Newsletter

Sign up for our quarterly newsletter and get the newest articles from acutecaretesting.org

Printed from acutecaretesting.org

Article

July 2011

Central venous blood gas analysis

Blood gas analysis (BGA) is a laboratory and point-of-care test routinely used to assess acid-base status along with adequacy of ventilation and oxygenation among predominantly critically/acutely ill patients. 

The ”gold standard” sample for BGA is arterial blood collected anaerobically by needle puncture of an artery or via an indwelling arterial catheter. BGA is unique among blood tests in its requirement for arterial blood; all other tests are performed on venous blood, collected usually by needle puncture of a peripheral vein (venepuncture); or less commonly on capillary blood obtained by finger prick. 

In intensive care settings most patients who require frequent blood gas monitoring have a central venous catheter inserted that allows easy and safe sampling of venous blood for laboratory testing, obviating the need for repeated venepuncture. It would be logistically convenient for clinical staff, and more comfortable and safer for the patient if this kind of venous blood sample could also be used for BGA. 

This article addresses the question: is central venous blood an acceptable alternative to arterial blood for blood gas analysis? The main focus of the article will be results of clinical studies that have compared BGA results derived from arterial blood with BGA results derived from simultaneously sampled central venous blood. Consideration will also be given to mathematical corrections that are intended to allow prediction of arterial blood gas values from measured venous blood gas values. 

The article begins with a very brief discussion of relevant physiological differences that distinguish arterial and venous blood.

 

THE ARTERIO-VENOUS (A-V) DIFFERENCE

Blood gas analysis (BGA) involves measurement of three parameters: the amount of free (unbound) oxygen (O2) and carbon dioxide (CO2) dissolved in blood, and the pH (acidity/alkalinity) of blood. 

The partial pressure (p) exerted by the two gases is what is actually measured so the three measured parameters are: pO2, pCO2and pH. A further parameter, bicarbonate (HCO3-) concentration is generated during blood gas analysis but this is calculated from pH and pCO2, rather than directly measured.

pO2 is used to assess patient oxygenation status; pCO2 is used to assess ventilation; and pH, pCO2 and HCO3- results together allow assessment of acid-base status. Another calculated parameter, base excess (BE), is also helpful, although often not necessary in this regard. Clearly, if the pO2 of arterial blood were the same as the pO2 of venous blood, then it would be immaterial which sample were used to assess oxygenation. 

Likewise, if the pH, pCO2 and HCO3- of arterial blood were the same as the pH, pCO2 and HCO3- of venous blood, then it would be immaterial which sample were used to assess ventilation and acid-base status.

Of course these equalities between arterial and venous blood do not exist because of the physiological exchange of oxygen and carbon dioxide that occurs as blood flows through the capillary bed of all tissues and the capillary bed of the alveoli of the lungs. 

It is this two-site gaseous exchange that fulfills a principal function of blood: delivery of inspired oxygen from lungs to all tissue cells and delivery of carbon dioxide (a waste product of cellular metabolism) from all tissue cells to lungs for excretion in expired air.

Veins convey blood from all tissues to the right side of the heart before onward journey via the pulmonary artery from heart to the lungs. This blood (venous blood) is relatively lacking in oxygen and relatively rich in carbon dioxide due to the gaseous exchange that has occurred in the capillary bed of tissue cells. 

As this blood flows through the alveoli of the lungs it gains oxygen (becomes oxygenated) and loses carbon dioxide before onward journey via the pulmonary veins to the left side of the heart. Non-pulmonary arteries convey blood from the left side of the heart via the aorta to the capillary bed of all tissues. This blood (arterial blood) is oxygenated but relatively lacking in carbon dioxide due to the gaseous exchange that has occurred in the alveoli of the lungs. 

The differences in the oxygen and carbon dioxide tensions of venous and arterial blood are reflected in the reference ranges of parameters generated during blood gas analysis (Table I). 

Of particular note for the following discussion it is evident from Table I that normal arterio-venous (A-V) difference is much greater for the measure of oxygenation (pO2) than for the measurements used to assess ventilation and acid-base status (pH,pCO2,HCO3-).

TABLE I: Arterial and venous blood gas reference range



Arterial Venous Arterio-venous
(A-V) difference

pH

7.35-7.45 7.31-7.41 ~ 0.04

pCO2 (kPa)

4.7 - 6.0 5.5 - 6.8 ~  0.6

pCO2 (mmHg)

35 -45  41 - 51 ~ 6

Bicarbonate (mmol/L)

22-28 23-29 ~ 1

PO2 (kPa)

10.6 - 13.3 4.0 -5.3 ~ 8.0

pO2 (mmHg)

 80-100 30 -40 ~ 55

sO2 (%)

> 95 75 > 20

Ever since BGA was first introduced to clinical care in the 1960s, arterial blood has been the standard sample; it reflects alveolar (pulmonary) gas exchange and all parameters generated by BGA are constant throughout the non-pulmonary arterial system. 

The great body of research that underlies the clinical application of BGA is based for the most part on studies conducted using arterial blood. Published reference ranges used to interpret patient blood gas values have been extensively validated using arterial blood, and clinicians are familiar with these rather than reference values derived from venous blood which, in any case, are less well validated. 

Despite this, over the past decade or two there has been an increasing level of clinical interest in the notion that it is worth investigating if venous blood might be a valid substitute for arterial blood in some circumstances. 

The impetus for this clinical interest centers largely on the practical disadvantages associated with sampling arterial rather than venous blood, but validation and development of pulse oximetry as an alternative means of assessing arterial oxygenation has been a significant factor in driving that interest.

DIFFICULTIES ASSOCIATED WITH USE OF ARTERIAL BLOOD FOR BGA

Collection of arterial blood for BGA is usually by needle puncture of a peripheral artery. The most common puncture site is the radial artery in the wrist; alternative sites include the brachial artery in the forearm and the femoral artery in the groin. 

Compared with venepuncture, arterial puncture is technically more demanding and significantly more painful and hazardous for the patient [1-3]. Specialist training in arterial puncture is essential for patient safety and comfort, and in many countries, obtaining arterial blood by arterial puncture remains the almost exclusive preserve of medically qualified staff. 

By contrast, venepuncture is a very commonplace procedure that can be easily and safely performed, after minimal training, by ancillary staff with no medical or nursing education.

In an intensive care setting patients often have an indwelling arterial catheter fitted principally to enable continuous blood pressure monitoring. These catheters also allow convenient and painless sampling of arterial blood for BGA. 

Although this method of arterial blood sampling obviates the need for repeated needle puncture of patients requiring frequent BGA, fitting of an arterial catheter is itself an invasive and technically difficult procedure [4] that is associated with risk of serious complications including systemic infection, hemorrhage, thrombosis and ischemia [5,6]. 

So common and serious are these complications that some have recently questioned whether the benefit of continuous blood pressure monitoring among the critically ill outweighs the considerable risk of arterial catheterization [7]. 

These concerns suggest that there might be more restricted use of the arterial catheter in the future. If so, then the only means of obtaining arterial blood for BGA, even in an intensive care setting, would be arterial needle puncture.

Patients who require BGA also require regular venous blood sampling for other blood tests. It would clearly be convenient, safer (for patients and staff) and more economic if a single venous sample could be used for all blood tests, including BGA.

IMPACT OF PULSE OXIMETRY

The contribution that BGA makes to the assessment of patient oxygenation status is measurement of pO2. sO2 determines the % of hemoglobin that is saturated with oxygen (sO2) and thereby the total amount of oxygen in blood. 

The relationship between pO2 and sO2, described graphically in the familiar sigmoidally shaped oxyhemoglobin dissociation curve, allows calculation of sO2 from measured pO2. Arterial blood gas analysis thus allows measurement of arterial pO2 (pO2(a)) and calculation of arterial sO2 (sO2(a)). 

In practice modern blood gas analyzers have an incorporated CO-oximeter that allows direct measurement of sO2(a).

Pulse oximetry, which has become ubiquitous in all areas of clinical medicine since the mid-1990s, provides an alternative entirely safe, non-invasive means of continuously monitoring arterial oxygen saturation and thereby roughly predicting pO2(a).

Although there is clinically acceptable agreement between arterial oxygen saturation measured by pulse oximetry (SpO2) and arterial oxygen saturation measured (or calculated) during BGA (sO2(a)) for most patient groups [8],thisis not necessarily the case [9]. 

There is for example conflicting evidence that SpO2 is a less-than-reliable measure of sO2(a) among critically ill patients with anemia, hypoxemia or acidosis [10]. Still, for many patients in whom the only reason for performing BGA is assessment of oxygenation status, pulse oximetry is a very convenient, reliable and safe alternative.

With pulse oximetry now providing an alternative means of assessing arterial oxygenation, studies aimed at consideration of the reliability of venous blood as a substitute for arterial blood have been able to focus principally on those blood gas parameters (pH,pCO2 and bicarbonate) that have lowest A-V difference (Table I) and therefore most likely to show agreement when arterial and venous values are compared.

PERIPHERAL VENOUS BLOOD, CENTRAL VENOUS BLOOD AND MIXED VENOUS BLOOD

Many (probably most) clinical studies investigating the validity of using venous blood for BGA have been conducted using venous blood obtained by conventional venepuncture of a peripheral vein (i.e. peripheral venous blood) [11-20]. 

This article is concerned only with studies [21-28] that have utilized central venous blood samples for comparison with arterial blood.

Central venous blood is the blood that is sampled via a central venous catheter (CVC). In addition to facilitating the means for easy sampling of venous blood for diagnostic testing, CVCs allow continuous monitoring of central venous pressure (vital in the hemodynamically unstable patient), and vascular access for administration of drugs, blood transfusion and other fluids.

Most patients (up to ~80 %) in intensive care have an indwelling CVC, but CVC use is not confined to this patient population so these studies [21-28] have relevance outside the intensive care unit, in emergency rooms, recovery rooms and some medical wards.

CVCs are usually inserted cutaneously via the jugular vein in the neck or subclavian vein in the upper chest to the superior vena cava, with the tip sited close to the point where the superior vena cava opens to the right atrium of the heart (Fig. 1), so that the blood sampled is the mixed venous blood from the upper half of the body. 

The inferior vena cava conveys mixed venous blood from the lower half of the body to the right atrium. Central venous blood is thus not truly mixed venous blood because it does not include that returning via the inferior vena cava. 

Mixing of venous blood from all parts of the body occurs as it flows from the right atrium to the right ventricle before journey from the heart via the pulmonary artery. 

Catheterization of the pulmonary artery provides the only means of sampling true mixed venous blood.

Peripheral blood obtained by venepuncture is different from central (”mixed”) venous blood and true mixed venous blood with regard to blood gas parameters (pH,pCO2,pO2) because as venous blood returns from the periphery back to the heart, it becomes mixed with venous blood from other tissues having differing levels of metabolic activity and therefore potentially differing pH,pO2, and pCO2

Unlike arterial blood, which remains constant with regard to these values until it reaches the capillary bed of tissues, venous blood gas values can potentially differ to some extent with site of sampling.

STUDIES COMPARING CENTRAL VENOUS AND ARTERIAL BLOOD GAS RESULTS

All clinical studies [11-28] investigating the validity of using venous blood for BGA share a simple and common design. In essence BGA results derived from arterial blood are compared with BGA results derived from simultaneously collected venous blood among a defined cohort of patients requiring BGA. 

It is of course vital for the validity of the comparison that both arterial and venous samples are collected anaerobically and analyzed within a common short time frame, using the same analyzer.

Of seven studies [21-27] that have examined the validity of using central venous blood for blood gases, all compared central venous and arterial pH; six [21-23,25-27] compared central venous and arterial pCO2; four [21,24,26,27] compared central venous and arterial bicarbonate (HCO3-); two [23,24] compared central venous and arterial base excess; and just one [25] compared central venous and arterial pO2.

Some details of these seven studies along with summary of the results are contained in Tables II-VI. The two most significant columns in these tables are the mean arterio-venous (A-V) difference along with range or SD of that difference; and the 95 % limits of agreement (LOA) on a Bland-Altman plot. 

A Bland-Altman plot is the accepted method for assessing the agreement between two tests and represents a clinically relevant measure of comparison. The difference between two paired (arterial and central venous) values are plotted against the mean of those two values. 

The derived 95 % LOA allowsestimation of the range of difference that can be expected between central venous and arterial values for all patients represented by the study population.

CENTRAL VENOUS pH VERSUS ARTERIAL pH

In all seven studies mean arterial pH was higher than the mean central venous pH (see Table II). The magnitude of this positive bias (mean A-V difference) ranged from 0.027 [26] to 0.05 pH units [21], but in most studies [23-27] mean bias was close to 0.03 pH units. 

Four of the seven studies provided 95 % LOA data. For the study showing best agreement [25] 95 % LOA was 0.008 to 0.063. This indicates that if measured central venous pH is 7.40, then in 95 % of patients arterial pH would lie within the range of 7.408 to 7.463, with most close to 7.43. 

For comparison, the study [23] showing the worst level of agreement, with 95 % LOA –0.03 to 0.09 indicates that for a measured central venous pH of 7.40 arterial pH would lie within the range of 7.37 to 7.49 for 95 % of patients, again with most close to 7.43.

TABLE II: Arterial versus central venous pH

Patient number and type No. of paired samples Mean Arterial (range or ±2SD) Mean Venous (range or ±2SD)   Mean A-V diff (range or ±2SD) Bland-Altman 95 % LOA Reference
55 "seriously ill" surgical patients 55 7.39 (7.15 to 7.55) 7.34 (7.12 to 7.48) 0.05
(0 to 0.13)
NR 21 (1967)
41 Critically ill adults in ICU 41 7.40 (6.97 to 7.56) 7.36 (6.95 to 7.51) 0.04
(-0.01 to 0.1)
NR 22 (1969)
25 adult trauma patients in ICU 99 7.39 (±0.14)  7.36 (±0.14)  0.032 (±0.052)  -0.03 to 0.09 23 (2005)
110 adult patients in ICU 168 7.37 (7.12 to 7.50) NR 0.03 (NR) -0.01 to 0.07 24 (2006)
73 adults from thoracic ICU, general ICU and pulmonary ICU 73 7.39 (7.24 to 7.54) 7.35 (7.21 to 7.45) 0.036 (±0.028) 0.008 to 0.063 25 (2008)
40 adults medical ICU, 72 % with sepsis 190 7.37 (±0.276)  7.34 (±0.268) 0.027 (±0.054)  -0.028 to 0.081 26 (2010)
187 adults medical and surgical ICU and cardiac catheteri -zation lab. 187 7.41 (±0.14) 7.37 (±0.14) 0.035 (±0.04) only venous adjusted LOA recorded - see text 27 (2010)
NR - not recorded

Given the narrow 95 % LOA and the consistency of mean A-V difference across nearly all studies, there is general agreement [23-27] that central venous pH is a clinically acceptable substitute for arterial pH after taking account of the systematic positive bias of ~0.03 pH units.

CENTRAL VENOUS pCO2VERSUS ARTERIAL pCO2

In all six studies mean arterial pCO2 was found to be less than mean central venous pCO2 (see Table III). The magnitude of this negative bias (mean A-V difference) ranged from 0.52 [26] to 1.22 kPa [21] (i.e. 3.9 to 9.2 mmHg) with the four most recent studies [23-27] indicating a negative bias in the narrower range of 0.52 to 0.79 kPa (3.9 to 5.9 mmHg). 

Three of the six studies provided 95 % LOA data. For the study showing best agreement [25] 95 % LOA was–1.3 to –0.28 kPa. This indicates that if measured central venous pCO2 is 5.0 kPa (38mmHg), then in 95 % of patients, arterial pCO2 would lie within the range of 3.70-4.72 kPa (28-35 mmHg) with most close to 4.2 kPa (31mmHg). 

For comparison, the study showing worst level of agreement [26] with 95 % LOA –1.63 to +0.64 kPa, a measured central venous pCO2 of 5.0 kPa predicts an arterial pCO2 in the range of 3.37 to 5.64 kPa (25 to 42 mmHg) for 95 % of patients with most close to 4.5 kPa (34 mmHg).

TABLE III: Arterial versus central venous pCO2 (kPa) ‡

Patient number and type No. of paired samples Mean Arterial (range or ±2SD) Mean Venous (range or ±2SD)   Mean A-V diff (range or ±2SD) Bland-Altman 95 % LOA Reference
55 "seriously ill" surgical patients 55 4.28 (1.99 to 9.31) 5.50 (2.66 to 10.3) -1.2 (range/
SD -NR)
NR 21 (1967)
41 Critically ill adults in ICU 41 4.52 (2.66 to 8.88) 5.58 (2.93 to 9.71) -1.06
(-2.39 to +0.27)
NR 22 (1969)
25 adult trauma patients in ICU 99 5.45 (±1.96)  5.98 (±1.83 )  -0.58 (±0.89)  -1.44 to -0.29 23 (2005)
73 adults from thoracic ICU. general ICU and pulmonary ICU 73 5.80 (3.98 to 10.81) 6.61 (4.64 to 10.9) -0.79 (±0.52)  -1.30 to +0.64 25 (2008)
40 adults medical ICU, 72% with sepsis 190 5.10 (±0.276) 5.62 (±0.268) -0.52 (±0.054)  -1.63 to +0.64 26 (2010)
187 adults medical and surgical ICU and cardiac cathete
-rization lab.
187 5.32 (±0.14)  5.98 (±0.14)  -0.59 (±0.04)  only venous adjusted LOA recorded - see text 27 (2010)
NR - not recorded
‡ to convert kPa to mmHg divide by 0.133

There is general agreement [22,25-27] that central venous pCO2is a clinically acceptable substitute for arterial pCO2 in most clinical contexts so long as the systematic negative bias of ~0.6 kPa (5.0mmHg) is taken into account. 

The authors of one study [23] consider the 95 % LOA too wide for general substitution of central venous values but concede that central venous pCO2 provides clinically valuable information that, for example, can guide weaning of trauma patients (the population they were studying) from mechanical ventilation.

In general one can be 95 % certain that after correction for systematic bias, central venous pCO2 is within ±0.52 kPa (i.e. ±3.9 mm Hg) of arterial pCO2[25].

CENTRAL VENOUS BICARBONATE VERSUS ARTERIAL BICARBONATE

Since bicarbonate (HCO3-) generated during blood gases is calculated from pH and pCO2 , it would be expected that if central venous pH and pCO2 are clinically acceptable substitutes for arterial pH and pCO2,then central venous HCO3-, too, would be an acceptable substitute for arterial HCO3-(see Table IV). 

This is borne out by the results of the four studies [21,24,26,27] that compared central venous and arterial HCO3-. All studies indicate that mean central venous HCO3- concentration is slightly higher than mean arterial HCO3- concentration. 

The magnitude of this negative bias (A-V difference) ranged from 0.52 mmol/L in one study [24] to 2.2 mmol/L in another [21]. Of the four studies, three returned negative bias of <1.2 mmol/L, which is clinically insignificant. Two studies [24,26] provided 95 % LOA data. 

The study showing best level of agreement with 95 % LOA –2.85 to +1.85 indicate that if measured central venous HCO3- is 25 mmol/L, then in 95 % of patients predicted arterial HCO3- would be in the range of 22 to 27 mmol/L with most close to 26 mmol/L. 

There is general agreement that central venous bicarbonate is a clinically acceptable substitute for arterial bicarbonate, especially if the small systematic positive bias of ~1mmol/L is taken into account.

TABLE IV: Arterial versus central venous HCO3- (mmol/L)

Patient number and type No. of paired samples Mean Arterial (range or ±2SD) Mean Venous (range or ±2SD)   Mean A-V diff (range or ±2SD) Bland-Altman 95 % LOA Reference
55 "seriously ill" surgical patients 55 NR NR -2,2 NR 21 (1967)
110 adult patients in ICU 168 25 (14.6 to 42.2) NR -0.52 (NR) -2.85 to +1.85 24 (2006)
40 adults medical ICU, 72% with sepsis 190 22.4 (±15.2) 23.2 (±15.6) -0.8 (±3.16) -4.0 to +2.4 26 (2010)
187 adults medical and surgical ICU and cardiac cathete
-rization lab.
187 25.4 (±8.4) 26.6 (±13.2) -1.13 (±8.6) only venous adjusted LOA recorded - see text 27 (2010)
NR - not recorded

CENTRAL VENOUS BASE EXCESS VERSUS ARTERIAL BASES EXCESS

Just three studies [22-24] compared central venous and arterial base excess (BE)(see Table V). Mean A-V difference was small (–0.19 mmol/L and –0.18 mmol/L) and 95 % limit of agreement was sufficiently narrow for one study author to conclude that central venous and arterial values are interchangeable [24].

TABLE V: Arterial versus central venous base excess (mmol/L)

Patient number and type No. of paired samples Mean Arterial (range or ±2SD) Mean Venous (range or ±2SD)   Mean A-V diff (range or ± 2SD) Bland-Altman 95 % LOA Reference
110 adult patients in ICU 165 -0.1
(-12 to +16)
NR -0.19 (range/SD -NR) -2.24 to +1.86 24 (2006)
25 adult trauma patients in ICU with sepsis 99 -0.01 (±7.76)  -0.34 (±7.44)  -0.34 (±2.06)  -2.20 to +1.80 26 (2010)
NR - not recorded

CENTRAL VENOUS pO2 VERSUS ARTERIAL pO2

Just one study [25] compared central venous O and arterial O(see Table VI). The large mean and range of A-V difference of 8.33kPa ± 7.88(2SD) (i.e. 63 ± 59 mmHg) confirms that it is not possible to use central venous O as a reliable substitute for arterial O. 

There is no correlation between arterial O and venous O (irrespective of the sampling site). The only reliable sample for accurately determining arterial oxygenation is arterial blood. Pulse oximetry provides an alternative means of assessing patients’ oxygenation status that requires no blood sampling.

TABLE VI: Arterial versus central venous pO2 (kPa) ‡

Patient number and type No. of paired samples Mean Arterial (range or ±2SD) Mean Venous (range or ±2SD)   Mean A-V diff (range or ±2SD) Bland-Altman 95 % LOA Reference
73 adults from thoracic ICU. general ICU and pulmonary ICU 73 11.32 (6.6 to 28.3) 5.41 (3.86 to 7.16) 8.33 (±7.88) Not calculated 24 (2008)
‡ to convert kPa to mmHg multiply by 0.133

PATIENTS IN SEVERE CIRCULATORY FAILURE - A SPECIAL CASE

The studies discussed thus far [21-27] have confirmed that the normal arterio-(central)venous (A-V) difference for pH and pCO2(~0.03 pH units and ~ –0.6 kPa respectively) are maintained within broadly clinically acceptable limits for the generality of patients requiring BGA. 

That is not the case for patients with severe circulatory failure (for example those suffering cardiac arrest). Adrouge et al [28] found much larger A-V differences in this small subset of very critically ill patients. 

His study revealed that mean difference between arterial pH and central venous pH ranged from 0.10 to 0.35 pH units depending on the severity of the circulatory failure, rather than ~0.03 pH units. 

Mean difference between arterial pCO2 and central venous pCO2 for the same group ranged from –3.2 to –7.4 kPa, rather than –0.6kPa. According to the authors of this report assessment of acid-base status in these patients requires consideration of both arterial and central venous blood gas results. 

Two further studies [29,30] confirm the much larger difference between arterial and central venous pH and pCO2 for patients in circulatory collapse.

MATHEMATICAL CORRECTIONS

There are three methods for mathematically converting measured central venous blood gas results to give”arterial” blood results. The first and most simple, which has already been hinted at, is to use the systematic differences between arterial and central venous blood that have been derived from the seven studies [23-27] thus:

”arterial” pH = measured central venous pH + 0.03

”arterial”pCO2(KPa) = measured central venous pCO2 –0.6

”arterial” HCO3-(mmol/L) = measured central venous HCO3- + 1.0

The capacity of this simple approach to improve diagnostic accuracy has been demonstrated by Walkey et al [27].

A second approach is to use regression equations generated during studies comparing central venous and arterial values. Treger et al [26] derived the following regression equations from their data:

”arterial” pH = –0.307+1.05 ×measured central venous pH

”arterial”pCO2(mmHg) = 0.805 + 0.936 ×central venous pCO2(mmHg)

”arterial” bicarbonate = 0.513 + 0.945 ×central venous bicarbonate

The validity (accuracy) of these two approaches dependson the assumption that the generality of patients are represented by the study population from which the systematic differences and regression equations are derived.

Toftegaard et al [31] have recently developed a novel much more sophisticated, patient-specific method of converting venous (either central, peripheral or mixed) to arterial values that depends on measuring arterial oxygenation by pulse oximetry at the time that venous blood is sampled for blood gases. 

The principle of the method is ”to calculate arterial values by simulating, with the help of mathematical models, the reverse transport of blood from the veins to the arteries until the simulated arterial oxygenation matches that measured by pulse oximetry”–effectively, a mathematical arterialization of venous blood. 

The complex mathematical transformation requires input of the following measured venous parameters all available on modern blood gas analyzers: pH, pCO2,pO2, sO2(a), hemoglobin, methemoglobin and carboxyhemoglobin; along with SpO2 determined by pulse oximetry.

A validation study of this method [31] indicates that calculated arterial values for pH and pCO2 by this method are essentially the same as measured arterial values. 

The transformation also allows for the first time a clinically useful estimation of arterial pO2 from central venous blood, although this clinical utility only applies to patients with SpO2<96 %. For those with SpO2>96 %, arterial pO2 cannot be estimated within an acceptable clinical range by this method. 

The imprecision in estimating arterial pO2 when SpO2 is >96 % is due to the flat shape of the oxyhemoglobin dissociation curve at high sO2 values where small changes in sO2 result in large changes in pO2

Although this limits the usefulness of this way of calculating pO2(a), the authors of this study observe that it is encouraging that the method is able to predict pO2(a) within clinically acceptable limits for patients with low SpO2because these are the clinically interesting patients. 

Previous study [32] has shown this method of estimating pO2(a) to be highly sensitive to error in SpO2 measurement. In the validation study [31] comparison of patients’ SpO2 and sO2(a) revealed a mean SpO2 bias (SD) of 0.4 %± 1.0 %. 

This favorable degree of accuracy/precision in SpO2 measurement allows calculation of pO2(a) within ±1.85 kPa (2SD) of measured value, if SpO2 is <96 %. This is judged clinically acceptable. Error ≥2 % (SD) in SpO2 measurement results in inaccurate (clinically unacceptable) pO2(a) estimation, even if SpO2 is <96 %.

SUMMARY

  • Central venous blood is unsuitable for determining patient oxygenation status. For many patients this can be determined sufficiently accurately using non-invasive pulse oximetry. If this is not the case, arterial blood must be sampled for measurement of pO2(a) and sO2(a).
  • Although central venous pH, pCO2(a) and bicarbonate are not interchangeable with arterial values, there is excellent correlation between the two for all three parameters.
  • With the exception of patients in severe circulatory failure, on average central venous pH is 0.03 pH units lower than arterial pH; central venous pCO2is 0.6 kPa (5mmHg) higher than arterial pCO2; and central venous and arterial bicarbonate are essentially the same.
  • Corrected central venous pH,pCO2 and bicarbonate provide results that are, in many cases, clinically insignificantly different from those obtained using the”gold standard” arterial blood sample.
  • Acidosis and alkalosis can be correctly diagnosed using central venous blood but severity may be under- or overestimated in some patients.
  • Central venous blood gases provide clinically useful information about patient acid-base status that can in some cases obviate the necessity for arterial sampling. Certainly the finding of corrected central venous pH,pCO2 and bicarbonate values within the normal arterial reference range is reliable evidence of normal acid-base status.
  • A recently developed, highly sophisticated mathematical conversion allows the most precise calculation of arterial pH,pCO2 and bicarbonate from measured central venous values. The conversion requires input of oxygen saturation measured by pulse oximetry (SpO2). The conversion also allows a clinically useful estimation of arterial pO2 so long as SpO2 is <96 %.
References
  1. Clinical and Laboratory Standards Institute (CLSI formerly NCCLS) Procedures for the Collection of Arterial Blood Specimens; Approved standard - 4th edition H11-A4 (ISBN 1-56238-543-3) Pennsylvania USA 2004
  2. Okeson G, Wullbrecht P. The safety of brachial artery puncture for arterial  blood sampling. Chest 1998;114:748-51
  3. Giner J, Casan P, Belda J et al. Pain during arterial puncture. Chest 1996;110:1443-45
  4. Eisen L, Miami T, Berger J et al.Gender disparity in failure rate for arterial catheter attempts. J Intensive Care 2007;22:166-72
  5. Wallach SG.Cannulation injury of the radial artery: diagnosis and treatmentalgorithm.Am J Crit Care 2004;13:315-19
  6. Scheer B, Perel A, Pfeifer U. Clinical review: complications and risk factors of peripheral arterial catheters used for haemodynamic monitoring in anaesthesia and intensive care medicine. Crit Care 2002;6:199-204
  7. Garland A, Connors A. Indwelling arterial catheters in the Intensive Care Unit: Necessary and beneficial, or a harmful crutch? Am J Respir Crit Care Med 2010; 182:133-37.
  8. Jensen L, Onysklw J, Prasad N. Meta-analysis of arterial oxygen saturation by pulse oximetry in adults. Heart Lung 1998;27:367-408
  9. Keogh BF. When pulse oximetry monitoring of the critically ill is not enough. AnesthAnalg 2002;94(Suppl 1):S96-99
  10. Wilson B et al. The accuracy of pulse oximetry in emergency department patients with severe sepsis and septic shock: a retrospective cohort study. BMC Emergency Medicine 2010;10:9
  11. Brandenburg M, Dire D. Comparison of arterial and venous blood gas values in the initial emergency department evaluation of patients with diabetic ketoacidosis. Ann Emerg Med 1998;31: 459-65
  12. Rang L, Murray H, Wells G et al.Can peripheral venous blood gases replace arterial blood gases in emergency department patients. CJEM 2002; 4: 7-15
  13. Chu YC, Chen CZ, Lee CH et al.Prediction of arterial blood gas values from venous blood gas values in patients with acute respiratory failure receiving mechanicalventilation. J Formos Med Assoc 2003;102:539-43
  14. Yildizds D, YapicogluH, Yilmaz H et al. Correlation of simultaneously obtainedcapillary, venous and arterial blood gases of patients in a paediatric intensive care unit.Arch Dis Child 2004;89:176-80
  15. Ak A, Ogun C, Bayir A et al.Prediction of arterial blood gas values form venous bloodgas values in patients with acute exacerbation of chronic obstructive pulmonarydisease. Tohoku J Exp Med 2006;210:285-90
  16. Kelly A. The case for venous rather than arterial blood gases in diabetic ketoacidosis. Emerg Med Australas 2006;18:64-67
  17. Malatesha G, Singh N, Bharija A et al.Comparison of arterial and venous pH, bicarbonate, PCO2 and PO2 in initial emergency department assessment. Emerg Med J 2007;24:569-71
  18. Bilan N Behbahan A Khosroshahi A Validity of venous blood gas analysis for diagnosis of acid-base imbalance in children admitted to pediatric intensive care unit World Journal of Pediatrics 2008;4:114-117
  19. O’Connor T Barry P Jehangir A Comparison of arterial and venous blood gases and the effects of analysis delay an air contamination on arterial samples in patients with chronic obstructive pulmonary disease and healthy controls Respiration 2011;81:18-25
  20. Koul P, Khan U, Wani A. Comparison and agreement between venous and arterial gasanalysis in cardiopulmonary patients in Kashmir valley of the Indian subcontinent. Ann of Thoracic Med 2011;6:33-37.
  21. Sutton R, Wilson R, Walt A. Differences in acid-base levels and oxygen saturationbetween central venous and arterial blood. Lancet 1967;2 ;748-51
  22. Phillips B, Peretz D. A comparison of central venous and arterial blood gas values inthe critically ill. Ann Intern Med 1969;70:745-49.
  23. Malinoski D Todd S Slone S et al C orrelation of central venous and arterial blood gas measurements in mechanically ventilated trauma patients Arch Surg 2005;140:1122-25
  24. Middleton P, Kelly A-M, Brown J et al. Agreement between arterial and centralvenous values for bicarbonate base excess and lactate. Emerg Med J 2006;23:622-24
  25. Toftegaard M, Rees S, Andreassen .S Correlation between acid-base parameters measured in arterial blood and venous blood sampled peripherally form vena cavae superior and from the pulmonary artery. Euro J Emerg Med 2008; 15:86-91
  26. Tregor R, Pirouz S, Kamanger N et al. Agreement between central venous and arterialblood gas measurements in the intensive care unit. Clin J Am Soc Nephrol  2010;5:390-94
  27. Walkey A, Farber H, O’Donnell C. The accuracy of the central venous blood gas foracid-base monitoring. J Intensive Care Med 2010; 25:104-10
  28. Adrougue H, Rashad N, Gorin A et al.Assessing acid-base status in circulatory failure - differences between arterial and central venous blood. NEJM 1989;320:1312-16
  29. Steedman D, Roberstson C. Acid base changes in arterial and central venous bloodduring cardiopulmonary resuscitation. Arch Emerg Med 1992;9:169-76
  30. Weil M, Rackow E, Trevino R et al.Difference in acid-base state between venous andarterial blood during cardiopulmonary resuscitation. NEJM 1986;315:153-56
  31. Toftegaard M, Rees S, Andreasson S. Evaluation of a method for converting venousvalues of acid-base and oxygenation status to arterial values. Emerg Med J 2009;26:268-72
  32. Rees S, Toftegaard M, Andreasson S. A method for calculation of the values of arterialacid-base chemistry form measurements in the peripheral venous blood. Comput Methods Programs Biomed 2006;81:18-25
+ View more
References
  1. Clinical and Laboratory Standards Institute (CLSI formerly NCCLS) Procedures for the Collection of Arterial Blood Specimens; Approved standard - 4th edition H11-A4 (ISBN 1-56238-543-3) Pennsylvania USA 2004
  2. Okeson G, Wullbrecht P. The safety of brachial artery puncture for arterial  blood sampling. Chest 1998;114:748-51
  3. Giner J, Casan P, Belda J et al. Pain during arterial puncture. Chest 1996;110:1443-45
  4. Eisen L, Miami T, Berger J et al.Gender disparity in failure rate for arterial catheter attempts. J Intensive Care 2007;22:166-72
  5. Wallach SG.Cannulation injury of the radial artery: diagnosis and treatmentalgorithm.Am J Crit Care 2004;13:315-19
  6. Scheer B, Perel A, Pfeifer U. Clinical review: complications and risk factors of peripheral arterial catheters used for haemodynamic monitoring in anaesthesia and intensive care medicine. Crit Care 2002;6:199-204
  7. Garland A, Connors A. Indwelling arterial catheters in the Intensive Care Unit: Necessary and beneficial, or a harmful crutch? Am J Respir Crit Care Med 2010; 182:133-37.
  8. Jensen L, Onysklw J, Prasad N. Meta-analysis of arterial oxygen saturation by pulse oximetry in adults. Heart Lung 1998;27:367-408
  9. Keogh BF. When pulse oximetry monitoring of the critically ill is not enough. AnesthAnalg 2002;94(Suppl 1):S96-99
  10. Wilson B et al. The accuracy of pulse oximetry in emergency department patients with severe sepsis and septic shock: a retrospective cohort study. BMC Emergency Medicine 2010;10:9
  11. Brandenburg M, Dire D. Comparison of arterial and venous blood gas values in the initial emergency department evaluation of patients with diabetic ketoacidosis. Ann Emerg Med 1998;31: 459-65
  12. Rang L, Murray H, Wells G et al.Can peripheral venous blood gases replace arterial blood gases in emergency department patients. CJEM 2002; 4: 7-15
  13. Chu YC, Chen CZ, Lee CH et al.Prediction of arterial blood gas values from venous blood gas values in patients with acute respiratory failure receiving mechanicalventilation. J Formos Med Assoc 2003;102:539-43
  14. Yildizds D, YapicogluH, Yilmaz H et al. Correlation of simultaneously obtainedcapillary, venous and arterial blood gases of patients in a paediatric intensive care unit.Arch Dis Child 2004;89:176-80
  15. Ak A, Ogun C, Bayir A et al.Prediction of arterial blood gas values form venous bloodgas values in patients with acute exacerbation of chronic obstructive pulmonarydisease. Tohoku J Exp Med 2006;210:285-90
  16. Kelly A. The case for venous rather than arterial blood gases in diabetic ketoacidosis. Emerg Med Australas 2006;18:64-67
  17. Malatesha G, Singh N, Bharija A et al.Comparison of arterial and venous pH, bicarbonate, PCO2 and PO2 in initial emergency department assessment. Emerg Med J 2007;24:569-71
  18. Bilan N Behbahan A Khosroshahi A Validity of venous blood gas analysis for diagnosis of acid-base imbalance in children admitted to pediatric intensive care unit World Journal of Pediatrics 2008;4:114-117
  19. O’Connor T Barry P Jehangir A Comparison of arterial and venous blood gases and the effects of analysis delay an air contamination on arterial samples in patients with chronic obstructive pulmonary disease and healthy controls Respiration 2011;81:18-25
  20. Koul P, Khan U, Wani A. Comparison and agreement between venous and arterial gasanalysis in cardiopulmonary patients in Kashmir valley of the Indian subcontinent. Ann of Thoracic Med 2011;6:33-37.
  21. Sutton R, Wilson R, Walt A. Differences in acid-base levels and oxygen saturationbetween central venous and arterial blood. Lancet 1967;2 ;748-51
  22. Phillips B, Peretz D. A comparison of central venous and arterial blood gas values inthe critically ill. Ann Intern Med 1969;70:745-49.
  23. Malinoski D Todd S Slone S et al C orrelation of central venous and arterial blood gas measurements in mechanically ventilated trauma patients Arch Surg 2005;140:1122-25
  24. Middleton P, Kelly A-M, Brown J et al. Agreement between arterial and centralvenous values for bicarbonate base excess and lactate. Emerg Med J 2006;23:622-24
  25. Toftegaard M, Rees S, Andreassen .S Correlation between acid-base parameters measured in arterial blood and venous blood sampled peripherally form vena cavae superior and from the pulmonary artery. Euro J Emerg Med 2008; 15:86-91
  26. Tregor R, Pirouz S, Kamanger N et al. Agreement between central venous and arterialblood gas measurements in the intensive care unit. Clin J Am Soc Nephrol  2010;5:390-94
  27. Walkey A, Farber H, O’Donnell C. The accuracy of the central venous blood gas foracid-base monitoring. J Intensive Care Med 2010; 25:104-10
  28. Adrougue H, Rashad N, Gorin A et al.Assessing acid-base status in circulatory failure - differences between arterial and central venous blood. NEJM 1989;320:1312-16
  29. Steedman D, Roberstson C. Acid base changes in arterial and central venous bloodduring cardiopulmonary resuscitation. Arch Emerg Med 1992;9:169-76
  30. Weil M, Rackow E, Trevino R et al.Difference in acid-base state between venous andarterial blood during cardiopulmonary resuscitation. NEJM 1986;315:153-56
  31. Toftegaard M, Rees S, Andreasson S. Evaluation of a method for converting venousvalues of acid-base and oxygenation status to arterial values. Emerg Med J 2009;26:268-72
  32. Rees S, Toftegaard M, Andreasson S. A method for calculation of the values of arterialacid-base chemistry form measurements in the peripheral venous blood. Comput Methods Programs Biomed 2006;81:18-25
Disclaimer

May contain information that is not supported by performance and intended use claims of Radiometer's products. See also Legal info.

Chris Higgins

has a master's degree in medical biochemistry and he has twenty years experience of work in clinical laboratories.

Articles by this author
Acutecaretesting handbook

Acute care testing handbook

Get the acute care testing handbook

Your practical guide to critical parameters in acute care testing. 

Download now
Preanalytical errors

Related webinar

Preanalytical errors related to venous blood sample collection and handling

Presented by Ana-Maria Simundic, PhD, Prof. of Medical Biochemistry, Zagreb University, Zagreb, Croatia

Watch the webinar
Webinar on demand: Evolution of blood gas testing - Part 1

Related webinar

Evolution of blood gas testing Part 1

Presented by Ellis Jacobs, PhD, Assoc. Professor of Pathology, NYU School of Medicine.

Watch the webinar
Webinar: Evolution of blood gas testing - Part 2

Related webinar

Evolution of blood gas testing Part 2

Presented by Ellis Jacobs, PhD, Assoc. Professor of Pathology, NYU School of Medicine.

Watch the webinar
Preanalytical errors handbook

Blood gas Preanalytics app

Get the Blood gas Preanalytics app for your smartphone

This smartphone app focuses on the preanalytical phase of blood gas testing and what operators can do to avoid errors.

Download app

Related Articles

Sign up for the newsletter

Sign up for the newsletter

Get the latest articles, journal scans and webinar news from acutecaretesting.org directly in your inbox - up to 12 times a year.

Get the latest articles, journal scans and webinar news from acutecaretesting.org directly in your inbox - up to 12 times a year.

About this site About Radiometer Contact us Legal notice
This site uses cookies Read more