. . . DIE SUIDER-AFRIKAANSE TYDSKRIF VIR KRITIEKE SORG The "Shunt Fraction": Accurate calculation and a graphical method o f interpretation R I STEWART MBCha PhD IMEO). FCC? Division of Med ica l Phys io lo gy and Re sp ir at or y Unit , Un iv er si ty of Ste lle nbosch and Tygerberg Hospita l. lY9er berg. Repub lic of South Africa. Correspond ence to : Prof RI Stewart, Department of Medical Physiolog y and Bio che mistry , University of Stellenbosch, AO Box 63, Tygerberg 7505, Republic of South Africa. a therapeutic protocol (1) 18 SUMMARY A method o f calculating th e venous admixture frac- tion Is proposed which avoids Inaccuracy at low F P, values. Inview of the quantitatively unpredictable consequences o f increasing th e "true shunt .. when th e F,O, Is 1.0, It Issuggested that th e 0",/0. be mea .... ured at th e therapeutic F,O,. In estimating Ideal alveolar P O, ( P ,OJ all the variables should be accurately measured except PH,0 reliably obtained from a regression equatlon. Actual measurement o f haemogl obin saturation with oxygen an d carboxyhaemoglobinIsessential. In estimating Ideal haemoglobin saturation with oxygen (ScOJ th e P,O, should first be corrected to /n vivo arterial conditions of pH, base excess and temperature prior to estimat In g th e Seo,. This latter value should also b e correct e d f or the presence of carboxyhemoglobln both in vivo an d In th e standanl oxyhemoglobln eqUilibrium curve. Interpretation of changes In arterial 0 , concentratlon (CaOJ Is facilitated by Identification Cao, along the venous 0 , concentration (CvOJ isopleths on a graph of 0 , content vs 0",/0.. Introduction The "shunt fraction- is wid ely use d in the assess men t of hypoxaemia. In th e measurement of true shunts (OJOJ and of the total venou s admixtu re (O",/OJ (s hu nt and shuntl ik e effects of Icm ventila- tionlperfusion ratios an d ox ygen dif fus ion li mitation). er rors ma y be inlroduced by boIh the administralion of oxygen'" and faulty calculation". The following appa rently si mp le equation is used: ecO,-CaO, a...o. CcO -CvO , , where CcO, is thei deat, end capillary oxyge n co ntent, CaO, the arterial and CVO, the mi xed venous oxygen content respectively. The si mp licity of this eq uation beli es the comp le xities inhe re nt not only in the me as ur ement of Ca 0 2 and evo 2 • but al so in the estima te of GcO,:;. These considerations assume importance at lower values of inspired ox yg en con centr ation as may be relevant in eval uat io n of intra cardi ac shu ntsand in the estimateof venousadmixture in patients with respiratory failure. Although the latter group of patients are fre que ntl y tre ated wi th hig h concentrations of 0 , this may not be the case in the earl y or reco ve ry stag es of the illness and is cert ainl y not so in patients tre ate d for paraqu at poisoni ng. Re-arrangement of th e equation (1l. thus, CaO, eco, -(eco, -CvOJ a.../O, (2) re-e mpha sis es the importance of both the eco, and the CVO, in the analysis of the ~ / 0 1 as a dete rminant of the arteri al oxyg enati on. Failure to consider changes in th es e other determinants of th e Ca 0 2 may also res ult in erro neou s int erpr etat ion of serial estimates of the venou s admixt ure fractionS. Th es e cons iderations are cl ear ly of importance sinceinterpretatio n of the "sh urt fraction" is frequentl y hcif y debated by clinicians'''. Alth ough the concept of venou s admixture is based on a simple, two compar tment model of gas excha nge'. is extremely useful in the analysis of art eri al hypoxia . lis usei ulnes s is, however, depen dentupon an ac cu rate valu of the 0'61/0.. It is the aim of thiS commun ica tionto prop osean acc ura temethod of ca lculat ing th e a...1O, at ION concertrations of inspired oxygen. and to present a graph icalapproach to the irterp<etation of this derived value. Blood sampling conditions Arterial hypoxia is the r e s u ~ of both shunt an d shunt-li ke effec ts on pUlmonary gas ex ch an ge . and in fact the la tt er is probably more important'. is well -esta blis hed that admi nistr ation of 100% 0 , prior 10 bloodsampling res ul ts in a n unpredi ctable overes timate of the Irue shu nt fra ct ion 2 .5- 8 due to conversion gas exchange units wi th ICM' vertlationperfusion ra ti os 0//Q) 10 units with a V lO o f zero (true shunts). is logical, theretore. 10 measure the 0",,/0,. the ven ous admixt ure , at the F,O, of therapeutic choice . Th e pro ble m of interpr_ng seriaJ ven ousadmixturea ttained at different inspired oxygen concentrati ons may be partfy overcome by the analys is pre sen ted below. Volume 5. No. 2 1989
4
Embed
1.2 the Shunt Fraction - Accurate Calculation & a Graphical Method of Interpretation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
8/3/2019 1.2 the Shunt Fraction - Accurate Calculation & a Graphical Method of Interpretation
. . . DIE SUIDER-AFRIKAANSE TYDSKRIF VIR KRITIEKE SORG
The "Shunt Fraction":Accurate calculation
and a graphicalmethod ofinterpretationR I STEWART MBCha PhD IMEO). FCC?
Division of Medical Physiology and Respiratory Unit, University ofStellenbosch and Tygerberg Hospital. lY9erberg. Republic of SouthAfrica.
Correspondence to:
ProfRIStewart, DepartmentofMedical Physiology and Biochemistry,University ofStellenbosch, AO Box 63, Tygerberg 7505, Republic of
South Africa.
a therapeuticprotocol
(1)
18
SUMMARY
A method of calculating th e venous admixture frac-
t ion Is proposed which avoids Inaccuracy at low FP,
values. In view of the quantitatively unpredictableconsequences of increasing the "true shunt.. when
the F,O, Is 1.0, It Is suggested that the 0",/0. be mea....
ured at the therapeutic F,O,. In estimating Ideal alveolar PO, (P,OJ all the variables should be accuratelymeasured except PH,0 which Is reliably obtained froma regression equatlon. Actual measurement ofhaemoglobin saturation with oxygen and carboxyhaemoglobin Is essential. In estimating Ideal hae
moglobin saturation with oxygen (ScOJ the P,O,
should first be corrected to /n vivo arterial conditions
of pH, base excess and temperature prior to estimatIng the Seo,. This latter value should also be corrected for the presence of carboxyhemoglobln both in
vivo and In the standanl oxyhemoglobln eqUilibrium
curve.
Interpretation of changes In arterial 0, concentratlon
(CaOJ Is facilitated by Identification Cao, along thevenous 0, concentration (CvOJ isopleths on a graph
of 0, content vs 0",/0..
Introduction
The "shunt fraction- is widely used in the assessment of hypox
aemia. In the measurement of true shunts (OJOJ and of the totalvenous admixture (O",/OJ (shunt and shuntlike effects of Icm ventila-
tionlperfusion ratios and oxygen diffusion limitation). errors may be
in lroduced by boIh the administralion of oxygen'" and faulty
calculation". The following apparently simple equation is used:
ecO,-CaO,
a...o. CcO -CvO, ,where CcO, is the ideat, end capillary oxygen content, CaO, thearterial and CVO, the mixed venous oxygen content respectively.
The simplicity of this equation belies the complexities inherent notonly in the measurement of Ca02 and evo 2• but also in the estimateof GcO,:;. These considerationsassume importanceat lower values ofinspired oxygen concentration as may be relevant in evaluation ofintracardiac shuntsand in theestimateof venousadmixture in patients
with respiratory failure. Although the latter group of patients arefrequently treated with high concentrations of 0, thismay not be thecase in the early or recovery stages of the illness and is certainly notso in patients treated for paraquat poisoning.
Re-arrangement of the equation (1l. thus,
CaO, eco, -(eco, -CvOJ a.../O, (2)
re-emphasises the importance of both the eco, and theCVO, in theanalysis of the ~ / 0 1 as a determinant of the arterial oxygenation.
Failure to consider changes in these other determinants of theCa02 may also result in erroneous interpretation of serial estimates ofthe venous admixture fractionS. These considerations are clearly ofimportancesinceinterpretation of the "shurt fraction" is frequently hcifydebated by clinicians'''.
Although theconcept of venous admixture is based on a simple,
two compartmentmodel of gasexchange'. is extremely useful in the
analysisof arterial hypoxia. lisuseiulness is, however, dependentuponan accurate value of the 0'61/0..
It is theaim of thiScommunicationto proposean accuratemethod
of calculating the a...1O, at ION concertrations of inspired oxygen. andtopresentagraphicalapproach to the irterp<etation of this derived value.
Blood sampling conditions
Arterial hypoxia is the r e s u ~ of both shunt and shunt-like effectson pUlmonary gas exchange. and in fact the latter is probablymoreimportant'. is well-established that administration of 100% 0, prior
10 bloodsampling results in an unpredictable overestimateof the Irueshunt fraction2.5-
8 due to conversion of gas exchange units with ICM'
vertlation perfusion ratios0//Q) 10unitswith aVlO of zero (true shunts).is logical, theretore. 10 measure the 0",,/0,. the venous admixture,
at the F,O, of therapeutic choice. The problem of interpr_ng seriaJ
venousadmixtureattained at differentinspiredoxygenconcentrationsmay be partfy overcome by the analysis presented below.
Volume 5. No. 2 1989
8/3/2019 1.2 the Shunt Fraction - Accurate Calculation & a Graphical Method of Interpretation
s i m u ~ e o u s l y inthe same patient In ananalysisof 30 s i m u ~ e o u s -Iy obtained arterial and venous samples (Instrument Laboratories IL
282 co-oximeter) I found a mean over-estimation of sea of 0.5% in
arterial blood, presumably due to light wave absorbtion overlap be-
tween HbO, and HbCa. Atthough the mean arteriovenous SCO
difference was only 0.5%, it was as great at 1% in individual patients.
Since overlap between absorption spectra of HbCO and deoxyhae
moglobin is less likely, tile I""""" venous value should bie used and
then the Sa02 corrected for the small over-estimate in sea using amodification of the formula ot Collier":
40 455
TEMPERATURE ('C)
JO
129"
IPM:zO .1 0 -0 J95:zT 'OOJ17 r2 1
'5
JO
70
PHoD
(mmHo)
so
Figure 1. Note theincrease ofPH,o with temperature. The two
data points shown are the extreme values recorded in our ceu:
PH,O values may differ significantly trom the conventional value
of 47 torr at 37"C. The equation for estimating PH,o is from Cotes
(10).
For example, with P A O , ~ l 0 0 torr, patient temperatureof 4O"C and ar
terial pH of 72 and base excess -7, t ile ideal end-eapillary SO, is
91.5%and not 975% as it would have been underconditions of stan
dard pH (7.4), temperature (32"C) and PCO, (40 torr).
The derived S02 value must then be further corrected for the
presence ofSCa. According to Bonsignoren the standard OHECwas
constructedwitll SCOof 1.0% bu t it is feasible thatat the higher levels
of PO,. this valueswould bie about 0.5%. For the purpose of estimat·
ing Seo" it is therefore proposed that the SCO (measured from a
venous sample and in calculating ecO,) bie reduced (C) by 1% at
F,O, < OA, and by 0.5% at F,O, > OA. Where venous seo is lessthan 1%, the measured value may bie used unaftered.
Care shouldalsobie taken toensure that tilehaemoglobin ismeasured from a blood samplethat is not e.xcessivelydiluted with hepariin,
as may readily occur in the critical care unit (CeU). The barometric
pressure should be measured and not guessed, and PH20 eitherread from tables or estimated from Cotes10 equation. In our CCU wehave had patients in cardicrrespiratory failure with temperatures (meas
ured from thethermistor of the pulmonary flotation catheter), ranging
from 29.3"C to 42_5"C. The change of PH,O with temperature is
graphically depicted in Figure 1.Finally the respiratoryexchange ranoshould biemeasured since
it may vary from 0.7to more than 1.0_ We obt,,;n VD, and veo, trom
application of the FlCk formula:
VO, or veo, 0, x(CaO, -CvO,) or 0, x(CvO, -CaCO,) (5)
R veO!'J0, (6)
0, - measured cardiac output (by thermodilution)
CaO" CvO, and CaCO, and CVCO, = measured 0 , and CO,
concenwations in
arterial and mixedvenous blood.
This approach obviates the necessity for inspiratory and expira
tory gas measurement(4)0,A - SO,T/{l00-SCO)
Where S0,A is the n"", value of SO"
SO,T is the measured SO,.
SCO is the arteriovenous SCO difference.
Calculation of CcO,
In order to accuratelyestimate theideal, end-eapillaryhaemoglobin
saturation with oxygen it is necessary to use either a table or amathematical model of tile standard OHEC". Furthermore, it IS necessary
Initially 10correct theP.o, for dElVlations of arterial pH (PGO, and base
excess effects) and temperature from tIlat of standard cond' o n s ~ .
PaCO,
Rp.
PH,OT
= the ideal, alveolar PO,
- % carboxyhaemoglobin
= a correction factor for the presence of sea in thestandard oxyhaemoglobin curve (q.v.)
= arterial carbon dioxide tension
= respiratory exchange ratio
= measured barometric pressure
water vapour pressure (mmHg)= patient lung temperature in OC
tioned aIbovemay bie small, especially at high Fp, I " " ".. the cumula
tive effect of minor inaccuracies may be substantial. The use of anFP2 of 1.0 does not necessarily imply uniformity of serial measured
conditions, since not only may the true shunt vary3-8. but in additionthere may be changes in CV02and/or Cc02 due to alterations in haemoglobin concentraton, cardiacoutputand oxygen consumption. Pul-
monary catheters are not only used In patients in severe respiratoryfaIlure and or a high FP , therapy (but also In patients In cardiac failure
and patients being treared WIth lowF,O, for paraquat poisoning). The
effect of calculation errors In eslJmatmg a..,/Q, are particularly impor·
Volume 5. No. 2 1989 19
8/3/2019 1.2 the Shunt Fraction - Accurate Calculation & a Graphical Method of Interpretation
tan! at low Fp,. An example of the variability in calculation of 0",/0,
dueto "errors" inmethdology is given in Table 1. Note not onlythediffer
ence in 0",/0, at F,O, of 0.21 but also the similarity of thevalues whena../O, is calculated by the convenlional method at Fp, of 0.21 and0-5. The falsely high a../o, (conventional method) at F,O, of 0.21 maywell have ledto averydifferentchoiceoftherapywhencompared with
thevalue oblained by themOfe accuratemethod. This exampledemon·strates the important implications tor the interpretation of serial estimatesof venous admixture at different therapeutic levels of Ft0 2 • It is alsoclear frOf" lable 1 that methodological "errors· in calculating 0",/0,
are relatively unimportant when F.02 is high.
500000
ITI I1SI7 0 . . . L
__. . . .,. -. . . ._ . . . . ,. -
__. . . .
__. . . . . ,
90
quantitatively most important cause of this deterioration was not theincrease in the venous admixture to 30%, but the decline in Sv0
2
from 750/0 10 50%. The Seo, (the y-inlercept) was constant The actual change in SaO, can be viewed as "a resultant" of these lorcesviz. changes in SV02 and a.•la l; in this particular example, therapymight usefully be directed at increasingSVO, by, tor example, increas
Figure 3. Graphical analysisofoxygendata in which oxygen con
tent is plotted against a../o,. The lineswith CaO, data points in·dlCaled along them are CVO, isopleths. The eco, value coincides with the intercept of the lsopleths.
Figure 2. The decline in Sao, IrOf" 93.7% 10 82.9% may be
viewed as the resultant of a decline in SV02and increase in
0",/0,. (See text for further details.)
Oxygen
Content
lmlJ1OOml)
15'5
In Rgure3 theoxygencontent is shCJNn on the y·axis. In essence,this is a graphical presentalion of data based on eeuation (2) whichhas themathematical farm of astraight line: y=mx+c, where the intercept is eco, and the slopeof the CvG, isopleth is CcO, -CVO,. Note
how the improvement in CaO, IrOf" 14.8 ml 0,1100 ml 10 16.3 ml0,1100 ml was the resull 01 a modest decline in OvJO, and substantial increases in CV02 andCc02• TheCc02 is represented by the intercept of theCv02 isoplethswith the O2 content axis; in clinical termsCc02 is increased mainly by increasing haemoglobin concentrationand the F,O,. ThUs, using the graphical analysis iI is possible to estimate quantitatively the relative importance of changes in the deter
minants 01 arterial oxygenation and thereby logically 10 direct furtherand assess the success therefore. Implementation of the formulaeandgraphics for clinical use in the CCU is a simplematter tn the current
CaO, (mV1oo m 14.8 14.8 13.8 13.8CVO, (mV1oo m 4.4 4.4 4.4 4.4
°VA/O, (%) 22.9 29.7 4.9 30.1(27.6) (34.0)
Even if the0",10, is accurately calculated, Ihe interpretation of thevalue as a determinant ofCaO, still remains problematical. 11 is evidentfrOf" InspeclJon of equation (2) that the CcO" the CvG, an d 0 ...10, alldeterminethe final arterial oxygencontErt All thesedeterminantsmayvary considerably throughout thecourse d thepatienlS' illness. In Ofder10 facilitate the pathophysiological ,nterpretalion of oxygenation varia·
bIesand to better un<Jersland themechanism d arterial hypoxia, graphIC presentationof the data becOfnes varyhelpful. A simplified versionofadiagram suitable for evaluation et dataotIalned when the patientbreathed room air, IS presented in Figure 2. The relative Influence of
changes in a../O, Of SVO, only is shown. While brealihing rOQfn ai ,
the patient's SaO, declined frOf" 93.7% 10 82.9%. 11 is clear that the
Interpretation of venous admixture values
Table 1. The "conventional method" of determining the0",10, IS
compared tothemOfeaccuratemethod described in thelext Note
theCNer-estimatlon of0 ...10, at Fp, 0.21 by the formermethodand thal this difference is reduced if the blood gases had beenoblained at a high F,O,. Note also the similarity of the two "conventional values" in spite of the great difference in FP2"PP, ~ i n s p i r e d PO,; "PACO, " ~ P a C O , (F,O, +(1-F,0,)/R).
The bracketed ScOz values are the ones derived from P..,02
without regard 10 the SCO; the bracket 0 .../0, values refer then10 the value of this variable using the bracketed Seo,.