Article Text

Download PDFPDF

Evaluation of a method for converting venous values of acid-base and oxygenation status to arterial values
  1. M Toftegaard1,2,
  2. S E Rees1,
  3. S Andreassen1
  1. 1
    Center for Model Based Medical Decision Support Systems, Aalborg University, Denmark
  2. 2
    Department of Anaesthesia and Intensive Care, Aalborg University Hospital, Aalborg, Denmark
  1. Dr M Toftegaard, Department of Anaesthesia and Intensive Care, Aalborg University Hospital, DK-9220 Aalborg, Denmark; marianne.toftegaard{at}rn.dk

Abstract

Objective: This paper evaluates a method in which arterial values of pH, carbon dioxide tension (Pco2) and oxygen tension (Po2) calculated from venous values and pulse oximetry are compared with simultaneously measured arterial values.

Methods: 103 adult patients from three departments (pulmonary medicine, thoracic intensive care and multidisciplinary intensive care) were studied. The patients belonged to three groups: (1) 31 haemodynamically stable patients with a diagnosis of chronic obstructive lung disease (COLD); (2) 49 haemodynamically stable patients without COLD; and (3) 23 haemodynamically unstable patients without COLD. Arterial and venous (peripheral and, where possible, central and mixed) blood samples were taken simultaneously and anaerobically. Peripheral arterial oxygen saturation was measured with a pulse oximeter. The principle of the method is to simulate the transport of venous blood back through the tissues using the respiratory quotient (adding oxygen and removing carbon dioxide) until simulated arterial oxygenation matches that measured by pulse oximetry.

Results: Calculated values of arterial pH and Pco2 had very small bias and standard deviations regardless of the venous sampling site. In all cases these errors were within those considered acceptable for the performance of laboratory equipment, and well within the limits of error acceptable in clinical practice. In addition, the standard deviation (SD) of calculated values of pH and Pco2 was similar to the variability between consecutive arterial samples. For peripheral oxygen saturation values ⩽96%, the method can calculate Po2 with an SD of 0.93, which may be useful in clinical practice. Calculations made from peripheral venous blood were significantly more accurate than those from central venous blood.

Conclusion: Arterial pH and Pco2 can be calculated precisely from peripheral venous blood in a broad patient population. The method has potential for use as a screening tool in emergency medical departments and in medical and surgical wards to assess a patient’s acid-base and oxygenation status prior to sampling arterial blood or to help in the decision to refer the patient to the ICU. In departments where arterial blood gas values are used to monitor patients (eg, pulmonary medicine), the method might reduce the number of arterial samples taken by replacing them with peripheral venous blood samples, thus reducing the need for painful arterial punctures.

Statistics from Altmetric.com

Arterial blood gas analysis is an important tool in assessing the status of patients in intensive care units and departments of pulmonary medicine, endocrinology and nephrology. In other medical and surgical departments arterial blood sampling is not routine, and is usually first performed when the patient deteriorates or when a newly admitted patient is in a very poor condition. At this point, blood is taken via an arterial puncture with the associated pain and risk of haematoma.13

In contrast, the sampling of peripheral venous blood is routinely performed in almost all patients on admission to hospital as part of the routine diagnostic procedures and acutely, depending on the status of the patient.

It is generally accepted that venous blood—especially peripheral venous blood—does not describe the metabolic and respiratory state of the patient.47 However, recent data show that peripheral venous blood sampled anaerobically correlates reasonably well with arterial values in intensive care, pulmonary medicine and emergency medicine patients, at least for values of pH, bicarbonate (HCO3) and carbon dioxide tension (Pco2).810 This has led some authors to conclude that these venous measurements could provide an acceptable replacement to arterial measurements in patients in emergency medicine. A method has recently been proposed which may further improve the clinical usefulness of peripheral venous blood.11 Using this method, the status of the patient can be evaluated from arterial blood values which have been calculated from values measured in anaerobically sampled peripheral venous blood and arterial oxygen saturation measured using a pulse oximeter. This calculation is performed using mathematical models of acid-base chemistry12 to convert venous values of pH, Pco2 and Po2 to arterial values.

This paper evaluates this method, where arterial values calculated from venous values are compared with simultaneously measured arterial values. This evaluation was performed in adult patients with a broad range of respiratory and metabolic states.

METHODS

Patients

Arterial and venous blood samples were collected from 112 patients (39 women) of median age 66 years (range 26–81) from three different departments (pulmonary medicine, thoracic intensive care and multidisciplinary intensive care) at Aalborg Hospital, Denmark.

Data from 9 patients were excluded because of sampling and/or handling errors. For 3 patients, arterial or venous samples were missed. In another patient the fraction of inspired oxygen (Fio2) was changed during the study. The remaining 5 patients were excluded because arteriovenous haemoglobin differences exceeded 0.5 mmol/l. In one of these patients the measured venous values were such that it was clear that the sample was taken erroneously (ie, base excess (BE) −19 mmol/l in a patient with chronic obstructive lung disease (COLD) with normal haemodynamics). For the remaining 103 patients, values of acid-base and oxygenation have been published previously.8

The patients belonged to one of three groups: group 1 comprised 31 haemodynamically stable patients diagnosed as having COLD; group 2 consisted of 49 haemodynamically stable patients without COLD; and group 3 comprised 23 haemodynamically unstable patients without COLD. The arterial acid-base and oxygenation status of these three groups is summarised in table 1. The patients were used to evaluate the method for converting venous to arterial values.

Table 1 Values of pH, Pco2 and Po2 measured in arterial blood for the three groups of patients

Data acquisition

Arterial and venous blood samples were taken simultaneously (ie, within 0–10 s) from each patient. All patients had peripheral venous blood taken with a further 73 and 18, respectively, having central venous and mixed venous blood samples taken. For the 73 patients with venous blood sampled from more than one site, two arterial blood samples were taken before and after the venous samples. These arterial values were not clinically different,8 illustrating stable acid-base and oxygenation status during the study.

All blood samples were taken anaerobically and immediately analysed for acid-base and oxygenation status as described previously.8 Peripheral venous blood samples were taken from a site considered to be well perfused as necessary for application of the method, which requires minimal strong acid production at the sample site.11 The peripheral arterial oxygen saturation was measured by a pulse oximeter (COSMO Plus, Novametrix Medical Systems, Wallingford, Connecticut, USA). Figure 1 illustrates the method for calculating arterial acid-base and oxygenation status from measurements in the venous blood and pulse oximetry, described in detail elsewhere.11

Figure 1

Calculation of arterial blood acid-base status from venous blood and arterial oxygen saturation measured using a pulse oximeter. A–E represent the mathematical steps included in the method.9 BE, base excess; COHb, carboxyhaemoglobin; DPG, 2,3-diphosphoglycerate; MetHb, methoxyhaemoglobin; NBB, non-bicarbonate buffer base; Pco2, carbon dioxide tension; Po2, oxygen tension; RQ, respiratory quotient; Sao2, arterial oxygen saturation; Spo2, peripheral oxygen saturation; V̇co2, carbon dioxide production; V̇o2, oxygen consumption; Reproduced with permission of Elsevier.

The principle of the method is to calculate arterial values by simulating, with the help of mathematical models,12 the transport of venous blood back through the tissues until simulated arterial oxygenation matches that measured by pulse oximetry. In this simulation, oxygen is added and carbon dioxide removed from venous blood using a constant value of the respiratory quotient (RQ) selected here to be 0.82,11 and either zero or a small amount of strong acid removed from the blood on its simulated passage from the veins to the arteries. For peripheral blood the latter is represented as a change in the base excess (BE) across the tissue sampling site (ΔBEav) equal to 0.09 mmol/l, which is the average change in base excess measured in this patient population.8 For central and mixed venous blood ΔBEav is set to zero.8

Calculated arterial values of pH, Pco2 and Po2 were compared with the simultaneously measured arterial values using Bland-Altman plots. Values of bias and standard deviation between measured and calculated arterial values are reported. This comparison was performed for arterial values calculated from peripheral venous, central venous and mixed venous values. Comparison of the mean difference between calculated and measured arterial values were analysed for different patient groups and venous sampling sites using one-sided ANOVA.

RESULTS

Figure 2 and table 2 illustrate the difference between measured arterial values (a) and calculated arterial values (ca) using the venous to arterial conversion method for peripheral (P), central (C) and mixed (M) venous blood.

Figure 2

Bland-Altman plots comparing measured arterial (a) and calculated arterial (ca) values of pH, carbon dioxide tension (Pco2) and oxygen tension (Po2) in peripheral venous blood (P), central venous blood (C) and mixed venous blood (M). Bias (mean difference) and 95% limits of agreement (±2SD) are shown for pH and Pco2. Solid circles, patients with chronic obstructive lung disease (COLD); crosses, haemodynamically stable patients without COLD; open circles, haemodynamically unstable patients without COLD.

Table 2 Bias (2SD) for the difference between arterial and calculated arterial blood values from venous blood taken from the three sampling sites compared with laboratory performance guidelines

Calculated values of arterial pH and Pco2 had very small bias and standard deviations regardless of the venous sampling site. In all cases these errors were within those considered acceptable for the performance of laboratory equipment (table 2),13 14 and well within the limits of error acceptable in clinical practice. In addition, the SD of calculated values of pH and Pco2 given in table 2 were similar to the variability between arterial samples in the 73 patients from whom two arterial samples were taken. For pH and Pco2 the variability between arterial samples was 0.027 (2SD) and 0.446 kPa (2SD), respectively.

For Po2 the situation is a little more complex. The error in calculated Po2 values has a standard deviation which is dependent upon the Po2 level, with higher values of calculated Po2 being associated with more calculation error. Table 2 illustrates this with bias and standard deviation reported for Po2 values corresponding to oxygen saturation (Spo2) ⩽96% and ⩽97%. Errors are outside those considered acceptable for the performance of laboratory equipment but, for Spo2 ⩽96%, an error in Po2 of 1.85 kPa (2SD) may be useful in clinical practice. This will be considered further in the discussion.

Table 3 illustrates the difference between measured arterial and calculated arterial values of pH and Pco2 in each of the patient groups. This comparison was not performed for Po2 owing to the non-random distribution of errors for Po2. Errors in the calculation of arterial pH and Pco2 were not significantly different between the patient groups, regardless of sampling site, except for pH sampled from the central vein where there was a significant difference in calculated pH values between groups 1 and 3 (p = 0.03).

Table 3 Comparison between the three patient groups expressed by bias (2SD) for the difference between arterial and calculated arterial blood values from venous blood taken from the three sampling sites

Table 4 shows the difference between measured and calculated arterial values of pH and Pco2 according to sampling site. The calculated values of pH and Pco2 were significantly different across the sampling sites (p<0.05). In particular, the accuracy of pH and Pco2 calculations made from peripheral venous blood was significantly better than those from central venous blood (p<0.005 for both pH and Pco2). Comparison of the accuracy in calculations made from peripheral and mixed venous blood did not show statistical significance.

Table 4 Comparison between the three sampling sites expressed by bias (2SD) for the difference between arterial and calculated arterial blood values from venous blood

DISCUSSION

Arterial blood gas analysis is an important tool in assessing patient status, but blood sampling from arterial puncture is associated with pain and risk of haematoma. This study has evaluated a new method for calculating arterial acid-base and oxygenation status from measurements taken from the venous blood combined with pulse oximetry measurements.11 Values of arterial pH, Pco2 and Po2 have been calculated using the venous to arterial conversion method and compared with simultaneously measured arterial values. The method has been tested on a broad patient population and using blood taken from different venous sampling sites.

When compared with measured arterial values of pH and Pco2, calculated values have a precision that is within laboratory acceptable performance criteria and similar to the variability between consecutive arterial blood samples. This is true for venous blood sampled from peripheral veins, central vein or the pulmonary artery and for patients with different respiratory and haemodynamic status. Arterial values of pH and Pco2 calculated from peripheral venous blood are significantly more precise than those calculated from central venous blood. This may seem surprising, given the clinical acceptance of central or mixed venous blood in the evaluation of the acid-base status of patients, but can be explained by the selection of the sampling site. Peripheral venous blood sampled from a well-perfused extremity is unlikely to have large metabolic disturbances, which means that the base excess of arterial and peripheral venous blood is likely to be similar. This may not be true for the comparison between arterial and either central or mixed venous blood, which include blood from different organs with different metabolic status.

For Spo2 values ⩽96%, the venous to arterial conversion method can calculate Po2 with an SD of 0.93, which may be useful in clinical practice (table 2) and is similar to the variability between consecutive arterial blood samples of 0.74 seen in our data. For Spo2 values >96%, Po2 values cannot be predicted within an acceptable clinical range. This is due to the flat shape of the oxygen dissociation curve at high levels of So2, meaning that small changes in So2 result in large changes in Po2. Patients with low values of Spo2 are clinically the most interesting, and it is therefore encouraging that calculations of Po2 are most precise at low levels. The imprecision of Po2 calculations at values of Spo2 <96% might, however, be seen as limiting the usefulness of the method in calculating Po2 in undifferentiated patients receiving supplementary oxygen. Direct measurement of arterial Po2 in these patients would not, however, provide substantial extra information about the status of the patient’s lungs unless the relationship between inspiratory and arterial oxygen can be determined. In these patients supplementary oxygen is usually delivered nasally, which means that inspired oxygen levels cannot be determined precisely.

Previously, the clinical usefulness of venous blood has been evaluated by comparing simultaneously measured arterial and venous values.8 9 The method described here is a significant improvement over direct correlation between arterial and venous values. For peripheral venous blood, the precision of calculated arterial values is 0.027 (2SD) for pH and 0.517 kPa (2SD) for Pco2. These SDs are half those seen for direct correlation between measured arterial and venous blood,8 9 illustrating that the venous to arterial conversion method substantially improves the information obtained from peripheral venous blood.

This method assumes that, for a peripheral limb with a clearly recognisable pulse and a normal capillary response, the amount of acid added to the blood as it passes through the tissues is small. The method also assumes that a fixed value of RQ can be used in the calculation. In addition, using sensitivity analysis, the method has previously been shown to be insensitive to errors in the measurement of venous blood and to the effects of bubbles.11 These assumptions seem justified, given the accuracy and precision with which the method can calculate arterial values in this heterogeneous patient group.

The method has, however, previously been shown—using sensitivity analysis—to be sensitive to errors in Spo2 when calculating arterial Po2.11 Figure 3 illustrates the accuracy and precision of the measurement of Spo2 in the patient population studied here, showing a mean (SD) bias of 0.0 (1.0)%. These measurements enable calculations of arterial Po2 within ±1.85 (2SD) for Spo2 levels of ⩾96%, values which can be considered clinically useful. The precision of Spo2 measurement seen here is, however, better than that seen in other studies15 16 where a SD closer to 2% has been found. These previous studies had patient groups of similar size to the present study and similar patient heterogeneity, so the explanation for the difference in precision is not clear. Errors of 2% (SD) in Spo2 would not enable accurate calculation of arterial Po2 using the method. Calculations of arterial pH and Pco2 could, however, be performed with clinically acceptable accuracy, even with these large errors in Spo2, as shown in the previous sensitivity analysis.11

Figure 3

Bland-Altman plot comparing measured arterial oxygen saturation (Sao2) and measured peripheral oxygen saturation (Spo2). Bias (mean difference) and 95% limits of agreement (±2SD) are shown.

This paper has evaluated a method for calculating arterial values of pH, Pco2 and Po2 from measurements in venous blood and a pulse oximeter. It has been shown that values of arterial pH and Pco2 can be calculated precisely from peripheral venous blood in a broad patient population. The method has potential for use as a screening tool in both emergency medical departments and in medical and surgical wards to assess a patient’s acid-base and oxygenation status prior to the decision to sample arterial blood or to help in the decision to refer the patient to the intensive care unit. In departments where arterial blood gas values are used to monitor patients (ie, pulmonary medicine), the method might be used to reduce the number of arterial samples taken, replacing them with peripheral venous blood samples and thereby reducing the need for painful arterial punctures.

REFERENCES

Footnotes

  • Funding: This work was partially supported by a grant awarded by the IT Committee under the Danish Technical Research Council.

  • Competing interests: All authors are board members and shareholders of OBI APS. OBI APS is currently applying for a patent on the method presented in this manuscript.

  • Ethics approval: Ethical approval was obtained from the ethics committee of North Jutland and Viborg Counties.

  • Patient consent: Parental/guardian consent obtained.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • Primary survey
    Kevin Mackway-Jones