alcohol, must be carefully documented.

The aim of the present work was to establish the timecourse of alcohol concentrations in specimens of venous andcapillary whole blood during absorption, distribution, andelimination stages of ethanol metabolism. We obtainednear-simultaneous specimens of cubital-vein blood fromboth arms and capillary blood from fingertips on both hands.Thus we could resolve the magnitude of variation caused bythe sampling procedure from that of the purely analyticalerrors.

Materials and Methods

Subjects and Conditions

The 12 healthy men who took part in this study as paidvolunteers were all medical students accustomed to moder-ate drinking: mean age 27.8 y (range 21-43), mean bodyweight 74.8 kg (range 71-81), and mean height 183 cm(range 180-193). The experiments started at about 0815hours after an overnight fast (10 h). Indwelling catheterswere inserted in a proximal direction into medial cubitalveins on the right and left arms. The catheters were keptpatent by flushing with isotonic saline containing a fewunits of heparin. Specimens of venous blood were collectedinto 5-mL Vacutainer Tubes (Becton Dickinson, Rutherford,NJ 07070) containing 20 mg of NaF and 75 mt. units of

sodium heparin.The subjects drank 0.80 g of ethanol per kilogram of body

weight during 30 mm. The drink was prepared from 950mJJL ethanol reagent, which we diluted fourfold with anorange-flavored drink to make a “cocktail.” Two or threesubjects took part in each experimental session. Drinkingwas begun by successive subjects at 15-mm intervals, start-ing at about 0900 hours.

Procedures

Blood sampling. Specimens of venous blood were obtainedat exactly 30, 60, 90, 120, 150, 180, 210, 270, 330, and 390mm, timed from the start of drinking, with the subjectslying quietly on a bed. Two assistants drew venous blood atexactly the same time from left- and right-arm cubital veins.Immediately thereafter (within about a minute), capillaryblood was obtained from a fingertip on the left and righthands. Getting samples of capillary blood took slightlylonger than venous-blood sampling. After the fingertipswere pricked with sterile mini-lancets, the free-flowingblood (100 mg) ifiled Widmark S-shaped capillary tubes, theopen ends of which were then closed with small rubberstoppers. The inside walls of these glass tubes are coatedwith fluoride and oxalate as preservative and anticoagulant.This sampling protocol gave us 10 specimens of venous bloodand 10 capillary blood specimens from the left and rightarms of each test subject.

CLINICAL CHEMISTRY, Vol.35,No. 3, 1989 401

Determination of ethanol in blood. The concentration ofethanol in venous whole blood was determined by headspacegas chromatography (HS-GC) after 11-fold dilution with anaqueous solution of n-propanol (80 mgfL), which served asinternal standard. The contents of the Vacutainer Tubeswere mixed for about 15 mm on a Rotamix mixer before weremoved 100-pL aliquots for dilution with internal stan-dard, using a fixed-volume dilutor dispenser. The dilutedblood was dispensed directly into 22-mL headspace vials,which we then closed with a rubber septum and made air-tight with a crimped-on aluminum cap, in preparation foranalysis.

Before analysis of capillary blood, we removedthe rubberstoppers from each Widniark tube and recorded the weightof the tube plus blood to the nearest milligram. The bloodwas then ejected from the Widmark tube into a small testtube containing exactly 200 pL of distilled water. We thenreweighed the empty Widmark tubes and calculated theamount of blood taken for analysis and the dilution factor.After this initial dilution, we again diluted the hemolyzedblood specimen 11-fold with n-propanol, exactly the same asfor venous blood specimens. The capillary blood specimenswere stored at 4#{176}Cuntil analysis the next day. The 11-folddilution of whole blood before analysis helps to abolish anyeffects of ethanol on the matrix during the headspaceequilibration procedure, so that aqueous ethanol standardscan be used for calibration.

Gas-chromatographic analysis. We used a Sigma 2000 gaschromatograph with flame ionization detector and HS-100headspace analyzer (all from Perkin-Elmer Corp., Norwalk,NJ) for measuring blood ethanol. The results were recordedby an LCI-100 electronic integrator (Perkin-Elmer). Thestationary phase was Carbopack C (80-100 mesh) coatedwith 0.2% Carbowax 1500 and held in a 2 m x 3 mm (i.d.)glass column maintained at 110 #{176}C.Hydrogen (30 mL/min)and air (300 mL/min) were supplied to the detector; nitrogenwas the carrier gas (flow rate, 20 mL’min). Under theseconditions the retention times of ethanol and n-propanolwere 0.78 mm and 1.52 mm, respectively; the peak heightratios for ethanolln-propanol were used for quantitativeanalysis.

All specimens of blood collected in this project wereanalyzed with the same HS-GC unit. The instrument wasnot switched off during the period of the study, which lastedseveral weeks, and recalibration was unnecessary betweenruns. The GC unit was first calibrated by single-pointcalibration based on the average result of six replicatedeterminations of a 100 mg/dL ethanol standard to calculatea calibration factor. This was subsequently used to derive aresponse factor to calculate unknown blood ethanol concen-trations and standards of known concentration (50, 100, or150 mg/dL) analyzed after every tenth unknown bloodspecimen. These solutions (Merck Ltd., Darmstadt, F.R.G.)were supplied already diluted to the stated concentrations inflame-sealed borosilicate glass ampules. The difference inconcentration between the known concentrations of thestandard (target value) and the results determined byanalysis was defined as the “residual,” with an expectedvalue of zero. The average residual was taken as an index ofaccuracy (systematic error) of the blood-alcohol measure-ments. The response of the detector was linear over a wideconcentration range, and the regression line passedthroughzero.

Control analysis of capillazy blood. To check the resultsobtained from analysis of capillary blood against those of

venous blood and to assess the precision of the methodwithout the influence of left vs right arm sampling differ-ences, we designed a small experiment. Empty Widmarkcapillary tubes were ifiled in duplicate with blood takenfrom 20 different specimens in Vacutainer Tubes, gatheredduring this study. These duplicate determinations wereused to get a measure of the analytical error inherent incapillary blood assays by eliminating biological samplingvariations. The venous blood specimens used to fill thecapillaries were also analyzed in duplicate within the samerun.

Statistical analysis. The time course of changes in venousand capillary BAC was plotted for each subject, and thepeak BAC and the time elapsed from end of drinking to thepeak concentration was read from the curves. The times atwhich the mean capillary BAC (left and right hands)exceeded that of the mean venous BAC (left and right arms)were noted for each subject. Differences between left armand right arm, left hand and right hand, and venous vscapillary BAC were evaluated by Student’s t-test for intra-individual comparisons. Analytical variations were calcu-lated, as usual, from differences between duplicate determi-nations on the assumption of a gaussian distribution ofrandom errors. The SD of a single determination wasworked out from the variance of this mean differencebetween duplicates. Variances were compared by F-test (7).

Results

Accuracy and Precision

Accuracy of the gas-chromatographic method. The meandifference (residual) between concentrations of ethanol de-termined in the control standards and the target values of50, 100, and 150 mg/dL was -0.137 (SD 0.625) mg/dL. Thedifference from zero was not statistically significant (P<0.05). These results were derived from 72 determinationsmade during the entire time that blood samples were beinganalyzed. Systematic errors of the HS-GC method aretherefore negligible.

Precision of analyzing alcohol in venous and capillaryblood. The mean difference between duplicate determina-tions of venous blood was 0.11 (SD 0.379) mg/dL, andtherefore a single assay has an SD of 0.267 mgldL. Themean difference between duplicate capillaries filled withvenous blood from 20 Vacutainer Tubes was 0.49 (SD 2.02)mg/dL. The SD of a single determination was therefore 1.42mg/dL. The precision of analyzing venous blood is considera-bly higher than for capillary blood (F = 18.9, P <0.001).

Precision of alcohol analysis, including sampling varia-tions. The differences in concentration of alcohol in bloodfrom left and right arms were not significant between 30and 390 mm after drinking. This held for both venous andcapillary specimens, as shown in Figure 1. The SD of theanalysis was highest during the absorption phase 30-90mm after drinking. The 30-min venous blood specimen hadan SD of 8.3 mg/dL compared with 1.6 mg/dL at 90 mm,when BAC was similar, and 1.0 mgldL at 390 mm, when themean BAC was lower. The difference in variance between30-mm and 90-min samples was statistically significant (F= 26.9, P <0.001). The maximum SD for capillary blood was6.0 mg/dL, which occurred at 60 mm after drinking, com-pared with 3.6 mg/dL at 90 mm and 1.1 mg/dL at 390 mm.The difference in variance between the 30-min and 90-minsamples was not significant (P >0.05).

ences are plotted as a function of time in Figure 4. The meancrossover point for capillary and venous BAC was 37 mmafter the end of drinking (range 6-87 mm), which corre-sponds to 67 mm (range 36-117 mm) from the start ofdrinking (Figure 4).

Discussion

Impairment by alcohol is a common cause of accidents atwork and on the highway (8), and reliability of chemicaltests for intoxication is often the subject of much controversyand debate (9,10). Statutory BAC limits for drivers refer towhole blood as the biological specimen, but the source ofblood-i.e., artery, capillary, or vein-is usually not definedprecisely (6).

Looking at the forensic alcohol literature, one finds thatthe source of blood was recognized early as a complicatingfactor that might account for discrepancies in reported BACresults (11). Specifically, large differences were observed ifvenous BAC was compared with arterial BAC or with theresults obtained indirectly from analysis of breath whenethanol was being absorbed from the gut (12). But thisimportant forensic science issue has received scant attentionin recent years, and it is widely understood that, in thepostabsorptive state, the capillary, venous, and arterialBAC’s are approximately equal and that any differences arenegligible or zero. A paper by Sedman et al. (13) questioned

.-. CAPILLARY BLOOD

mL

_I0

mgMt.

.t0

S-.5

<1 0

0-10 .10

0 2 3 5 6 7 h=S 0 2 3

TIME FROM START OF DRINKING

Fig. 1. Differencesin concentrationsof ethanol in blood as a function oftime afterdnnking: (left) venousblood taken from the rightand leftarmcubital veins, and (right) fingertipcapillary blood taken from flght andlefthands

Concentration-Time Profiles

Alcohol in venous blood. Figure 2 (left) shows the meanconcentration-time proffles of ethanol for venous bloodspecimens obtained from left- and right-arm veins. Thesemeasurements agree well on average, although five of the12 subjects had small but systematic differences in concen-tration between both arms. During the postabsorptive state,the mean left-right arm difference for these five subjectswas 1.4 (SD 0.89) mg/dL (P <0.01).

Alcohol in capillary blood. Figure 2 (right) shows themean concentration-time profiles of ethanol for fingertipcapillary blood obtained from the left and right hands. Insome individuals there were marked left vs right handdifferences in capillary BAC at certain sampling times, butthis might be associated with problems in getting adequateblood for analysis. The mean-value curves agreed well.

Capillary vs venous blood-alcohol. Figure 3 shows meanvenous BAC (left and right arms) and mean capillary BAC(left and right hands) plotted as a function of time afterdrinking. These time proffles differ, depending in part onsampling time after drinking. During rapid absorption ofalcohol from the gut, i.e., 30-60 mm after the start ofdrinking, the capillary BAC exceeded the venous BAC. Thecapillary-venous difference was 13.6 (SD 7.8) mg/dL at 30mm and 3.3 (SD 6.9) mg/dL at 60 mm after the start ofdrinking. Later than 60 mm, the curves for venous/capillarysamples crossed, and for the remainder of the eliminationphase the venous BAC exceeded the capillary BAC; between90 and 390 mm, the mean difference was 5.8 (SD 3.4) mg/dL.The differences between capillary and venous BAC weresignificantly different from zero except at the 60- and 90-mm sampling intervals. The mean intra-individual differ-

.-I.EFT ARM VEIN -..o LEFT HAND CAPILLARY

.-.RIGHT ARM VEIN - RIGHT HAND CAPILLARY

o---o VENOUS BLOOD

-J0I0(2-J

a00-Jcc

0 1 2 3 4 5 6 7 BhoursTIME FROM START OF DRINKING

Fig. 3. Concentration-time profiles of ethanol in capillary and venousblood from 12 fasting men afterthey drank0.8 got ethanol per kilogramof body weight

‘ 100

50

0

S4

000

.5

5.E.1 2 3 1. 5 6 7hours

TIME FROM START OF DRINKING

Fig. 2. Mean concentration-time profiles of ethanol (left)in venousblood forsamples simultaneously taken from the left and right armcubital veinsand in fingertip capillary blood (right) taken from the leftand righthandsThe tastingsubjectsdrank 0.8g ofethanol per kilogramofbody weight in 30 mm

012345678TIME FROM START OF DRINKING hours

Fig. 4. Differencesin concentrationof ethanolin capillaryand venousbloodas a function of time afterdrinkingThe mean differences werestatistically different from zero at all sampling timesstudied, exceptat 60 and 90 mm

MEAN DIFFERENCE SE.

402 CLINICAL CHEMISTRY, Vol. 35, No. 3, 1989

CLINICAL CHEMISTRY, Vol. 35, No. 3, 1989 403

this notion, but their findings have failed to influence bloodsampling routines or the interpretation of BAC results inforensic science practice. More recently, Martin et al. (14)compared venous and arterial BAC after subjects drank abolus dose of ethanol 5 h after their last meal. Clear-cutarteriovenous differences in BAC were reported duringabsorption of alcohol from the gut, but later on, during theelimination phase, fairly good agreement was found be-tween the source of blood and the alcohol concentration.

Ethanol is completely miscible with water and distributesinto the water compartment of body fluids and tissue (3).During absorption of alcohol from the gut, the arterial BACequilibrates with the water fraction of all body fluids andtissue according to an exponential function. The speed of thedistribution process is determined in part by the blood flow(F) and the size of the distribution volume (V), and the ratioV/F expresses the equilibration time constant. For the wholebody, F/V is about 0.07 L/(min x L of tissue), whichcorresponds to a time constant of 14 mm and a half-time of10 mm for the equilibrium process (15-17). This suggeststhat if BAC changes rapidly as compared with the speed ofequilibration of alcohol between blood and tissue water, asignificant arterial-venous concentration gradient shouldexist. Furthermore, the arteriovenous difference will dependon the rate of blood flow and the vascularity and mass oftissue involved. The significance of nonequilibrium betweenblood and tissue and fluctuations in blood flow on themagnitude of arteriovenous differences were analyzed indetail by Zierler (15). Except for situations with a restrictedskin perfusion, capillary fingertip blood will be more repre-sentative of arterial BAC. The different time profiles forethanol in capillary and venous BAC can be attributed, atleast in part, to the distribution kinetics of ethanol.

The large superficial cubital veins drain blood fromsurface tissue as well as deep tissue in the forearm andhand. The venous anatomy is not necessarily identical onboth sides, and the relative contributions of blood fromsurface and deep tissue might differ, depending on thesampling site on each arm. We found that five of 12 subjectshad small but systematic differences in BAC between sam-ples from their right- and left-arm cubital veins. Thesedifferences ranged from 0.8 to 2.4 mg/dL in the postabsorp-tive phase. One explanation for this might be differentperfusion-distribution volume ratios for the vascular areasbeing drained into the particular veins catheterized. Wefound a significantly higher variance in left vs right armdifferences 30 mm after drinking compared with 90 mm,presumably because the BAC changes much faster shortlyafter drinking (30-mm specimen) when alcohol is still beingrapidly absorbed from the gut. A different perfusion-distri-bution volume ratio on each side will exert a greaterinfluence at 30 mm after drinking than at 90 mm.

The BAC differences in various parts of the venous systemcomplicate the interpretation of results for legal purposes.The contribution of sampling errors to uncertainties inreported BAC has been touched upon by others but neverclearly established. Several studies from Germany havefocused on left vs right arm differences in concentrations ofethanol (18-21). One study (18) reported that there were nophysiological BAC differences between left and right armsand that the maximum deviation was only 1 mg/dL. Incontrast, Naeve et al. (19) reported significant left vs rightarm differences, which were most pronounced when alcoholwas being rapidly absorbed from the gut.

During the absorption phase of the BAC time proffle the

concentration of alcohol in the arterial blood is higher thanin the venous blood and the magnitude of the arteriovenousdifference depends in part on the rate of absorption ofalcohol from the gut: rapid absorption exaggerates thedifference; slow absorption minimizes it. During the post-absorptive state the alcohol concentration gradients be-tween blood in artery and vein are reversed and venousBAC now exceeds arterial and capillary BAC. The onset ofthe postabsorption period, as reflected in the time of cross-over for capillary and venous BAC, began at 36 to 107 mm(mean, 67 nun) after subjects began drinking and therefore6 to 77 mm (mean, 37 mm) after the drink was finished(drinking time, 30mm). A difference in the water content ofvenous and capillary blood might provide an explanation forthe different BAC in postabsorptive samples. However, onthe basis of blood hematocrit values and hemoglobin mea-surements, the differences in water were considered negligi-ble under proper conditions of sampling (12, 13). A recentpaper, however, reports a 3% higher hematocrit for capillaryblood as compared with venous blood (22).

When BAC measurements are used for legal purposes, itseems advisable to make an adjustment to the averageresult of several determinations to compensate for inherentanalytical and sampling variations. The amount subtractedwill depend on the accuracy and precision of the analyticaltechniques used and the experience of the analysts involved.The degree of confidence in the final prosecution BAC-whether set at 95%, 99%, or 99.9%-will also determine thesize of the subtraction term. These levels of probabilityimply that a person with a true BAC just below thestatutory limit for driving runs a 1 in 20, 1 in 100, or 1 in1000 chance, respectively, that the reported BAC exceedsthe true value. Results of BAC measurements for legalpurposes in Sweden are reported as not less than theprosecution value with 99.9% confidence. Litigation in driv-ing-while-intoxicated trials in Sweden and defense chal-lenges directed at the reliability of forensic alcohol analysisare virtually nonexistent.

The systematic differences observed between venous andcapillary BAC is a more difficult problem to tackle, becausethis depends in part on the phase of alcohol metabolismwhen the samples are taken. This will require some carefulconsideration by forensic experts called upon to evaluateblood-alcohol measurements for legal purposes. We concludethat both analytical precision and sampling-site variationsshould be defined precisely and the source of blood obtainedspecified on the analytical report.

This work was supported by grants from the Swedish MedicalResearch Council(nos. 7214 and 4139) and the County CouncilofOstergotland.We wish to thank our research nurses, Ingrid Petter-son and Susanne Tillmar, for expert technical assistance.

ReferencesL Williams DL, Nunn RF, Marks V. Scientific foundationsofclinicalbiochemistry. Vol. 1: analytical methods. London: WilliamHeinemann Medical Books Ltd, 1978.2. Larsson-Cohn V. Differences between capillary and venous bloodglucose during oral glucose tolerance tests. Scand J Clin Lab Invest1976;36:805-8.3. Kalant H. Absorption, diffusion, distribution and elimination ofethanol: effects onbiological membranes. In: Kissin B, Begleitter H,eds. The biology of alcoholism, Vol 1: biochemistry, New York:Plenum Press, 1971.4. Westgard JO, Carey RN, Wold S. Criteria for judging precisionand accuracy in method development and evaluation. Clin Chem1974;20:825-33.

404 CLINICALCHEMISTRY, VoL35, No.3, 1989

5. Dubowski KM. Manual for analysis of ethanol in biologicalliquids. Report No. DOT-TSC-NHTSAO-76-4, 1977. Washington,DC: U.S. Dept. of Transportation.6. Jones AW. Enforcement of drunk-driving laws by use of per selegal alcohol limits; blood and/or breath concentration as evidenceof impairment [Review]. Alcohol Drugs Driving 1988;4:99-112.7. Youden WJ, Steiner EH. Statistical manual of the Association ofOfficial Analytical Chemists. Washington, DC: Association of Offi-cial Analytical Chemists, 1977.8. Forney RB. Alcohol, an abused drug [Review]. Clin Chem1987;33:82B-6B.9. Simpson G. Accuracy and precision of breath alcohol measure-ments for subjects in the absorption state. Clin Chem 1987;33:753-6.10. Hume DN, Fitsgerald EF. Chemical tests for intoxication; whatdo the numbers really mean [Review]. Anal Chem 1985;57:876A-80A.11. Harger RN. Blood source and alcohol level; errors from usingvenous blood during active absorption. In: Havard J, ed. Alcoholand road traffic. London: British Medical Association, 1963:212-9.12. Forney RB, Hughes FW, Harger RN, Richards AB. Alcoholdistribution in the vascular system; concentrations of orally admin-istrated alcohol in blood from various points in the vascular systemand in rebreathed air during absorption. J Stud Alcohol1964;25:205-17.13. Sedman AJ, Wilkinson PK, Wagner JG. Concentrations ofethanol in two segments of the vascular system. J Forensic Sd1976;21:315-22.

14. Martin E, Moll W, Schmid P, Dettli L. The pharmacokinetics ofalcohol in human breath, venous and arterial blood after oralingestion. Eur J Clin Pharmacol 1984;26:619-26.15. Zierler KL. Theory of the use of arterio-venous concentrationdifferences for measuring metabolism in steady and non-steadystates. J Clin Invest 1961;40:2111-25.16. Rowland M, Tozer TM. Clinical pharmacokinetics; concepts andapplications.Philadelphia:Lea & Febiger, 1980:1-331.17. Jorfeldt L. Measurement of skeletal muscle blood flow inhumans: lethysmographicbolus and continuous infusion technique[Review]. Am J Cardiol 1988;62:25E-9E.18. Zink P, Blauth M. Zur Frage Beeinflussung der Alkoholkon-zentration im Cubitalvenenblut durch die Blutentnahmetechnik.Blutalkohol 1982;19:75-87.19. Naeve W, Koops E, Audrlicky I, Brinkmann B. VenoVenoseDifferenzen der Blutalkoholkonzentration nach “Sturatrunk.” Blut..alkohol 1971;8:451-6.20. GoetomzykJG, Reineke H, Neumann GK, Leithoff H. Bestehteine veno-ven#{246}seRechts-links-Differenz in der Blutalkoholkonzen-tration? Z Rechtsmed 1974;75:37-42.21. Teige K, Niermeyer H, Gerlach D, Bohn G, Brauer G. Unter-schiede der Blutalkoholkonzentration bei gleichzeitiger, quasi-kon-tinuierlicher Blutabnahme aus den korrespondierenden Venen desrechten und des linken Armes. Blutalkohol 1974;11:29-39.22. Dane LNW, Halvorsen S, Mathisen PM, Mironska K. A com-parison between haematological parameters in ‘capillary’ and ve-nous blood from healthy adults. Scand J Clin Lab Invest1988;48:723-6.