SEARCH

SEARCH BY CITATION

Keywords:

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

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Authors' declaration of interests
  8. Source of funding
  9. Acknowledgements
  10. Authorship
  11. Manufacturers' addresses
  12. References

Reasons for performing study: The stability of total CO2 concentration (ctCO2) in plasma is influenced by storage temperature and handling during sample processing. Conflicting information exists regarding the stability of ctCO2 in equine plasma over time, and the effect of centrifugation on the measured value for plasma ctCO2 is unclear.

Objectives: To determine plasma ctCO2 stability over 5 days when equine blood is collected into Vacutainer tubes, centrifuged within 30 min of collection, and stored at 4°C; and to determine whether a delay in centrifugation increases the rate at which plasma ctCO2 decreases over time.

Methods: Blood was collected from 6 adult horses into 3 ml plastic Vacutainer tubes and randomly assigned to be centrifuged immediately, or after, storage. Plasma ctCO2 was measured in triplicate at 0, 24, 48, 72 and 96 h after collection using a NOVA-4 analyser. Data were analysed using multivariable linear regression, with P<0.05 being defined as significant.

Results: Plasma ctCO2 decreased linearly over time during storage at 4°C. The measured value for ctCO2 decreased at a faster rate (-0.28 mmol/l/day; P<0.0001) when centrifugation was delayed, compared with immediate centrifugation (-0.10 mmol/l/day). There was a significant effect of sequence of sample analysis on the ctCO2 value when measured in triplicate: the second and third measurements were 0.31 and 0.41 mmol/l lower than the first measurement, respectively.

Conclusions: Blood samples collected from horses into Vacutainer tubes should be centrifuged immediately after collection and analysed as soon as possible to ensure accurate values for plasma ctCO2.

Potential relevance: Failure to centrifuge, or excessive delay in measuring ctCO2 after centrifugation, produces values significantly lower than the true value.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Authors' declaration of interests
  8. Source of funding
  9. Acknowledgements
  10. Authorship
  11. Manufacturers' addresses
  12. References

Total CO2 (tCO2) is the combination of all forms of carbon dioxide in plasma that are in equilibrium in blood [1]. As such, the plasma or serum concentration of total CO2 (ctCO2) provides a clinically useful screening test for the presence of metabolic acid base disturbances, including the prerace administration of alkalinising agents to racehorses [2]. Underestimation of ctCO2is a common preanalytical error [3–6] and optimised collection and storage methods of blood samples for ctCO2 determination in horses are deemed ‘critical’[7,8].

The Clinical and Laboratory Standards Institute Procedures for the Handling and Processing of Blood Specimens recommend that serum or plasma be physically separated from contact with cells as soon as possible, unless conclusive evidence exists indicating that longer contact times do not contribute to inaccuracies [1,9]. A maximal interval of 2 h from sample collection to serum or plasma separation has been recommended [9,10], with a suggested minimum contact time of 20–30 min for serum [11]. Separation should occur as quickly as possible to prevent ongoing metabolism of cellular constituents, as well as the active and passive movement of analytes and free water between the plasma or serum and cellular compartments [12]. The International Federation of Clinical Chemistry recommends that plasma, serum or blood samples be handled and centrifuged anaerobically whenever ctCO2 is determined [13]. Most guidelines recommend centrifuging samples on a swing-out rotor centrifuge, at room temperature, for at least 10 min and with a relative centrifugal force of 1200 ± 100 g[14].

Several racetrack studies have demonstrated a marked effect of storage temperature and a mild effect of storage time on ctCO2[15–17]. Small decreases (1.1 and 1.5 mmol/l) in mean ctCO2 were observed when equine blood samples were stored at 4°C for up to 5 days with delayed centrifugation, compared with storage at 20°C or ambient temperature for 5 days with delayed centrifugation (1.6 or 5 mmol/l decrease) [15,16]. Another study documented a 1.6 mmol/l decrease in ctCO2 over 3 days in equine blood samples that were centrifuged immediately and stored at 4°C. Based on the above, we hypothesised that plasma ctCO2 would decrease slowly over time when equine blood was collected into Vacutainer tubes containing lithium heparin, centrifuged within 30 min of collection, and stored at 4°C. We also hypothesised that delayed centrifugation would increase the rate of ctCO2 decrease during storage because of the larger surface area for contact between plasma and the erythrocytes, leucocytes and platelets. Accordingly, our specific objectives were to determine plasma ctCO2 stability over 96 h when equine blood was stored at 4°C and to determine whether delayed centrifugation increases the rate at which ctCO2 decreases over time. We elected to store blood samples at 4°C because previous studies have documented increased ctCO2 stability at this temperature [15–17].

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Authors' declaration of interests
  8. Source of funding
  9. Acknowledgements
  10. Authorship
  11. Manufacturers' addresses
  12. References

Animals

Six horses (2 Standardbreds, 2 Quarter Horses and 2 Thoroughbreds, 10–23 years of age, 4 females and 2 males) were recruited from the teaching herd at Purdue University, Indiana, USA for blood collection. All horses sampled had varying degrees of recurrent airway obstruction but were in clinical remission at the time of collection. Horses were housed in a paddock and maintained on pasture with ad libitum access to water.

Sample handling and storage

The venipuncture site over the left jugular vein of each horse was cleaned with a 70% alcohol swab and venous blood collected using a 20 gauge, 1 inch (2.54 cm) Vacutainer needle (No. 367214)a and Vacutainer tube holder into 10 3 ml green top sterile (unopened) plastic Vacutainer tubes containing 51 u lyophilised (dried) lithium heparin (lithium heparin spray coated 3 ml Vacutainer tubes No. 366667)a. Blood was collected into each tube until it reached the designated line on the tube and there was approximately 2.3 ml of blood. The final mean concentration of heparin is 22.2 u/ml. Care was taken to remove the last tube from the needle before removing the needle from the jugular vein in order to prevent aspiration of atmospheric air into the Vacutainer tube. Tubes were gently inverted 8 times immediately after collection to ensure appropriate mixing of blood and anticoagulant according to the manufacturer's recommendations. Filled Vacutainer tubes were stored vertically during transport (10 min) to the laboratory in an insulated container that had an interior temperature of approximately 4°C.

Five Vacutainer tubes for each horse were randomly selected and immediately centrifuged using a fixed angle centrifuge (Group IC= immediate centrifugation) for 10 min at 1000 g and stored upright in a refrigerator at 4°C where they remained undisturbed until analysed. Total CO2 values were measured and calculated (based on measured values for plasma pH and PCO2) after 0, 24, 48, 72 and 96 h of storage at 4°C, with one tube being analysed and then discarded at each time. Five Vacutainer tubes for each horse were not centrifuged immediately (Group DC= delayed centrifugation); instead, the tubes were stored upright at 4°C and centrifuged immediately before analysis. Plasma remained in contact with the erythrocyte, leucocyte and platelet-rich fraction during storage in both groups.

Total CO2 analysis

Plasma ctCO2 was measured using a total CO2 analyser (NOVA-4)b at 0, 24, 48, 72 and 96 h after the Vacutainer tubes arrived in the laboratory (within 15 min after blood collection for time = 0 h), with time = 0 h being immediately after centrifugation in Group IC. This analyser measures ctCO2 using indirect potentiometry; the analyser acidifies the plasma sample, thereby releasing CO2 and changing the pH of the sample. The ctCO2 is calculated based on the assumption that the pH decrease is proportional to the amount of tCO2 in the sample.

The total CO2 analyser was maintained and calibrated according to manufacturer's recommendations. Aqueous tCO2 standards (5, 10, 20, 30, 40 mmol/l)c were used to confirm linearity and calibrate the analyser each day. Specifically, the measured ctCO2 value (dependent variable) was regressed against the nominal ctCO2 value (independent variable) for the 5 tCO2 standards. The regression equation estimates for the coefficient and intercept values were then used to correct the measured ctCO2 value, provided the R2 value for the linear regression equation exceeded 0.990. Upon meeting calibration requirements, the top of the Vacutainer tube was carefully removed after the appropriate storage time and 0.5 ml plasma gently aspirated from the middle third of the plasma fraction using a calibrated pipette and transferred to a new 0.5 ml polystyrene sample cupd on the total CO2 analyser. The sample cup had a height (20 mm) to diameter (7 mm) ratio of 2.86 with a measured free water loss of 2.6–2.9% per hour at room temperature [18,19]. The decrease in serum ctCO2 for 0.5 ml sample cups is 14.4% per hour at room temperature [20]; this is approximately 5 times the rate of evaporative water loss. Total CO2 was measured in triplicate by the analyser within 6 min of removing the cap of the Vacutainer tube, after which the mean ctCO2 was calculated. The analyser aspirated 220 µl of plasma and took 52 s to complete each analysis. The cap was replaced on the Vacutainer tube between successive aspirations of the three 0.5 ml plasma samples.

Plasma pH and gas analysis

pH, PCO2 and PO2 were measured on plasma from the 5 tubes that were centrifuged immediately using a blood gas analyser (ABL5 pH and blood gas analyser)e at 0, 24, 48, 72 and 96 h after arriving in the laboratory. Plasma samples for gas and pH analysis were obtained within 15 min of first removing the cap from each Vacutainer tube to obtain plasma for ctCO2 determination and then analysed immediately. This analysis was completed to determine whether the change in plasma ctCO2 during storage calculated from measured pH and PCO2 values was similar to the change in measured plasma ctCO2.

The blood gas analyser was calibrated according to manufacturer's recommendations and calibration verified with 3 quality control samples (Qualicheck)f that included low, median and high values for pH, PCO2 and PO2. The bicarbonate concentration (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 in separated plasma (pK1′= 6.105) for plasma H2CO3 and solubility of CO2 (S = 0.0307 [mmol/l]/mmHg of PCO2) in plasma at 37°C and at one atmosphere (760 mmHg) [1,21], whereby cHCO3-= S × PCO2× 10(pH − pK1′). The value for pK1′ was adjusted for changes in plasma pH, whereby pK1′= 6.105–0.044 × (pH -7.40) [22]. The value for S (measured at 760 mmHg [21]) was adjusted for the barometric pressure, Pi in mmHg, whereby Si= (Pi/760) × S. The ctCO2 (in units of mmol/l) at Pi was calculated using the Henderson–Hasselbalch equation, whereby ctCO2= Si× PCO2× (10(pH − pK1′)+ 1).

Data analysis

Data were expressed as mean ± s.d. and P<0.05 was considered significant. Multivariable linear regression analysis (PROC REG, SAS 9.2)g was used to determine the linear association between ctCO2 and storage time at 4°C separately for measured ctCO2 in Groups IC and DC (from total CO2 analyser) and calculated ctCO2 (from plasma pH and gas analysis) 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 linear regression line can be estimated [23]. Dummy variables (H1 to Hn) were defined as previously described [24]. Mixed models analysis of variance (PROC MIXED, SAS 9.2)g was used to determine whether the main effects of group (2 levels, IC, DC), time (5 levels, 0, 24, 48, 72 and 96 h), the interaction between group and time and order of analysis (3 levels, first, second and third) on ctCO2 were significant. Mixed models analysis of variance was also used to determine whether the main effect of time (5 levels) on pH, PCO2 and PO2 was significant. Bonferroni-adjusted P values were used for significance whenever post tests were conducted based on the F test for each main effect.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Authors' declaration of interests
  8. Source of funding
  9. Acknowledgements
  10. Authorship
  11. Manufacturers' addresses
  12. References

Total CO2 analysis

The interassay coefficient of variation for the 2 NOVA-4 standards over 5 consecutive days was 2.6% for ctCO2= 11.0 mmol/l and 1.9% for ctCO2= 20.6 mmol/l. The mean intra-assay coefficient of variation for triplicate measurement of ctCO2 was 1.3% (range 0.2–3.0%). A linear relationship existed between the measured ctCO2 and nominal tCO2 standards for each day, with r2 ranging from 0.9976 to 0.9998.

Plasma ctCO2, as measured by the NOVA-4, decreased linearly over time (Fig 1) and decreased at a greater rate (P<0.0001) in Group DC(-0.28 mmol/l/day) than in Group IC (-0.10 mmol/l/day). For Group IC, ctCO2= -0.0043 × time + 28.9 (r2= 0.96; P = 0.037 for slope coefficient ≠ 0) with ctCO2 in mmol/l and time in hours. For Group DC, ctCO2= -0.0118 × time + 29.0 (r2= 0.92, P = 0.0003 for slope coefficient ≠ 0).

image

Figure 1. Effects of 2 types of sample handling (delayed centrifugation, filled circles; immediate centrifugation, open circles) and time of storage on measured plasma total CO2 concentration in equine plasma (6 horses per group) as measured by the NOVA-4 analyser. Values are mean ± s.d. The linear regression line for both groups is presented (solid = group DC; dashed =Group C). The effect of storage time on the value for plasma total CO2 was calculated from plasma pH and PCO2 as measured by an ABL5 blood pH and gas analyser (filled triangles). Mean values for each day are slightly offset to improve legibility.

Download figure to PowerPoint

There was a significant effect of order on the measured value for ctCO2 (P = 0.0018 for the main effect of order); the second measurement was 0.31 mmol/l lower than the first measurement (P = 0.0043) and the third measurement was 0.41 mmol/l lower than the first measurement (P = 0.0007).

Plasma pH and gas analysis

The interassay coefficients of variation for the calculated value for ctCO2 from the blood gas analyser standards over 5 consecutive days were 8.4% for ctCO2=13.5 mmol/l, 0.9% (17.3 mmol/l) and 1.1% (26.8 mmol/l).

Blood gas analysis of plasma harvested anaerobically from centrifuged Vacutainer tubes indicated that PCO2 and PO2 increased and pH decreased (Table 1) after one day of storage at 4°C. In contrast to the decrease in measured ctCO2 over time, plasma ctCO2 calculated from blood gas and pH analysis results increased linearly over time (+0.22 mmol/l/day; Fig 1), whereby ctCO2=+0.0093 × time + 31.1, (R2= 0.95, P = 0.0009 for slope coefficient ≠ 0).

Table 1. Change in blood gas and pH values in horse blood collected into Vacutainers, immediately centrifuged and stored at 4°C for up to 4 days
TimepHPCO2PO2
  • *

    Significantly different (P<0.0125, Bonferroni adjusted) compared with time = 0 h value.

(h)(mmHg)(mmHg) 
07.485 ± 0.02140.8 ± 1.6141 ± 20
247.428 ± 0.028*46.5 ± 1.5*176 ± 14*
487.423 ± 0.023*47.5 ± 2.4*189 ± 12*
727.423 ± 0.024*48.0 ± 0.6*215 ± 17*
967.402 ± 0.028*50.8 ± 0.8*179 ± 23*

For the 6 horses at time = 0 h, the calculated value for ctCO2 (31.3 ± 1.6 mmol/l) was significantly lower (P = 0.011, paired t test) than the measured value for ctCO2 (32.7 ± 2.2 mmol/l) using the total CO2 analyser without correction using aqueous standards, and significantly higher (P = 0.0005, paired t test) than the measured value for ctCO2 (28.9 ± 2.1 mmol/l) using the corrected values for aqueous standards.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Authors' declaration of interests
  8. Source of funding
  9. Acknowledgements
  10. Authorship
  11. Manufacturers' addresses
  12. References

The first major finding of the study reported here was that blood samples collected into Vacutainer tubes should be centrifuged immediately after collection and analysed as soon as possible if accurate values for plasma ctCO2 are required. Failure to centrifuge, or excessive delay in measuring ctCO2, will result in tCO2 values significantly lower than the true value, even when samples are stored at 4°C. The second major finding was that when plasma ctCO2 was measured consecutively in triplicate from an open 0.5 ml sample cup, there was a decrease in the measured value from the first to the second measurement and an additional decrease to the third measurement. This finding raises the question as to whether duplicate, triplicate or quadruplicate measurement of ctCO2 provides the optimal method for measuring ctCO2. The third major finding was that the plasma ctCO2 calculated from pH and PCO2 values measured by a blood pH and gas analyser increased during storage at 4°C; we attributed this finding to changes in the value of S or pK1′ in plasma from the assumed fixed values of 0.0307 and 6.105 because of the movement of free water from plasma into erythrocytes.

The decrease in cHCO3 and ctCO2 in stored blood is due to anaerobic glycolysis and the formation of L-lactate and hydrogen ions; the latter is buffered by bicarbonate, resulting in decreased cHCO3 and ctCO2 and the production of CO2 gas [12]. Higher storage temperatures support a faster rate of glycolysis and therefore a faster rate of decline in cHCO3 and ctCO2[24]. Sample storage at 4°C appears optimal for minimising the rate of decline of ctCO2 in equine plasma during storage [15–17,24], as cooling reduces the metabolic rate and inhibits anaerobic glycolysis [25]. The linear decrease in plasma ctCO2 over time and the faster rate of decrease in plasma ctCO2 in the DC group most likely reflected prolonged contact time with leucocytes, erythrocytes and platelets. This result was consistent with those of 3 previous studies of human serum or plasma samples [9,10,12]. Storage of human serum or plasma samples in contact with blood cells at 25°C resulted in a significant decrease in ctCO2 after 56 h, with a faster rate of decline noted in plasma samples [12]. In addition, serum cHCO3 decreased by 4 mmol/l when human serum samples in contact with blood were stored for 24 h at 30, 32 and 37°C [9,10]. However, early centrifugation, separation and transfer of plasma to an empty tube in order to store plasma at 4°C until later analysis would also not be appropriate owing to exposure of the sample to atmospheric air.

An ongoing challenge with automated methods for measuring ctCO2 is that some of the dissolved CO2 is lost from the sample into the atmosphere because anaerobic conditions cannot be maintained during and after sample placement on the analyser [26]. The loss of dissolved CO2 reduces the measured value for ctCO2. We attempted to minimise the loss of CO2 from the measurement cup by keeping the time from opening the Vacutainer tube to tCO2 measurement to less than 10 min [26]. However, we still observed a statistically significant decrease in ctCO2 of 0.31 mmol/l and 0.10 mmol/l between sequential measurements that occurred at approximately 1 min intervals for the NOVA-4. At room temperature, the predicted decrease in ctCO2 for plasma stored in the 0.5 ml sample cups used in this study was 14.4% per hour [20], equivalent to a decrease of 0.07 mmol/l for each minute. The rate and magnitude of CO2 loss is proportional to the storage temperature and difference in sample and room air PCO2[27–29] consequently, sample cups with a larger surface area lead to greater rates of CO2 loss to the atmosphere [27]. Interestingly, open sample cups also lose H2O, resulting in volume reduction and increased ctCO2, which can partially counter CO2 loss. Evaporation can have clinically significant effects on analytical results after just 15 min [20]. Evaporative losses can be minimised by choosing sample cups of smaller diameter relative to height, as in this study, and filling the sample cup to approximately 50% [18,19]. The latter was not possible because of challenges with consistently aspirating 220 µl from a half-filled 500 µl sample cup. We attributed the higher rate of ctCO2 decrease from the anticipated value of 0.07 mmol/l to successive aspiration and pipetting of 0.5 ml plasma samples and transfer to a new sample cup [29]. We anticipate that direct aspiration of plasma from the Vacutainer tube will minimise sample to sample decreases in ctCO2.

Measurement of ctCO2 has a number of advantages over calculating of ctCO2 from the measured values for pH and PCO2[30] in that: 1) measured ctCO2 is independent of temperature, whereas calculated ctCO2 is dependent on temperature and incorrect thermostating of the measuring system may lead to inaccuracies; 2) measured ctCO2 is much less affected by metabolism than PCO2 and therefore the calculated value for ctCO2 and 3) measured ctCO2 is less influenced by the escape of CO2 than calculated ctCO2. The increase in calculated ctCO2 over time observed in this study was most likely due to buffering of protons by haemoglobin and the incorrect assumption that the value for S remained constant during storage. It has been known since 1928 that an increase in plasma protein concentration reduces the solubility of CO2 in plasma [31] without changing the value for pK1[21]. Specifically, the corrected value for S (Sc) can be calculated from the total protein concentration ([TP]) in g/l as: Sc= S × (H2Oi/H2Ostandard) = S × (1000 −[TP])/(1000–65) [31–33], where 65 g/l is the [TP] for the standardised human serum sample used to experimentally determine S [21]. In other words, the calculated value for ctCO2 will be greater than the true value whenever [TP]>6.5 g/l.

Differential effects of changes in total protein concentration on the measured values for ctCO2 have been previously reported [34]. It should be noted that the total CO2 content as measured by the NOVA-4 reflects the content per total volume of plasma and not the aqueous volume (plasma water) [33,35]. In other words, changes in plasma free water will not change the measured ctCO2 value, but will change the calculated ctCO2 value because pH and PCO2 reflect measurements of the aqueous volume (i.e. the plasma water) [35]. The difference in coefficient values between measured and calculated ctCO2 is 0.0093 − (-0.0043) = 0.0136, which is equivalent to a daily difference in the 2 methods of 0.33 mmol/l or 1.1%. Part of this increase may be the result of an increase in the total protein concentration in plasma during storage. Plasma or serum concentrations of albumin or total protein increase to a small but significant extent due to a decrease in plasma free water when unseparated blood is stored at 4°C [36,37], 25°C [12] or 32°C [9]. The decrease in plasma free water that occurs during storage is attributed to the movement of water from plasma into erythrocytes following failure of the sodium/potassium pump to maintain osmotic balance. This results in swelling of erythrocytes, haemoconcentration and a decrease in volume of the plasma water compartment after 24 h [9,36,37].

The increase in PO2 observed on blood gas analysis was most likely due to the presence of O2 in the partial vacuum in the Vacutainer tube as well as influx of atmospheric O2 across the semipermeable plastic wall of the tube [24]. The permeability of plastic to oxygen increases with cooling, with a significant increase in PO2 when Vacutainer tubes were stored at cooler temperatures, as cooling is thought to widen the pore size of plastic, thereby increasing the rate of oxygen flux based on its partial pressure gradient [24].

Protocols and laboratory equipment for measuring ctCO2 in racehorses should be standardised to minimise variability in results [7]. In some race track testing programmes, blood samples are collected on race day but analysed several days later due to transportation to off-site laboratories. We observed a slow rate of decline in ctCO2 in the study reported here despite having high precision in measured values for ctCO2 (1.9% at 20.6 mmol/l) for control samples as well as the calculated values for ctCO2 (1.1% at 26.8 mmol/l). While the overall decrease in ctCO2, the rate of decrease with delayed centrifugation and cup to cup decreases were numerically small and apparently clinically insignificant at first glance, the cumulative effects of small decreases in tCO2 due to under-filling or small Vacutainer tube size [8], too high a storage temperature, exposure of the sample to air, sample cup dimensions and increased number of analyses with subsequent exposure to air, is of concern, in that the cumulative effect could manifest as a clinically significant decrease in ctCO2. The cumulative effect of these preanalytical errors could affect the accuracy of testing programmes designed to detect the prerace administration of alkalinising agents to racehorses.

Authorship

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Authors' declaration of interests
  8. Source of funding
  9. Acknowledgements
  10. Authorship
  11. Manufacturers' addresses
  12. References

The study was designed by Drs Constable and Tinkler. Experimental data was obtained by Dr Tinkler and data analysis was completed by Dr Constable. All three authors assisted in interpreting the data and writing the manuscript.

Manufacturers' addresses

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Authors' declaration of interests
  8. Source of funding
  9. Acknowledgements
  10. Authorship
  11. Manufacturers' addresses
  12. References

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

b Nova Biomedical Corporation, Waltham, Massachusetts, USA.

c Verichem Laboratories Inc, Providence, Rhode Island, USA.

d Globe Scientific Inc, Paramus, New Jersey, USA.

e Radiometer, Copenhagen, Denmark.

f RadiometerAmerica, Westlake, Ohio, USA.

g SAS Inc, Cary, North Carolina, USA.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Authors' declaration of interests
  8. Source of funding
  9. Acknowledgements
  10. Authorship
  11. Manufacturers' addresses
  12. References
  • 1
    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.
  • 2
    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.
  • 3
    Mohler, J.G., Mohler, P.A. and Pallivathucal, R.G. (1987) Failure of the serum CO2 determined by automation to estimate the plasma bicarbonate. Scand. J. Clin. Lab. Invest. 47, 61-67.
  • 4
    Lowe, R.A., Arst, H.F. and Ellis, B.K. (1991) Rational ordering of electrolytes in the emergency department. Ann. Emerg. Med. 20, 16-20.
  • 5
    Herr, R.D. and Swanson, T. (1992a) Pseudometabolic acidosis caused by under-filling of Vacutainer tubes. Ann. Emerg. Med. 21, 177-180.
  • 6
    Herr, R.D. and Swanson, T. (1992b) Serum bicarbonate declines with sample size in Vacutainer tubes. Am. J. Clin. Pathol. 97, 213-216.
  • 7
    Ryan, D. (2000) An Australian perspective of tCO2 in harness racing. Available at: http://www.harness.org.au/hra/papers/TC02RYAN.HTM. Accessed February 13, 2011.
  • 8
    Tinkler, S.H., Couetil, L.L., Kennedy, S.A. and Constable, P.D. (2012) Effect of vacutainer tube size on total carbon dioxide concentration in equine plasma. J. Am. Vet. Med. Ass. In Press.
  • 9
    Zhang, D.J., Elswick, R.K., Miller, W.G. and Bailey, J.L. (1998) Effect of serum-clot contact time on clinical chemistry laboratory results. Clin. Chem. 44, 1325-1333.
  • 10
    Rehak, N.N. and Chiang, B.T. (1988) Storage of whole blood: effect of temperature on the measured concentration of analytes in serum. Clin. Chem. 34, 2111-2114.
  • 11
    Young, D.S. and Bermes, E.W. (1999) Specimen collection and processing: sources of biological variation. In: Tietz Textbook of Clinical Chemistry, Eds: C.A. Burtis and E.R. Ashwood, W.B. Saunders Company, Philadelphia. pp 42-72.
  • 12
    Boyanton Jr., B.L. and Blick, K.E. (2002) Stability studies of twenty-four analytes in human plasma and serum. Clin. Chem. 48, 2242-2247.
  • 13
    Burnett, R.W., Covington, A.K., Fogh-Anderson, N., Kulpmann, W.R., Lewenstam, A., Maas, A.H.J., Van Kessel, A.L. and Zijlstra, W.G. (2001) IFCC reference measurement procedure for substance concentration determination of total carbon dioxide in blood, plasma, or serum. Clin. Chem. Lab. Med. 39, 283-289.
  • 14
    Lippi, G., Salvagno, G.L., Montagnana, M. and Guidi, C. (2007) Preparation of a quality sample: effect of centrifugation time on stat clinical chemistry testing. Lab Med. 38, 172-176. DOI: 10.1309/D8TJCARUW575CXYH.
  • 15
    Lloyd, D.R., Reilly, P.J. and Rose, R.J. (1992) The detection and performance effects of sodium bicarbonate administration in the racehorse. In: Proceedings 9th International Conference of Racing Analysts and Veterinarians, New Orleans, USA, pp 131-135.
  • 16
    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.
  • 17
    Reilly, P., Duffield, A.M., Suann, C.J., Vine, J., Batty, D., Auer, D.E., Skelton, K., Tay, S., Stenhouse, A. and Ralston, J. (1997) The analysis of plasma total carbon dioxide in racehorses. In: Proceedings 11th International Conference of Racing Analysts and Veterinarians, Queensland, Australia, pp 238–240.
  • 18
    Burtis, C.A. (1990) Sample evaporation and its impact on the operating performance of an automated selective-access analytical system. Clin. Chem. 36, 544-546.
  • 19
    Burtis, C.A., Begovich, J.M. and Watson, J.S. (1975) Factors influencing evaporation from sample cups, and assessment of their effect on analytical error. Clin. Chem. 21, 1907-1917.
  • 20
    Schouwers, S., Cuypers, E., Vervaet, S., Uyttenbroeck, W. and Neels, H. (2010) Sample evaporation from pierceable cups: still an important source of analytical error. Clin. Biochem. 43, 1464-1467.
  • 21
    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.
  • 22
    Severinghaus, J.W., Stupfel, M. and Bradley, A.F. (1956) Variations of serum carbonic acid pK1′ with pH and temperature. J. Appl. Physiol. 9, 197-200.
  • 23
    Glantz, S.A. and Slinker, B.K. (1990) Primer of Applied Regression and Analysis of Variance, McGraw-Hill, New York. pp 1-777.
  • 24
    Noel, P.G., Couetil, L. and Constable, P.D. (2010) Effects of collecting blood into plastic heparinized vacutainer tubes and storage conditions on blood gas analysis values in horses. Equine Vet. J. 42 Suppl . 38, 91-97.
  • 25
    Paulsen, L. (1957) Comparison between total CO2 content (Total CO2) in plasma/serum from blood collected with or without paraffine oil. Scand. J. Clin. Lab. Invest. 9, 402-405.
  • 26
    Gambino, R.S. and Schreiber, H. (1966) The measurement of CO2 content with the Autoanalyzer. Am. J. Clin. Pathol. 45, 406-411.
  • 27
    Kirschbaum, B. (2003) Loss of carbon dioxide from serum samples exposed to air. Effect on blood gas parameters and strong ions. Clin. Chim. Acta 334, 109-113.
  • 28
    Zazra, J.J., Jani, C.M. and Rosenblum, S. (2001) Are the results of carbon dioxide analysis affected by shipping blood samples? Am. J. Kidney Dis. 37, 1105-1106.
  • 29
    Bandi, Z.L. (1981) Estimation, prevention, and quality control of carbon dioxide loss during aerobic sample processing. Clin. Chem. 27, 1676-1681.
  • 30
    Rispens, P., Kampen, E.J. and Zijlstra, W.G. Determination of total CO2 concentration in blood or plasma. National Bureau of Standards special publication 450. Proceedings of a workshop on pH and Blood Gases held at NBS, Gaithersburg, Maryland, July 7-8, 1975, Issued June 1977. pp 27-31.
  • 31
    Van Slyke, D.D., Sendroy, J., Hastings, A.B. and Neill, J.N. (1928) Studies of gas and electrolyte equilibria in blood. X. The solubility of carbon dioxide at 38°C in water, salt solution, serum, and blood cells. J. Biol. Chem. 78, 765-799.
  • 32
    Tibi, L., Bhattacharya, S.S. and Fleur, C.T.G. (1982) Variability in pK1′ of human plasma. Clin. Chim. Acta 121, 15-31.
  • 33
    Maas, A.H.S., Rispens, P., Siggaard-Andersen, O. and Zijlstra, W.G. (1984) On the reliability of the Henderson-Hasselbalch equation in routine clinical acid-base chemistry. Ann. Clin. Biochem. 21, 26-39.
  • 34
    Lustgarten, J.A., Creno, R.J., Byrd, C.G. and Wenk, R.E. (1976) Evaluation of contemporary methods for serum CO2. Clin. Chem. 22, 374-378.
  • 35
    Pichette, C., Chen, C.B., Goldstein, M., Stinebaugh, B. and Halperin, M. (1983) Influence of solutes in plasma on the total CO2 content determination: implications for clinical disorders. Clin. Biochem. 16, 91-93.
  • 36
    Clark, S., Youngman, L.D., Palmer, A., Parish, S., Peto, R. and Collins, R. (2003) Stability of plasma analytes after delayed separation of whole blood: implications for epidemiological studies. Int. J. Epidemiol. 32, 125-130.
  • 37
    Wei, Y.H., Zhang, C.B., Yang, X.W. and Ming-de, J. (2010) The feasibility of using lithium-heparin plasma from a gel separator tube as a substitute for serum in clinical biochemical tests. Lab Med. 41, 215-219.