Stability of blood gases when refrigerated.
Blood gas analysis is a widely used procedure; however, in clinical practice, the physicians do not always have a blood gas analyzer in their proximity and, not infrequently, the samples are stored in a fridge or on ice and read retrospectively.
Continued anaerobic and aerobic metabolism in the blood after collection may alter blood gas composition in the interval between drawing arterial blood and its analysis, such as a fall in the Pa[O.sub.2] and pH and a rise in the PaC[O.sub.2]. These changes are temperature and time dependent (1-3). In addition, oxygen will diffuse into the sample, particularly in plastic syringes (2). However, previous research has indicated that if blood gas analysis is done within 30 minutes, there seems to be no reason to keep it on ice (2), but we found no previous studies comparing longer periods of preservation on ice versus in the fridge.
Based on the limited existing information, we designed this study to answer the following questions:
* How long does it take for blood gas composition to change?
* Which parameters change and when does that change become statistically significant?
* Are there any significant differences between keeping the samples in ice or in the fridge?
Arterial blood samples were obtained randomly from hospitalized patients as part of their normal clinical management. All patients had an arterial line placed. One milliliter (mL) of blood was collected in 1 mL heparinized (2 IU/mL) plastic syringes (RAPID Lyte[R] 1 mL L/S Syringe, Siemens).
For each patient, two arterial blood samples were collected at the same time using the arterial line. Samples were mixed and care was taken to ensure no air entered the syringes. If air bubbles were found, these were carefully removed from the syringe after each analysis to avoid equilibration with room air. Samples were analysed immediately. After the initial analysis, one sample was placed on ice (0 to +1 [degrees]C) and the other one was placed in the fridge (+4 to +8 [degrees]C). The samples were capped using filter caps that came with the syringes. These samples were submitted to serial testing at 30 minutes, 1 hour and 2 hours after the initial analysis. The pH, Pa[O.sub.2], PaC[O.sub.2], HC[O.sub.3]-, [Na.sup.+], [K.sup.+], [Ca.sup.2+] and lactate were measured.
All the determinations were performed using a RAPIDLab[R] 1200 Blood Gas Analyzer (Siemens Healthcare Diagnostics Inc., Deerfield, IL, USA) according to the manufacturer's instructions. All samples were analysed under a standard temperature (4).
The number of samples needed (n=25) to obtain potentially significant results was determined using Piface software (3). Statistical analysis was performed using Software IBM SPSS Statistics 19[R] (IBM Corporation, Somers, NY, USA).
Mean values of all parameters at 0 min, 30 min, 1 hr and 2 hr were compared between groups by an independent two-tailed t test within each group, and the magnitude of change was examined by comparing the differences in the measurements for each value at 0, 30 min, 1 hr, and 2 hr using a paired two-tailed t test. A p value of [less than or equal to]0.05 was defined as significant.
Ethics approval was obtained from the local ethics committee. Informed consent was obtained from the patients.
Two hundred arterial blood determinations from 25 patients were entered in the study - 100 measurements for each of the two experimental conditions. The mean values of Pa[O.sub.2], PaC[O.sub.2], HC[O.sub.3]-, [Na.sup.+], [K.sup.+], [Ca.sup.2+] and lactate at 0 min, 30 min, 1 hr and 2 hr were not significantly different between the two groups (Table 1).
However, within each group, significant changes were found over time in the blood gases and electrolytes. The P[aO.sub.2], [K.sup.+], [Na.sup.+] and lactate increased; PaC[O.sub.2] and [Ca.sup.2+] decreased (Table 2). The increase in the Pa[O.sub.2] was significant in both groups, compared with baseline values. These changes were noticed at 30 min in both groups, increasing even further afterwards. The decrease in PaC[O.sub.2] was statistically significant at 30 min and 2 hr in the fridge, compared to baseline. No significant differences were found when these samples were kept on ice. The pH remained stable in both groups; however the lactate levels increased significantly in both groups after 30 min. The decrease in HC[O.sub.3]- was statistically significant only after 2 hr storage in the fridge.
The [K.sup.+] started to increase significantly after 30 min on ice, but when kept in the fridge significant changes were observed only after 2 hr. [Na.sup.+] increased in samples stored on ice at 30 min and 1hr, whereas in the fridge [Na.sup.+] increased significantly throughout the three time periods. The decrease in ionized [Ca.sup.2+] was sustained in both groups, but this change appeared to begin later in the samples kept on ice.
Refrigerating the blood may delay changes in arterial blood gas composition (5,6). A robust and well conducted study by Liss and Payne analysed the changes in Pa[O.sub.2], PaC[O.sub.2] and pH at 0, 15 and 30 min (7). They found a statistically significant increase in the Pa[O.sub.2] at 15 and 30 min in both groups, and a statistically significant decrease in the PaC[O.sub.2] at 15 min in both groups. There was also a statistically significant decrease in the pH at 15 min in both groups. There were no differences between the samples stored at room temperature or in ice. In our study the changes in the blood gas composition over time were not in the direction as previously found (5, 8-9). The authors of those studies attributed their finding to the use of plastic syringes. Plastic syringes can act as semipermeable membranes and allow diffusion of gases (10). This was corroborated by another study (12).
According to Fletcher et al, the Pa[O.sub.2] of oxygenated water stored in glass syringes remains stable for 1 hr if kept on ice or at room temperature, but not in plastic syringes (8). Another study comparing plastic versus glass syringes regarding blood gas tensions in samples with high oxygen partial pressures concluded that glass syringes are superior to plastic syringes in preserving samples with a high P[aO.sub.2], and prompt and adequate cooling of such samples is essential for accurate blood gas analysis (12).
Our findings are similar to those described by Liss and Payne (7). However, instead of comparing ice with room temperature, we compared ice with refrigeration using baseline measurements as standard. We also extended the time length, and analyzed changes in the ions and lactate.
Comparisons between groups (ice vs fridge) did not show statistically significant differences. The most significant changes in both groups were a rise in lactate levels and Pa[O.sub.2] and a decrease in [Ca.sup.2+] over 2 hr. As in another study, no changes in pH, pC[O.sub.2] and HC[O.sub.3]- were found in samples stored on ice (13).
The changes in Pa[O.sub.2] are consistent and significant in both groups, i.e. independently of storing blood gas samples on ice or in the fridge, the Pa[O.sub.2] rose significantly at 30 min, 1 hr and at 2 hr. As other authors have already observed, when plastic syringes are placed on ice there is an inflow of oxygen into the samples, because iced water exhibits a very high oxygen concentration (13). Besides, as the temperature decreases, there is a shift in the oxygen-hemoglobin dissociation curve towards the left and an increase in the solubility of oxygen in the plasma, resulting in an increase of the measured Pa[O.sub.2]. A limitation is that we cannot exclude the possibility that sample manipulation and homogenization may have contributed to air entry into the sample.
The decrease in PaC[O.sub.2] could be explained using the same principle of gas equilibration between the syringe and the environment. However, the decrease in PaC02 and HC03- was significant only after 2 hr in the fridge, explaining why there were no major changes in pH.
The increase in [K.sup.+] and decrease in ionized [Ca.sup.2+] can be explained by cell lysis that occurs over time, releasing [K.sup.+] and P[0.sub.4.sup.-]. P[0.sub.4.sup.-] binds to [Ca.sup.2+], decreasing ionized [Ca.sup.2+]. In addition [K.sup.+] leaks out of red blood cells at temperatures around 4[degrees]C due to the inability of the membrane [Na.sup.+][K.sup.+]-ATPase pump to work correctly and this could also be an explanation for the [K.sup.+] results.
The increase in lactate was also expected because of anaerobic metabolism, particularly by red blood cells that utilize anaerobic pathways preferentially in their metabolism (11). Unexpectedly, a decrease in both the HC[0.sub.3.sup.-] and pH was not observed and this interesting finding may be explained by the release of intracellular HC[0.sub.3.sup.-] as cell lysis occurs.
When using plastic syringes, the arterial blood gas analysis should be processed shortly after the sample is collected. Despite the fact that low temperatures can slow down metabolism, neither the ice nor the fridge preserved all the sample characteristics. If however, the analysis of the blood gases needs to be postponed, the clinician should be aware that storage in ice or in the fridge are not very different (even though the former is possibly slightly better) and that the levels of Pa[0.sub.2] and lactate are probably overestimated.
The authors acknowledge Dr. Alvaro Ferreira for his support and help in recruiting the participants for the study.
(1.) Foster JM, Terry ML. Studies on the energy metabolism of human leukocytes. 1. 0xidative phosphorylation by human leukocyte mitochondria. Blood 1967; 30: 168-175.
(2.) Shapiro BA. Clinical application of blood gases. Year Book Medical Publishers, Chicago. 1973: 100-101.
(3.) Milkulcik P. Analises Rapid--gases no sangue e muito mais (1st ed.). Siemens Healthcare Diagnostics Inc., 2008.
(4.) Lenth RV. (2006-9). Java Applets for Power and Sample Size [Computer software]. Retrieved from http://www.stat.uiowa.edu/~rlenth/Power.
(5.) Lenfant C, Aucutt C. 0xygen uptake and change in carbon dioxide tension in human blood stored at 37 C. J Appl Physiol 1965; 20: 503-508.
(6.) Fell WL Sr. Sampling and measurement of blood gases. In: Lane EE, Walker JF Jr, Eds. Clinical Arterial Blood Gas Analysis. CV Mosby, St. Louis; 1987: 199-229.
(7.) Liss HP, Payne CP Jr. Stability of blood gases in ice and at room temperature. Chest 1993; 103: 1120-1122.
(8.) Fletcher G, Barber JL. Effect of sampling technique on the determination of Pa02 during oxygen breathing. J Appl Physiol 1966; 21: 463-468.
(9.) Andersen 0S. Sampling and storing of blood for determination of acid-base status. Scand J Clin Lab Invest 1961; 13: 196-204.
(10.) Mahoney JJ, Harvey JA, Wong RJ, Van Kessel AL. Changes in oxygen measurements when whole blood is stored in iced plastic or glass syringes. Clin Chem 1991; 37: 1244-1248.
(11.) Hess CE, Nichols AB, Hunt WB, Suratt PM. Pseudohypoxemia secondary to leukemia and thrombocytosis. N EngI J Med 1979; 301: 361-363.
(12.) Pretto JJ, Rochford PD. Effects of sample storage time, temperature and syringe type on blood gas tensions in samples with high oxygen partial pressures. Thorax 1994; 49: 610-612.
(13.) Beaulieu M, Lapointe Y, Vinet B. Stability of P02, PC02 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: 101-107.
Joao Pedro Ferreira, MD, Internal Medicine Resident  Sara Vieira Silva, MD, Internal Medicine Resident  Patrfcia Rodrigues, MD, Cardiology Resident [1,2] Miguel Araujo Abreu, MD, Infectious Diseases Resident [1,3] Jose Miguel Maia, MD, Internal Medicine Resident (1) Daniela Carvalho, MD, Internal Medicine Resident  Lufsa Carvalho, MD, Internal Medicine Hospital Assistant 
 Internal Medicine Department,  Cardiology Department and  Infectious Diseases Department, Centro Hospitalar do Porto, Porto, Portugal.
JPF conceived the study, contributed to the analytic work, data acquisition and data analysis, and substantially drafted the main article. SVS contributed to the analytic work, data acquisition and data analysis, and added critical content to the article. PR contributed to data acquisition and added critical content to the article. MAA conceived the study and added critical content to the article. JMM, DC and LC added critical content to the article. The authors declare no conflicts of interest.
Author for correspondence
Dr Joao Pedro Ferreira, Internal Medicine Department, Centro Hospitalar do Porto, Largo Prof. Abel Salazar, 4099-001 Porto, Portugal. Email: firstname.lastname@example.org
Table 1. Blood gas values at 0, 30 min, 1 hr and 2 hr. Comparisons between groups. Blood gas parameter Time Ice pH 0 7.39 [+ or -]0.09 30 min 7.40 [+ or -]0.09 1 hr 7.40 [+ or -]0.09 2 hr 7.39 [+ or -]0.10 Pa[O.sub.2] mm hg 0 71.96 [+ or -]14.0 30 min 77.09 [+ or -]16.22 1 hr 84.62 [+ or -]21.29 2 hr 92.49 [+ or -]25.16 PaC[O.sub.2] mm hg 0 42.25 [+ or -]25.54 30 min 44.66 [+ or -]24.86 1 hr 44.56 [+ or -]25.36 2 hr 44.44 [+ or -]25.31 HC[O.sub.3]- 0 25.99 [+ or -]8.05 mmol/L 30 min 25.84 [+ or -]8.13 1 hr 25.92 [+ or -]8.02 2 hr 25.54 [+ or -]7.99 [K.sup.+] 0 4.17 [+ or -]0.91 mmol/L 30 min 4.28 [+ or -]0.94 1 hr 4.41 [+ or -]0.93 2 hr 4.53 [+ or -]0.94 [Na.sup.+] 0 138.08 [+ or -]5.96 mmol/L 30 min 138.90 [+ or -]5.58 1 hr 139.95 [+ or -]6.24 2 hr 139.02 [+ or -]5.24 Ca[2.sup.+] 0 1.13 [+ or -]0.09 mmol/L 30 min 1.10 [+ or -]0.10 1 hr 1.07 [+ or -]0.10 2 hr 1.01 [+ or -]0.10 Lactate 0 1.31 [+ or -]0.60 mmol/L 30 min 1.47 [+ or -]0.61 1 hr 1.59 [+ or -]0.60 2 hr 1.83 [+ or -]0.52 Blood gas parameter Time Fridge p pH 0 7.40 [+ or -]0.09 NS 30 min 7.41 [+ or -]0.09 NS 1 hr 7.40 [+ or -]0.09 NS 2 hr 7.40 [+ or -]0.10 NS Pa[O.sub.2] mm hg 0 72.25 [+ or -]12.90 NS 30 min 78.75 [+ or -]17.07 NS 1 hr 84.84 [+ or -]21.55 NS 2 hr 95.39 [+ or -]27.15 NS PaC[O.sub.2] mm hg 0 43.92 [+ or -]25.26 NS 30 min 43.07 [+ or -]24.05 NS 1 hr 43.84 [+ or -]24.99 NS 2 hr 42.96 [+ or -]24.64 NS HC[O.sub.3]- 0 25.10 [+ or -]7.47 NS mmol/L 30 min 24.87 [+ or -]7.10 NS 1 hr 24.99 [+ or -]7.62 NS 2 hr 24.32 [+ or -]7.04 NS [K.sup.+] 0 4.03 [+ or -]1.00 NS mmol/L 30 min 3.98 [+ or -]0.99 NS 1 hr 4.04 [+ or -]1.01 NS 2 hr 4.37 [+ or -]1.19 NS [Na.sup.+] 0 138.58 [+ or -]6.00 NS mmol/L 30 min 139.42 [+ or -]6.49 NS 1 hr 140.34 [+ or -]6.75 NS 2 hr 141.84 [+ or -]6.34 NS Ca[2.sup.+] 0 1.11 [+ or -]0.09 NS mmol/L 30 min 1.06 [+ or -]0.11 NS 1 hr 1.06 [+ or -]0.12 NS 2 hr 0.99 [+ or -]0.18 NS Lactate 0 1.32 [+ or -]0.59 NS mmol/L 30 min 1.47 [+ or -]0.64 NS 1 hr 1.64 [+ or -]0.62 NS 2 hr 1.93 [+ or -]0.67 NS NS=not significant Table 2. Changes in blood gas values from 0 min to 30 min, 1 hr and 2 hr. Comparisons within groups. Values in Blood gas parameter comparison Ice p pH 0 and 30 min 0.01 [+ or -]0.03 NS 0 min and 1hr 0.01 [+ or -]0.03 NS 0 min and 2 hr 0.01 [+ or -]0.03 NS Pa[O.sub.2] mm hg 0 and 30 min 5.02 [+ or -]4.31 <0.001 0 min and 1hr 12.39 [+ or -]13.21 <0.001 0 min and 2hr 20.53 [+ or -]15.71 <0.001 PaC[O.sub.2] mm hg 0 and 30 min -0.58 [+ or -]2.16 NS 0 min and 1hr -0.68 [+ or -]2.23 NS 0 min and 2hr -0.80 [+ or -]2.65 NS HC[O.sub.3]-mmol/L 0 and 30 min -0.14 [+ or -]0.93 NS 0 min and 1hr -0.07 [+ or -]1.12 NS 0 min and 2hr -0.45 [+ or -]1.21 NS [K.sup.+] mmol/L 0 and 30 min 0.11 [+ or -]0.26 =0.05 0 min and 1hr 0.24 [+ or -]0.28 <0.001 0 min and 2hr 0.36 [+ or -]0.34 <0.001 [Na.sup.+] mmol/L 0 and 30 min 0.84 [+ or -]1.00 <0.001 0 min and 1hr 1.86 [+ or -]1.97 <0.001 0 min and 2hr 0.94 [+ or -]5.41 NS Ca[2.sup.+] mmol/L 0 and 30min -0.03 [+ or -]0.06 NS 0 min and 1hr -0.06 [+ or -]0.05 <0.001 0 min and 2hr -0.12 [+ or -]0.07 <0.001 Lactate mmol/L 0 and 30min 0.15 [+ or -]0.13 <0.001 0 min and 1hr 0.27 [+ or -]0.18 <0.001 0 min and 2hr 0.46 [+ or -]0.22 <0.001 Values in Fridge p Blood gas parameter comparison 0.02 [+ or -]0.04 NS pH 0 and 30 min 0.01 [+ or -]0.03 NS 0 min and 1hr 0.01 [+ or -]0.04 NS 0 min and 2 hr 6.28 [+ or -]7.10 <0.001 Pa[O.sub.2] mm hg 0 and 30 min 12.34 [+ or -]11.19 <0.001 0 min and 1hr 23.27 [+ or -]17.01 <0.001 0 min and 2hr -0.84 [+ or -]2.09 0.05 PaC[O.sub.2] mm hg 0 and 30 min -0.07 [+ or -]3.86 NS 0 min and 1hr -1.04 [+ or -]2.27 =0.03 0 min and 2hr -0.26 [+ or -]0.89 NS HC[O.sub.3]-mmol/L 0 and 30 min -0.09 [+ or -]1.78 NS 0 min and 1hr -0.80 [+ or -]0.74 <0.001 0 min and 2hr -0.07 [+ or -]0.18 NS [K.sup.+] mmol/L 0 and 30 min 0.01 [+ or -]0.19 NS 0 min and 1hr 0.32 [+ or -]0.77 0.05 0 min and 2hr 0.83 [+ or -]1.90 0.04 [Na.sup.+] mmol/L 0 and 30 min 1.71 [+ or -]2.58 0.003 0 min and 1hr 3.28 [+ or -]2.90 <0.001 0 min and 2hr -0.05 [+ or -]0.05 <0.001 Ca[2.sup.+] mmol/L 0 and 30min -0.06 [+ or -]0.05 <0.001 0 min and 1hr -0.15 [+ or -]0.12 <0.001 0 min and 2hr 0.15 [+ or -]0.12 <0.001 Lactate mmol/L 0 and 30min 0.31 [+ or -]0.17 <0.001 0 min and 1hr 0.55 [+ or -]0.29 <0.001 0 min and 2hr NS=not significant
|Printer friendly Cite/link Email Feedback|
|Author:||Ferreira, Joao Pedro; Silva, Sara Vieira; Rodrigues, Patricia; Abreu, Miguel Araujo; Maia, Jose Migu|
|Publication:||New Zealand Journal of Medical Laboratory Science|
|Date:||Aug 1, 2012|
|Previous Article:||Attenuation of chloroquine-induced hepatoxicity and renal damage by Gnetum bucholzianum leaf extract.|
|Next Article:||Evaluation of laboratory request forms for incomplete data at a rural tertiary hospital in Nigeria.|