SEARCH

SEARCH BY CITATION

Keywords:

  • pH;
  • PCO2;
  • PO2;
  • bicarbonate;
  • total CO2

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Manufacturers' addresses
  9. References

Reasons for performing study: Plastic heparinised vacutainer tubes are used for blood gas analysis in horses. This collection method may not be ideal because influx of atmospheric O2 through the permeable plastic wall of the vacutainer tube and loss of CO2 into the gas phase above the blood sample should increase blood PO2 and decrease PCO2, respectively.

Objectives: To determine the effects of collecting blood into plastic vacutainer tubes and storage conditions on blood gas analysis values.

Methods: Blood was obtained from 6 healthy horses and tonometered at 37°C with 12% O2 and 5% CO2. Three ml aliquots of tonometered blood were collected using a glass syringe or vacutainer tube and stored in iced water or at room temperature for 0, 5, 15, 30, 60 and 120 min. Blood samples from vacutainer tubes were collected aerobically (tube opened for 5 s) or anaerobically (tube remained closed). Blood gas analysis was performed in duplicate using a Radiometer ABL5. Data was analysed using repeated measures analysis of variance and P<0.05 was significant.

Results: Compared to the glass syringe, tonometered blood collected in vacutainer tubes had an immediate, significant, sustained and marked increase in PO2 and an immediate, significant, transient but small decrease in PCO2. Blood PO2 and PCO2 were higher when vacutainer tubes were stored in iced water instead of at room temperature. Measured blood pH and calculated values for plasma bicarbonate and total CO2 concentration and base excess of extracellular fluid were similar when blood was collected in glass syringes or vacutainer tubes and values were not altered by storage temperature or time.

Conclusions: Plastic heparinised vacutainer tubes should not be used to collect samples for measurement of blood PCO2 and PO2. Vacutainer tubes provide an accurate method for measuring plasma bicarbonate concentration, total CO2 concentration and base excess.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Manufacturers' addresses
  9. References

The gold standard preanalytical method for blood gas analysis is anaerobic collection and storage of blood in a glass syringe placed in iced water (Anon 2009). This method is preferred because storage at cold temperatures reduces the rate of leucocyte, thrombocyte and erythrocyte metabolism and because glass has a very low permeability to oxygen and carbon dioxide (Mahoney et al. 1991; Muller-Plathe and Heyduck 1992), thereby minimising post collection changes in pH, PCO2 and PO2. Despite the analytical advantages of glass syringes for blood gas analysis, there is continuing interest in developing alternative methods for collection and storage of blood samples because glass syringes are expensive, require sterilisation before use, are easily broken and are not widely available. Moreover, glass syringes require coating of the syringe barrel with liquid heparin which can potentially dilute the blood sample and alter measured values for blood pH, PCO2 and PO2, thereby altering the calculated values for plasma bicarbonate concentration (cHCO3), plasma total CO2 concentration (ctCO2), and base excess of extracellular fluid (BE(ecf); Hansen and Simmons 1977; Hopper et al. 2005).

Extensive research has been conducted on the use of plastic syringes for collection and storage of blood for gas analysis (Mahoney et al. 1991; Muller-Plathe and Heyduck 1992; Wu et al. 1997; Deane et al. 2004; Knowles et al. 2006; Picandet et al. 2007; Piccione et al. 2007; Rezende et al. 2007). The results of these studies have led to recommendations that plastic syringes with low gas permeability can be used for blood gas analysis as long as blood samples are stored at room temperature and analysed within 30 min of collection. Blood gas analysis should not be performed on blood samples collected in plastic syringes and stored in iced water because such storage facilitates oxygen diffusion from the atmosphere across the semi-permeable plastic barrel of the syringe into the blood sample, thereby increasing PO2 (Mahoney et al. 1991; Piccione et al. 2007; Rezende et al. 2007; Anon 2009).

Heparinised plastic vacutainer tubes are widely used to collect prerace jugular venous blood samples for ctCO2 determination (Auer et al. 1993) and a small number of investigators have used heparinised vacutainer tubes to collect jugular venous bloodsamples for blood gas analysis (Kallings and Persson 1994; Soma et al. 1996; Donovan et al. 2003). The main advantages of vacutainer tubes for blood gas analysis are wide availability, low cost, familiarity with use and ease of sample collection because aspirated bubbles do not need to be immediately expelled from the sample. Moreover, vacutainer tubes contain lyophilised heparin which prevents dilution of the blood sample by liquid heparin when blood is collected in syringes (Hansen and Simmons 1977). However, collecting blood samples into plastic vacutainer tubes may be suboptimal for blood gas analysis and measurement of ctCO2 because of the semi-permeable properties of plastic and because it is impossible to maintain an anaerobic sample (Anon 2009). Oxygen diffuses from the atmosphere across the plastic vacutainer tube wall, thereby increasing blood PO2. This phenomenon led Mueller and Lang (1973) to state that ‘the use of vacutainers for determining PO2 and per cent haemoglobin saturation is condemned’. Carbon dioxide is lost from the blood sample into the gas phase above the sample within the vacutainer tube (Mueller and Lang 1973; Berns and Matchett 1998) or into the atmosphere after opening the vacutainer tube, thereby decreasing blood PCO2, increasing blood pH and potentially changing the calculated values for cHCO3, ctCO2 and BE(ecf). Accordingly, the main objectives of the study reported here were to determine the effects of collecting blood into plastic vacutainer tubes and storage under different conditions on measured blood gas analysis values (pH, PCO2 and PO2) and calculated values (cHCO3, ctCO2, BE(ecf)).

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Manufacturers' addresses
  9. References

Animals

Six horses (2 Standardbreds, 2 Quarter Horses, 1 Arabian and 1 Paint; 4 mares and 2 geldings; ages 2–20 years; mean 12 years) were recruited from the University teaching herd. Horses were maintained on pasture and supplemented with grass hay and grain when green pasture was not sufficient to provide the necessary nutritional requirements. Animals were deemed to be healthy based on normal physical examination, normal haematology and serum biochemistry and no history of illness in the 3 months before conducting the study.

Experimental method

The venipuncture site over the jugular vein was cleaned with 70% alcohol gauze swabs and approximately 120 ml of venous blood collected from each horse using a 20 gauge 1 inch vacutainer needle (20 gauge needles #367214)1 and vacutainer tube holder into 10 ml plastic green top vacutainer tubes containing 144 USP lyophilised lithium heparin (Lithium heparin spray coated vacutainer tubes #367880)1. Blood was collected into each vacutainer tube until reaching the designated line on the tube; which was approximately 9.1 ml of blood. Care was taken to remove the last vacutainer tube from the needle before removing the needle from the jugular vein in order to prevent aspiration of atmospheric air into the vacutainer tube. Vacutainer tubes were gently inverted 8–10 times immediately after collection to ensure appropriate mixing of blood and anticoagulant according to manufacturer's recommendations. Blood samples were stored vertically during transport (10 min) to the laboratory in an insulated container with an interior temperature of approximately 22°C.

Vacutainer tubes were equilibrated in a heated water bath to approximately 37°C, gently inverted 5 times and 7 ml gently poured into a temperature-controlled rotating tonometer (IL 237 Tonometer)2 that permitted rapid equilibration of a blood sample with a calibrated gas at a specific temperature without foaming or otherwise denaturing the blood sample. Blood was tonometered for at least 18 min at 37°C with 5% CO2, 12% O2, balance N2 calibration gas3 at a flow rate of 400–450 ml/min. This calibration gas provided blood gas values within the reference range for arterial blood samples. Although vacutainer tubes are used to collect jugular venous blood and not arterial blood, a calibrated gas containing 12% O2 and 5% CO2 was selected for tonometry because of cost and availability and to assist comparison to other studies. Preliminary studies verified that equilibration of the 7 ml blood sample was complete within 18 min. Tonometry was preferred to collection of sequential arterial blood samples because tonometry provides more repeatable values for PCO2 and PO2 at 37°C; excitement of a horse during arterial blood sampling can lead to hyperventilation and decreased blood PCO2 as well as increasing the risk of aspirating air bubbles during sample collection, thereby increasing blood PO2. Moreover, sequential collection of arterial blood samples into vacutainer tubes would be problematic, even with surgical translocation of a carotid artery.

Tonometered blood samples were randomly assigned to sampling method and storage conditions using a duplicated Latin square. Three ml samples of the tonometered blood were collected by aspiration directly into a 3 ml luer slip glass syringe (3 ml luer tip glass syringe, Yale #512306)1 that did not contain heparin or into a 3 ml green top vacutainer tube containing 51 USP units of lyophilised lithium heparin (Lithium heparin spray coated vacutainer tubes #366667)1 through a 20 gauge needle (Kendall Monoject needle)4. The gas leaving the tonometer was aspirated and expelled 3 times with the glass syringe before anaerobic collection of the tonometered blood sample; all visible gas bubbles were expelled from the glass syringe within 20 s of collection and a cork placed on the end of the 20 gauge needle attached to the glass syringe in order to maintain an anaerobic condition. Tonometered blood collected into vacutainer tubes was gently inverted 8 times before storage according to manufacturer's recommendations. Care was taken to separate the needle from the vacutainer tube before removal from the tonometer in order to prevent aspiration of air. The 3 ml tonometered blood samples in glass syringes or vacutainer tubes were then stored in iced water (0°C, mixture of water and ice cubes providing sufficient time to equilibrate) or at room temperature (22°C, stored in room air) for 0, 5, 15, 30, 60 and 120 min before blood gas analysis was performed. Storage in iced water was utilised instead of placement on ice cubes because the latter method produces a wide variation in the temperature of the blood sample in the syringe (Harsten et al. 1988).

Blood gas analysis

The glass syringe and vacutainer tubes were removed from their storage conditions immediately before blood gas analysis and gently inverted and rolled 10 times in order to ensure a homogenous sample. One ml of blood was aspirated through a 20 gauge needle (Kendall Monoject needle)4 into a 1 ml plastic luer slip tuberculin syringe (1 ml Monoject luer tip tuberculin syringe, #8881501400)4 from the glass syringe (GS) via the hub of the glass syringe. Great care was taken to expel all visible air bubbles in the aspirated sample as quickly as possible (within approximately 20 s); frothing of the sample or a delay of greater than 30 s in expelling visible air bubbles can result in an increase in PO2 and a decrease in PCO2 (Biswas et al. 1982; Harsten et al. 1988).

Tonometered blood samples in vacutainer tubes were collected aerobically (vacutainer open, VO) by removing the rubber stopper for 5 s and then aspirating 1 ml of blood using a plastic tuberculin syringe (1 ml Monoject luer tip tuberculin syringe, #8881501400)4. Tonometered blood samples in vacutainer tubes were collected anaerobically (vacutainer closed, VC) by leaving the rubber stopper in place and injecting 1 ml of tonometered gas into the closed vacutainer tube and then immediately aspirating 1 ml of blood using a plastic luer tip tuberculin syringe (1 ml Monoject luer tip tuberculin syringe, #8881501400)4 attached to a 20 gauge 1 inch needle.

Tonometered blood samples were analysed in duplicate using a blood gas analyser (ABL5 pH and blood gas analyser)5. Care was taken to remove all visible air bubbles between duplicate analyses and maintain an anaerobic sample between analyses by capping the syringe with a 20 gauge needle inserted into a cork.

The cHCO3 (in units of mmol/l) was calculated using the Henderson-Hasselbalch equation, measured values for pH and PCO2 and documented values for the negative logarithm of the apparent dissociation constant (pK1′= 6.095) for plasma H2CO3 and solubility of CO2 (S= 0.0307 mmol/l/mmHg) in plasma at 37°C (Anon 2009), whereby: cHCO3-=S× PCO2× 10(pH – pK1′). The plasma total CO2 (in units of mmol/l) was calculated using the Henderson-Hasselbalch equation, whereby: ctCO2=S× PCO2× (10(pH – pK1′)+ 1) (Anon 2009). Base excess of extracellular fluid (in units of mmol/l) was calculated from the measured pH and PCO2 and documented values for pK1′ and S, such that: BE(ecf)=cHCO3- - 24.8 + 16.2 × (pH-7.40) (Anon 2009).

The blood gas analyser was calibrated using the manufacturer's recommendations. Three quality control samples (representing low, medium and high values in the physiological range for pH, PCO2 and PO2) (Qualicheck blood gas ampoules, acidaemia, normal, alkalaemia)6 were analysed each day. Barometric pressure was measured each day using a mercury barometer.

Statistical analysis

Data were expressed as least squares mean and standard error and P<0.05 was considered significant. Repeated measures analysis of variance (PROC MIXED, SAS 9.2)7 with an autoregressive covariance structure was used to examine the fixed effects of sampling method (3 levels; GS, VO, VC), storage temperature (2 levels; 0°C, 22°C), storage time (6 levels; 0, 5, 15, 30, 60, 120 min) and all possible interactions. Horse was included as a random effect. When indicated by a significant F test for a fixed effect or interaction effect, Bonferonni adjusted post test comparisons were conducted within a group to the storage time = 0 h value, or between groups at the same storage time to the value for the glass syringe stored in iced water, as this was the gold standard method.

Multivariable regression analysis (PROC REG, SAS 9.2)7 was used to determine the linear association between measured blood gas values (pH, PCO2, PO2) and storage time in the glass syringe at 0 and 22°C using dummy variable coding for each horse. This analysis of covariance approach accounts for between-subjects variability, thereby increasing the precision with which slope and intercept coefficients for the regression line can be estimated (Glantz and Slinker 1990). The approach enforces a uniform slope but a different intercept value for each horse; this approach is reasonable whenever the slopes are similar as in this study. Dummy variables (H1 through Hn) were defined in the following way: Hi= 1 if horse i (i<n), -1 if horse = n and 0 otherwise. The following regression equation was used to analyse the linear relationship between the dependent variable (y) and storage time: y = bo+ΣbiHi+ bt(time), where bo is the intercept value, bi is the coefficient value for ith horse and bt is the coefficient value for storage time. Dummy variables were entered into the model first to account for between horse differences before analysing the main factor of interest. Coefficients associated with Hi describe how much the intercept values for each horse vary from the average, but this information was of minimal interest in this study and only the estimated values of bo and bt are reported.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Manufacturers' addresses
  9. References

The blood gas analyser had acceptable precision as assessed by the coefficient of variation (CV) for quality control samples that had values approximating that of arterial blood (pH, mean = 7.348, CV = 0.09%; PCO2, mean = 43.6 mmHg, CV = 1.19%; PO2, mean = 111.5 mmHg, CV = 1.29%). Of the calculated indices of the strong ion (metabolic) component of acid base homeostasis, the CV of BE(ecf) was much greater than that of cHCO3 and ctCO2 (BE(ecf), mean =−1.7 mmol/l, CV = 29.5%; cHCO3, mean = 24.0 mmol/l, CV = 1.70%; ctCO2, mean = 25.3 mmol/l, CV = 1.64%). Barometric pressure ranged from 743–750 mmHg. Tonometered blood gas values remained stable for at least 120 min when stored in glass syringes at 0°C (Table 1, Figs 1 and 2, Fig 3).

Table 1. The effect of sampling method (3 levels; glass syringe, GS; vacutainer open, VO; vacutainer closed, VC), storage temperature (2 levels; 0°C, 22°C), time of storage (6 levels; 0, 5, 15, 30, 60, 120 min) and all possible interactions on measured blood gas values (pH, PCO2, PO2) or calculated blood gas values (plasma bicarbonate concentration, cHCO3-; plasma total CO2 concentration; ctCO2; Base excess of extracellular fluid, BE(ecf))
ItemSampling methodStorage temperatureP value
GSVCVOs.e.0oC22oCs.e.Sampling methodStorage temperatureTimeSampling method × storage temperatureSampling method × timeStorage temperature × timeSampling method × storage temperature × time
  1. Values are least squares mean. s.e. = standard error. Significantly different from GS value.

Measured blood gas values
pH7.4467.4587.4530.0017.4517.4540.0070.620.620.00470.360.460.430.14
PCO2 (mmHg)37.335.735.70.336.835.60.30.00820.0130.00070.0280.0250.130.62
PO2 (mmHg)95.4147.5145.62.0135.5123.51.6<0.00010.0021<0.00010.062<0.00010.00710.037
Calculated blood gas values
cHCO3- (mmol/l)25.925.125.10.625.824.90.50.570.0720.560.580.230.320.048
ctCO2 (mmol/l)27.026.226.20.627.026.00.50.540.0670.520.560.210.300.055
BE(ecf) (mmol/)1.91.31.10.81.91.00.60.770.120.670.660.390.410.024
image

Figure 1. Effects of 3 types of sampling method (glass syringe, GS; vacutainer open, VO; vacutainer closed, VC), 2 types of storage temperature (0°C, 22°C) and time of storage on measured blood pH. Values are least squares mean and standard error. Open circles represent blood collected in a GS and stored at 0°C, which is considered the reference method. Filled circles represent blood collected in a GS and stored at 22°C. Open triangles represent blood collected in VC and stored at 0°C. Filled triangles represent blood collected in a VC and stored at 22°C. Open squares represent blood collected in VO and stored at 0°C. Filled squares represent blood collected in a VO and stored at 22°C.

Download figure to PowerPoint

image

Figure 2. Effects of 3 types of sampling method (glass syringe, GS; vacutainer open, VO; vacutainer closed, VC), 2 types of storage temperature (0°C, 22°C) and time of storage on measured blood PCO2. Values are least squares mean and standard error. † Significantly different from blood stored at 0°C in a GS at the same time.

Download figure to PowerPoint

image

Figure 3. Effects of 3 types of sampling method (glass syringe, GS; vacutainer open, VO; vacutainer closed, VC), 2 types of storage temperature (0°C, 22°C) and time of storage on measured blood PO2. Values are least squares mean and standard error. SeeFigure 1 legend for symbol key. † Significantly different from blood stored at 0°C in a GS at the same time.

Download figure to PowerPoint

Blood pH

Least squares mean blood pH was not altered by sampling method or storage temperature (Table 1; Fig 1), but was significantly altered by time, with least squares mean pH (adjusted for sampling method and storage temperature decreasing over time (data not shown). The interaction term between sampling method and storage temperature, sampling method and time and storage temperature and time and the interaction term between sampling method, storage temperature and time were not significant (Table 1).

Blood PCO2

Blood PCO2 remained stable when stored in the glass syringe at 0°C for 120 min (P value for coefficient = 0.45), but increased linearly with time when stored in the glass syringe at 22°C: PCO2= 37.6 + 0.010 × time (P value for coefficient = 0.045) (Fig 2).

Least squares mean blood PCO2 was altered by sampling method, storage temperature and storage time (Table 1, Fig 2). There were significant interactions between sampling method and storage temperature, as well as between sampling method and storage time.

Least squares mean blood PCO2 was 1.6 mmHg lower when stored in VC or VO than in glass syringes (Table 1, Fig 2). Least squares mean blood PCO2 was 1.2 mmHg higher when stored at 0°C than at 22°C. Least squares mean blood PCO2 was lower in VC and VO at 22°C than GS at 0°C for the first 5 min and lower in VO than GS at 0°C at time = 0 min.

Blood PO2

Blood PO2 remained stable when stored in the glass syringe at 0°C for 120 min (P value for coefficient = 0.78), but decreased linearly with time when stored in the glass syringe at 22°C: PO2= 93.9 − 0.042 × time (P value for coefficient = 0.012) (Fig 3).

Least squares mean blood PO2 was altered by sampling method, storage temperature and storage time (Table 1, Fig 3). There were significant interactions between sampling method and storage time, as well as between storage temperature and storage time.

Least squares mean blood PO2 was more than 50 mmHg higher when stored in vacutainer tubes than in glass syringes, with no difference between VC and VO (Table 1, Fig 3). Least squares mean blood PO2 was 12.0 mmHg higher when stored at 0°C than at 22°C. Blood PO2 was immediately increased after collection in VC and VO and continued to increase in the following 5 min, but PO2 appeared to stabilise after this time.

cHCO3-, ctCO2, BE(ecf)

Sampling method, storage temperature and time did not change the values of cHCO3-, ctCO2 and BE(ecf) (Table 1). The interaction term between sampling method, storage temperature, and time was significant for cHCO3- (P = 0.048) and BE(ecf) (P = 0.024) but significant differences compared to blood stored in a glass syringe at 0°C were not detected for any time points.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Manufacturers' addresses
  9. References

The major findings of the study reported here were that plastic heparinised vacutainer tubes should not be used to collect blood samples for measurement of PO2 and PCO2 in horses. Collection of tonometered equine blood into plastic vacutainer tubes resulted in an immediate, sustained, marked and statistically and clinically significant increase in PO2 of more than 50 mmHg within 15 min. Collection of tonometered equine blood into plastic vacutainer tubes resulted in an immediate, transient, small and statistically significant, but clinically insignificant, decrease in PCO2. These effects were more prominent on PO2 when blood was stored at 0°C than 22°C. On the contrary, the effects of sampling method were more prominent on PCO2 when blood was stored at 22°C than 0°C. However, despite the presence of preanalytical changes in PO2 and PCO2, 3 widely used indices of the strong ion (metabolic) component of an acid-base disturbance (cHCO3, ctCO2, BE(ecf); Constable 1999a, 2000) were not affected by collection in plastic vacutainer tubes.

The between day precision of the blood gas analyser used in this study, as assessed by CV, was similar to that reported previously for pH, PCO2, and PO2 (van Kessel et al. 1987; Mahoney et al. 1991). Collection of blood into a glass syringe and storage in iced water is considered the gold standard (reference) method for blood gas analysis (Mahoney et al. 1991; Smeenk et al. 1997; Anon 2009) and was used in this study on this basis. Consistent with the findings of other investigators (Deane et al. 2004; Picandet et al. 2007), pH, PCO2 and PO2 of tonometered equine blood remained constant for at least 120 min when collected in a glass syringe and stored in iced water. We did observe a significant decrease in blood PO2 and an increase in PCO2 when the glass syringe sample was stored at 22°C. This finding was consistent with temperature-dependent metabolism of O2 by leucocytes, platelets and erythrocytes, leading to a decrease in blood PO2 and increase in PCO2. The estimated linear slope for the PO2-time relationship of equine blood at 22°C in the glass syringes used in the study reported here (−0.042 mmHg/min) were much lower than that reported for human blood kept at room temperature (−0.84 to −1.37 mmHg/min; Pretto and Rochford 1994; Smeenk et al. 1997) or horse blood kept at room temperature (−0.11 mmHg/min; Deane et al. 2004).

The marked, immediate and sustained increase in PO2 when blood was collected into plastic vacutainer tubes was most likely due to the presence of O2 in the partial vacuum within the vacutainer tube as well as influx of atmospheric O2 across the semi-permeable plastic wall of the tube. Similar findings have been observed previously when blood is collected into semi-permeable plastic syringes (Mahoney et al. 1991; Muller-Plathe and Heyduck 1992). The permeability of plastic to oxygen is increased with cooling, as observed in the study reported here with a significant increase in PO2 when vacutainer tubes were stored at 0°C instead of 22°C. Cooling of plastic is thought to widen the pore size of plastic on a nanomolar scale, thereby increasing the rate of oxygen flux based on its partial pressure gradient (Knowles et al. 2006). A relatively smaller increase in PO2 would be expected when jugular venous samples are collected in plastic vacutainer tubes because haemoglobin binds the exogenous oxygen as oxyhaemoglobin (Mahoney et al. 1991).

The small and immediate decrease in PCO2 when blood was collected into plastic vacutainer tubes was attributed to equilibration of CO2 between the plasma and overlying gas phase in the vacutainer tube, as previously described (Mueller and Lang 1973). Under filling of a vacutainer tube with blood can lead to marked decreases in ctCO2 (James et al. 1997; Berns and Matchett 1998) and PCO2 (Still and Rodman 1962); the amount lost depends on the ratio of sample volume to total tube volume (Bandi 1981). All vacutainer tubes should therefore be filled to capacity and remain capped in order to optimise the accuracy of ctCO2 and PCO2 measurement. The same equilibration phenomena should occur for oxygen, in that PO2 should drop immediately after collection as O2 in the blood diffuses into the vacutainer deadspace as demonstrated by Mueller and Lang (1973). Our failure to observe an initial decrease in PO2 indicates that the partial vacuum in the tube must contain oxygen or that oxygen flux into the tube from the atmosphere is extremely rapid.

The decrease in PCO2 when blood was collected into plastic vacutainer tubes was transient (less than 15 min) and more pronounced when vacutainer tubes were stored at room temperature rather then 0°C (Fig 2). Interestingly, the transient decrease in PCO2 has not been observed when blood is collected into glass syringes (this study) or glass, plastic or polypropylene syringes (Mahoney et al. 1991; Muller-Plathe and Heyduck 1992; Deane et al. 2004), indicating that the presence of a gas phase within the vacutainer tube plays a central role in the transient decrease in PCO2. The most likely reason for the initial decrease in PCO2 was Gay-Lussac's law, which states that the pressure of a gas in a fixed volume varies with temperature, so that the PCO2 in the gas phase is higher at 22°C than at 0°C, thereby leading to a greater reduction in PCO2 of blood in the vacutainer tube at 22°C, relative to 0°C. We believe the subsequent increase in PCO2 after 5 min of storage in vacutainer tubes is due to the differential effects of temperature on blood pH, pK1′ and S which influence the partial pressure of CO2 at equilibrium in the gas phase in the vacutainer tube. Rearrangement of the simplified strong ion equation (Constable 1997, 1999b; Constable et al. 1998) provides: PCO2= 10(pH – pK1′)× {SID –[Atot/(1 + 10pKa-pH)]}/S, where SID is the plasma strong ion difference, Atot is the total plasma concentration of nonvolatile buffers (albumin, globulins and inorganic phosphate) and pKa is negative logarithm of the effective dissociation constant for nonvolatile buffers in plasma. The value for the term {SID −[Atot/(1 + 10pKa-pH)]} will remain constant with a change in temperature (T in °K) because the value for SID and Atot will remain constant in a closed system as long as there is no change in plasma free water and because the value for pKa-pH remains constant because ΔpKa/ΔT ≈ΔpH/ΔT (Reeves 1976; Cameron 1989; Constable 1997). The simplified strong ion equation for blood stored in a vacutainer tube can therefore be expressed as: PCO2= 10(pH – pK1′)× k/S, where k is a constant. Values for pH, pK1′ and S are available at different temperatures (Austin et al. 1963; Anon 2009). The value for 10(pH – pK1′) increases with decreasing temperature (thereby increasing PCO2) because ΔpH/ΔT > ΔpK1′/ΔT (Constable 1997); however, the value for S also increases with decreasing temperature (thereby decreasing PCO2) according to Henry's law (Austin et al. 1963). The value for 10(pH – pK1′) is calculated to increase from 10(7.466 – 6.095)= 23.5 at 37°C (values in this study; pH of 7.466 is the least squares mean value at time = 0, Fig 1) to 10(7.687 – 6.145)= 34.8 at 22°C, with blood pH being corrected for temperature using Rosenthal's equation (Anon 2009) and pK1′ being corrected for pH and temperature (Austin et al. 1963). At the same time, the value for S increases from 0.0307 mmol/l/mmHg at 37°C to 0.0440 mmol/l/mmHg at 22°C. The net result of these temperature induced changes in the values for pH, pK1′ and S is that PCO2 is calculated to be 3.3% higher at 22°C than 37°C and therefore presumably higher at 0°C than 22°C, which is in agreement with our experimental findings (Table 1; Fig 2). The kinetics of the temperature-induced changes in pK1′ and S are rapid (Austin et al. 1963); the less than 15 min required to reach steady state was attributed to the time required for the temperature of the tonometered blood sample and gas within the vacutainer tube to approach that of the new storage temperature.

Statistically significant differences were not observed for blood pH in any of the samples analysed during this study (Fig 1; Table 1). This finding was similar to that in other studies (Knowles et al. 2006; Picandet et al. 2007; Piccione et al. 2007) and partially reflects the excellent buffering capacity of blood. Although a statistically significant effect of sampling method or storage temperature on blood pH was not detected (Table 1, Fig 1), the main effect of time was significant when adjusted for sampling method and storage temperature. This result was expected because erythrocyte, leucocyte and thrombocyte metabolism increases PCO2 and blood lactate concentration which are accompanied by formation of a proton. Horse-to-horse variation in blood pH due to individual variation in SID and the protein concentration in plasma (Atot; Constable 1997) is the likely reason for the observed variability in blood pH values when blood from different horses was tonometered at constant PCO2.

An important finding of this study was that the values for cHCO3, ctCO2 and BE(ecf) were not altered by collection in plastic vacutainer tubes or by storage temperature. All 3 indices of the strong ion (metabolic) component of an acid-base disturbance are heavily dependent on the value for cHCO3, which is dependent on the value for {SID –[Atot/(1+10pKa - pH)]} (Constable 1997). As discussed above, the value for [Atot/(1+10pKa - pH)] for each horse remains constant with changes in temperature and over time, and the value of SID for each horse remains constant with changes in temperature. However, because SID reflects the difference in concentration between plasma strong cations (Na, K, Ca, Mg) and strong anions (Cl, L-lactate, sulphate, ketoacids, nonesterified fatty acids, etc), SID varies from horse to horse and glycolysis during storage will lead to an increase in plasma L-lactate concentration and a decrease in SID and therefore the value for cHCO3 (Constable 1997, 1999a,b, 2000). The value for cHCO3 tended (P = 0.072) to be lower when stored at 22°C instead of 0°C and the value for ctCO2 tended (P = 0.067) to be lower when stored at 22°C instead of 0°C, presumably as a result of the higher rate of glycolysis at 22°C. These findings indicate that collecting blood into plastic vacutainer tubes and calculating values for cHCO3, ctCO2 and BE(ecf) from the results of blood gas analysis provides an accurate method for the race day detection of the administration of systemic alkalinising agents to racehorses, provided blood gas analysis is completed within 2 h of collection. Our findings also suggest that if analysis of blood samples collected into vacutainer tubes is delayed, that it would be preferable to store the vacutainer tubes at 0°C because this will minimise the rate of metabolism and subsequent reduction in cHCO3, ctCO2 and BE(ecf). Finally, because the CV for BE(ecf) was much greater than that of cHCO3 and ctCO2, calculation of cHCO3 and ctCO2 may provide more precise measurements of the strong ion (metabolic) component of an acid-base disturbance than calculation of BE(ecf). This area requires further investigation.

The effects of collecting blood into lyophilised heparin glass vacutainer tubes instead of plastic vacutainer tubes on blood gas analysis values remains to be determined. The results of the study reported here suggest that transient changes in PCO2 and PO2 will occur when blood is collected into glass vacutainer tubes because of equilibration of gas into the partial vacuum above the blood in the tube. It should be recognised that the 2009 Clinical and Laboratory Standards Institute document on approved guidelines for blood gas and pH analysis does not recommend the use of vacutainer tubes to collect samples for blood gas analysis. This is because the vacutainer sampling method does not maintain an anaerobic sample (Anon 2009).

Conflicts of interest

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Manufacturers' addresses
  9. References

Peter Noel received a stipend for a summer research fellowship from Merck-Merial in the past 2 years. The stipend was not related to the present study.

Dr Couetil has received grants from the following commercial entities in the past 2 years: Boehringer-IngelheimVetmedica, Trudell Medical International, and VibraLung Inc. These grants were not related to the present study.

Dr Constable has received a grant from Boehringer-Ingelheim in the last 2 years. This grant is unrelated to the article and deals with cattle.

Manufacturers' addresses

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Manufacturers' addresses
  9. References

1 Becton Dickinson, Franklin Lakes, New Jersey, USA.

2 Instrumentation Laboratory Inc, Lexington, Massachusetts, USA.

3 Scott Specialty Gases Inc, Plumsteadville, Pennsylvania, USA.

4 Tyco Healthcare Group LP, Mansfield, Massachusetts, USA.

5 Radiometer, Copenhagen, Denmark.

6 RadiometerAmerica, Westlake, Ohio, USA.

7 SAS Inc, Cary, North Carolina, USA.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Manufacturers' addresses
  9. References
  • Anon (2009) Blood Gas and Ph Analysis and Related Measurements; Approved Guidelines, 2nd edn., CLSI document C46-A2. Clinical and Laboratory Standards Institute, Wayne, PA.
  • Auer, D.E., Skelton, K.V., Tay, S. and Baldock, F.C. (1993) Detection of bicarbonate administration (milkshake) in Standardbred horses. Aust. vet. J. 70, 336-340.
  • Austin, W.H., Lacombe, E., Rand, P.W. and Chatterjee, M. (1963) Solubility of carbon dioxide in serum from 15 to 38 C. J. appl. Physiol. 18, 301-304.
  • Bandi, Z.L. (1981) Estimation, prevention, and quality control of carbon dioxide loss during aerobic sample processing. Clin. Chem. 27, 1676-1681.
  • Berns, S.D. and Matchett, J.L. (1998) Effect of phlebotomy technique on serum bicarbonate values. Acad. Emerg. Med. 5, 40-44.
  • Biswas, C.K., Ramos, J. and Agroyannis, B. (1982) Blood gas analysis: effect of air bubbles in syringe and delay in estimation. Br. med. J. 284, 923-927.
  • Cameron, J.N. (1989) Acid-base homeostasis: past and present perspectives. Physiol. Zool. 62, 845-865.
  • Constable, P.D. (1997) A simplified strong ion model for acid-base equilibria: application to horse plasma. J. appl. Physiol. 83, 297-311.
  • Constable, P.D. (1999a) Clinical assessment of acid-base status. Strong ion difference theory. Vet. Clin. N. Am.: Food Anim. Pract. 15, 447-471.
  • Constable, P.D. (1999b) The physicochemical approach for evaluating acid-base balance in exercising horses. Equine vet. J., Suppl. 30, 636-638.
  • Constable, P.D. (2000) Clinical assessment of acid-base status: comparison of the Henderson-Hasselbalch and strong ion approaches. Vet. clin. Pathol. 29, 115-128.
  • Constable, P.D., Hinchcliff, K.W. and Muir, W.W. (1998) Comparison of anion gap and strong ion gap as predictors of unmeasured strong ion concentration in plasma and serum from horses. Am. J. vet. Res. 59, 881-887.
  • Deane, J.C., Dagleish, M. and Benamou, A. (2004) Effects of syringe material and temperature and duration of storage on the stability of equine arterial blood gas variables. Vet. Anaesth. Analg. 31, 250-257.
  • Donovan, D.C., Hippen, A.R., Hurley, D.J. and Chase, C.C. (2003) The role of acidogenic diets and β-hydroxybutyrate on lymphocyte proliferation and serum antibody response against bovine respiratory viruses in Holstein steers. J. anim. Sci. 81, 3088-3094.
  • Glantz, S.A. and Slinker, B.K. (1990) Primer of Applied Regression and Analysis of Variance, McGraw-Hill, New York. pp 1-777.
  • Hansen, J.E. and Simmons, D. (1977) A systematic error in the determination of blood PCO2. Am. Rev. Respir. Dis. 115, 1061-1063.
  • Harsten, A., Berg, B., Inerot, S. and Muth, L. (1988) Importance of correct handling of samples for the results of blood has analysis. Acta Anaesth. Scand. 32, 365-368.
  • Hopper, K., Rezende, M.L. and Haskins, S.C. (2005) Assessment of the effect of dilution of blood samples with sodium heparin on blood gas, electrolyte, and lactate measurements in dogs. Am. J. vet. Res. 66, 656-660.
  • James, K.M., Polzin, D.J., Osborne, C.A. and Olson, J.K. (1997) Effects of sample handling on total carbon dioxide concentrations in canine and feline serum and blood. Am. J. vet. Res. 58, 343-347.
  • Kallings, P. and Persson, S. (1994) Effects of sodium bicarbonate on total carbon dioxide in blood and plasma lactate in the horse. In: Proceedings 10th International Conference of Racing Analysts and Veterinarians, Stockholm, Sweden, pp 90-93.
  • Van Kessel, A.L., Eichhorn, J.H., Clausen, J.L., Stone, M.E., Rotman, H.H. and Crapo, R.O. (1987) Inter-instrument comparison of blood gas analyzers and assessment of tonometry using fresh heparinized whole human blood. Chest 92, 418-422.
  • Knowles, T.P., Mullin, R. and Hunter, J. (2006) Effects of syringe material, sample storage time, and temperature on blood gases and oxygen saturation in arterialized human blood samples. Respir. Care 51, 732-736.
  • Mahoney, J.J., Harvey, J. and Wong, R. (1991) Changes in oxygen measurements when whole blood is stored in iced plastic or glass syringes. Clin. Chem. 37, 1244-1248.
  • Mueller, R.G. and Lang, G. (1973) Phase equilibrium of oxygen in ‘nitrogen-filled vacutainers’. Clin. Chem. 19, 1198-1200.
  • Muller-Plathe, O. and Heyduck, S. (1992) Stability of blood gases, electrolytes and haemoglobin in heparinized whole blood samples: influence of the type of syringe. Eur. J. Clin. Chem. Clin. Biochem. 30, 349-355.
  • Picandet, V., Jeanneret, S. and Lavoie, J. (2007) Effects of syringe type and storage temperature on results of blood gas analysis in arterial blood of horses. J. vet. intern. Med. 21, 476-481.
  • Piccione, G., Bertolucci, C., Grasso, F. and Giudice, E. (2007) Changes in gas composition and acid-base values of venous blood samples stored under different conditions in 4 domestic species. Vet. clin. Pathol. 36, 358-360.
  • Pretto, J.J. and Rochford, P.D. (1994) Effects of sample storage time, temperature and syringe type on blood gas tensions in samples with high partial pressures. Thorax 49, 610-612.
  • Reeves, R. (1976) Temperature-induced changes in blood acid-base status: pH and PCO2 in binary buffer. J. appl. Physiol. 40, 752-761.
  • Rezende, M.L., Haskins, S.C. and Hopper, K. (2007) The effects of ice-water storage on blood gas and acid-base measurements. J. vet. emerg. crit. Care 17, 67-71.
  • Smeenk, F.W.J., Janssen, J.D.J., Arends, B.J., Harff, G.A., Bosch, J.A., Schönberger, J.P. and Postmus, P.E. (1997) Effects of four different methods of sampling arterial blood and storage time on gas tensions and shunt calculation in the 100% oxygen test. Eur. Respir. J. 10, 910-913.
  • Soma, L.R., Uboh, C.E., Nann, L. and Gerber, A.L. (1996) Prerace venous blood acid-base values in standardbred horses. Equine vet. J. 28, 390-396.
  • Still, G. and Rodman, T. (1962) The measurement of the content of carbon dioxide in plasma. Am. J. clin. Pathol. 38, 435-439.
  • Wu, E.Y., Barazanji, K.W. and Johnson, R.L. (1997) Sources of error in A-aDO2 calculated from blood stored in plastic and glass syringes. J. appl. Physiol. 82, 196-202.