To the Editor,
Venous blood gas (VBG) analysis including pH, partial pressure of carbon dioxide (pCO2) and calculated actual bicarbonate (cHCO3 −) have become a common alternative to arterial blood gas (ABG) analysis during investigation and monitoring of metabolic disturbances and pulmonary function in hospitalized patients, when arterial oxygen is not required. In practice venous punctures are less technically demanding and associated with lower levels of pain and fewer complications than arterial punctures . Studies have also documented a close correlation between arterial and venous acid/base-values, including lactate, in stable patients .
Recommended sample handling and pre-analytical factors that impact the stability of ABG-samples are well documented  and are likely interchangeable in with VBG-samples if factors related to oxygen stability are ignored. A search through laboratory handbooks from 19 clinical laboratories operating in Norwegian hospitals revealed a common non-compliance to these recommendations when handling VBG-samples (literary search autumn 2018). Six out of the eight laboratories with published procedures for VBG-testing opted to use evacuated lithium-heparin tubes instead of recommended air-tight blood gas syringes (BGS), and two out of eight laboratories stored VBG-samples on ice.
Accordingly, we undertook a study investigating the stability of the acid/base parameters pH, pCO2, cHCO3 −, and lactate in VBG-samples at room temperature vs. ice and BGS vs. lithium-heparin tubes. We also included both healthy subjects and end-stage renal disease (ESRD) pre-dialysis patients with metabolic acidosis. Current knowledge indicates that biological differences between patients (e.g. increased leukocyte concentration) may be important . To our knowledge no studies have investigated if an acid/base disturbance may influence the stability of VBG.
A total of 26 healthy subjects and 10 dialysis patients volunteered to participate. Exclusion criteria for the healthy participants were: pregnancy, breastfeeding or smoking. None of the healthy volunteers consumed alcohol and all felt well the last 24 h before blood sampling. The dialysis patient should feel clinically stable at the day of sampling, and stable condition was also confirmed by the attending nurse. The project was considered a quality assurance projected by the regional Ethics Committee and approved by the local data protection officer. All participants were informed about the project and provided oral consented to participate.
Venipuncture was performed with Vacutainer, Safety-Lok (BD). A discard tube was used to remove air from the Safety-Lok tubing. For each healthy subject blood samples were collected in both lithium-heparin vacutainer tubes (Vacuette Ref. 454,029, 72 IU lithium heparin, Greiner Bio-One, Kremsmünster, Austria) and BGS (SafePico Ref. 956-62280, 80 UI electrolyte balanced heparin, Radiometer, Copenhagen, Denmark). Half of the samples were stored on crushed ice (0 °C) and half at room temperature (18–22 °C). We analyzed one lithium-heparin tube and one BGS immediately, to be used as the reference measurement. Remaining samples were analyzed consecutively on 15 min intervals up to 60 min. Blood sampling of the ESRD-patients were performed preceding dialysis treatment. The sampling protocol included only BGS-samples and were otherwise identical as for the healthy participants.
For the blood gas analysis two identical blood gas analyzers were used (AB L - 800 FLEX, Radiometer Copenhagen, Denmark). Both were located near the sampling site.
The instrument performance were validated with daily internal quality control materials and routine correlation measurements, using patients ABG-samples, between the blood gas analyzers.
During analysis samples where rejected for one of the following reasons: clot formation, presence of air bubbles in BGS, poor blood flow during sampling or instrumental errors. All discarded samples belonged to healthy subjects, leaving n=26 heparin tubes in room temperature, n=25 heparin tubes on ice, n=22 BGS in room temperature and n=23 BGS on ice to be included in this group.
For all stored samples, the results for each component were calculated as a percentage change from the reference sample. To avoid the logarithmic scale of pH the results were converted to hydrogen (H+) concentration before calculating percentage change.
Two quality criteria had to be met before a component was considered stable: (1) the mean percentage change, including a 90% confidence interval, should be within limits of the allowable bias, and (2) the percentage change in 95% of individual samples had to be within limits of allowable total error (TEa).
Maximum allowable bias and TE, were calculated based on within-subject- and between-subject biological variation as described by Fraser .
Searching the European Federation of Laboratory Medicine biological variation database and PubMed only two applicable biological variation studies were found. One investigating the biological variation of H+, pCO2, and cHCO3 −  and the other of lactate .
To determine if there were significant differences in the bias between BGS from healthy- and ERSD patients, a statistical comparison was performed using an independent t-test (p<0.05) or Shapiro-Wilk test (p<0.05). Statistical calculations were performed with Microsoft Excel (2008 version) and IBM SPSS statistics (version 26). Baseline results in BGS for the two groups, and results from statistical comparison, can be seen in Figure 1.
The main finding in this study is that almost all venous acid/base components were unstable after a short period (<15 min) when collected in heparin tubes (Table 1), the only exception was cHCO3 − being stable up to 15 min at room temperature. This rapid decline in stability is likely caused by evaporation of dissolved CO2 in the blood into the small air volume that remains in the tube, which is observed as a gradual reduction of pCO2 (Table 1). The loss of CO2 will further influence both pH and cHCO3 − as H+ and HCO3 − are linked to CO2 in the bicarbonate-buffer system.
Chosen quality criteria states that the mean deviation from the reference sample should be less than the allowable bias (including a 90% confidence interval) and 95% of individual samples should be within the limits of allowable total error (TEa). Conditions where both criteria were met are marked with a grey background, whereas aIndicates an exceeded quality criterion. Absolute quality criteria (in brackets) are estimated by quality criteria % *mean of reference samples.
The use of BGS improved stability for most components. At room temperature pH (H+) was stable for 15 min, pCO2 for 45 min, and cHCO3 − for 45 min (Table 1). When BGS-samples were stored on ice, pH (H+) and cHCO3 − were stable for 60 min. Bias for pCO2 was within the acceptance limit, while the TEa-limit was marginally exceeded with 91% of samples within the limit (Table 1). For lactate none of the acceptance criteria were met, although cooling on crushed ice caused a clear reduction in lactate increase (Table 1) as cell metabolism is slowed, leading to stable concentrations after 15 min of cooling.
A notable difference between healthy and ESRD-patients is the increase in pCO2 and cHCO3 − which is seen exclusively in samples from the last group (Figure 1). This could arise from a reduced capacity in the closed protein/phosphate blood-buffer systems (caused by the chronic metabolic acidosis) leading to more H+ being buffered through the open mechanism of the bicarbonate-buffer system. Although these differences where statistically significant the confidence intervals are generally overlapping. From a clinical perspective it is unclear if the magnitude of bias could be important, suggesting that larger studies are needed to draw a conclusion.
Based on our findings BGS on ice are considered the optimal storage condition for VBG-samples. Sixty minutes storage for these samples, as previously suggested for ABG-samples , , , could be considered if wider criteria are found acceptable; as only small changes occur in pCO2 and lactate in the interval between 15 and 60 min in BGS when stored on ice. Based on the data provided laboratories may interact with local clinicians and consider widening the acceptance criteria based on local clinical needs.
We would like to thank the staff of the dialysis units at Haukeland university hospital for coordinating the subjects and all biomedical laboratory scientists assisting in blood sampling.
Research funding: None declared.
Author contributions: All authors have accepted responsibility for the entire content of this manuscript and approved its submission.
Competing interests: Dr. Aakre has served on advisory boards for Roche Diagnostics and received lecture fees from Siemens Healtiners. Remaining authors state no conflict of interest.
Informed consent: Informed consent was obtained from all individuals included in this study.
Ethics approval: The regional Ethics Committee deemed this study a quality assurance project and exempt from review. The study was approved by the local institutional data protection officer.
1. Chauvin, A, Javaud, N, Ghazali, A, Curac, S, Altar, A, Ali, T, et al.. Reducing pain by using venous blood gas instead of arterial blood gas (VEINART): a multicentre randomised controlled trial. Emerg Med J 2020 Aug 5. https://doi.org/10.1136/emermed-2019-209287 [Epub ahead of print]. Search in Google Scholar
2. Middleton, P, Kelly, A, Brown, J, Robertson, M. Agreement between arterial and central venous values for pH, bicarbonate, base excess, and lactate. Emerg Med J 2006;23:622–4. https://doi.org/10.1136/emj.2006.035915. Search in Google Scholar
4. Fraser, C. Biological variation: from principles to practice. Washington, DC: AACC Press; 2001. Search in Google Scholar
5. Harding, P, Fraser, C. Biological variation of blood acid-base status: consequences for analytical goal-setting and interpretation of results. Clin Chem 1987;33:1416–8. https://doi.org/10.1093/clinchem/33.8.1416. Search in Google Scholar
7. Arbiol-Roca, A, Imperiali, CE, Dot-Bach, D, Valero-Politi, J, Dastis-Arias, M. Stability of pH, blood gas partial pressure, hemoglobin oxygen saturation fraction, and lactate concentration. Ann Lab Med 2020;40:448456. https://doi.org/10.3343/alm.2020.40.6.448. Search in Google Scholar
8. Mohammadhoseini, E, Safavi, E, Seifi, S, Seifirad, S, Firoozbakhsh, S, Peiman, S. Effect of sample storage temperature and time delay on blood gases, bicarbonate and pH in human arterial blood samples. Iran Red Crescent Med J 2015;17:e13577. https://doi.org/10.5812/ircmj.13577. Search in Google Scholar
9. Beaulieu, M, Lapointe, Y, Vinet, B. Stability of PO2, PCO2, and pH in fresh blood samples stored in a plastic syringe with low heparin in relation to various blood-gas and hematological parameters. Clin Biochem 1999;32:101–7. https://doi.org/10.1016/s0009-9120(98)00098-8. Search in Google Scholar
© 2021 Eirik Åsen Røys et al., published by De Gruyter, Berlin/Boston
This work is licensed under the Creative Commons Attribution 4.0 International License.