Portal de Periódicos da CAPES

Sending Blood Gas Specimens Through Pressurized Transport Tube Systems Exaggerates the Error in Oxygen Tension Measurements Created by the Presence of Air Bubbles

1995; Lippincott Williams & Wilkins; Volume: 81; Issue: 1 Linguagem: Inglês

10.1213/00000539-199507000-00037

1526-7598

Michael H. McKane, Peter A. Southorn, Paula J. Santrach, Mary F. Burritt, David J. Plevak,

Clinical Laboratory Practices and Quality Control

Our institution uses dedicated tube systems (Translogic Corp., Denver, CO) to transport laboratory specimens from patient care areas to the main laboratory. Specimens are placed in closed Perspex containers which are then moved through the tubes at approximately 24 ft/s by a combination of applied positive pressure (3-4 lb/sq in.) and subatmospheric pressure. To expedite reporting the results of arterial blood gas analysis and to improve the efficient use of our laboratory technicians' time, we investigated whether such a tube system could be used to convey arterial blood gas samples from the operating suite to a stat laboratory located in the building, 10 floors above. During this evaluation, we noted that some specimens sent through the tube system had unexpectedly high oxygen tensions. This study was designed to verify and quantify this observation. Methods This study conformed to our institution's Institutional Review Board guidelines. It was performed using surplus waste arterial blood made available after blood-gas, acid-base, and electrolyte analyses performed for patient care. Forty 4-mL pooled specimens were used for the study with this blood having no patient identifiers. Using 3-mL polypropylene heparinized (120 IU) syringes, each 4-mL aliquot was subdivided into identical 2-mL samples. Each syringe was held up-right, tapped to expel any air bubbles from the blood sample, and then capped. One of these samples was designated as the control and was hand carried on ice by a technician to the laboratory. The other sample of each pair was identified as belonging to one of four specimen groups, each group comprising 10 specimens. These groups consisted of 1) a group which was transported on ice by hand to the laboratory (manual transport); 2) a group in which the syringe cap was removed, 0.2 mL air was added using a separate tuberculin syringe, and the blood-containing syringe was then recapped and transported by hand on ice to the laboratory (manual transport with air bubble); 3) a group sent to the laboratory through the tube system (pressure tube transport); and finally, 4) a group to which 0.2 cc of air was added to the syringe prior to recapping and transported through the tube system (pressure tube transport and air bubble). Immediately as they arrived in the laboratory, each blood gas specimen was analyzed using an IL 1312 blood gas manager Trademark (Instrumentation Laboratories, Lexington, MA). The difference in the oxygen tension (PO2) and carbon dioxide tensions (PCO2) between each experimental specimen and its control was calculated (experimental-control). A one-sample paired t-test of experimental versus control was performed in each study group. Using the paired difference (experimental-control), a two-way analysis of variance was accomplished to test for interaction between the transport method (manual versus tube) and the absence or presence of air bubbles. Welch's procedure was used to accommodate unequal variances [1]. In all cases, two-sided tests were used with P values <or=to 0.5 denoting findings not attributable to chance. Results The data for PO2 are summarized in Table 1 and Figure 1. The presence of air bubbles was associated with a significant increase in the PO2 measurement with both manual and pressure transport (paired t-test and two-sample t-test: P < 0.001). A significant interaction was found between the effects of transport method and air bubbles. In the presence of air bubbles, the mean change in PO2 measurement was significantly larger with pressure tube transport (two-sample t-test: P < 0.001).Table 1: PO2: Effect of a 0.2-mL Air Bubble and Pressure Tube TransportationFigure 1: Bland-Altman difference plots showing effect of pressurized tube transportation and an air bubble on PO2 determination. Horizontal lines in each panel represent the mean and mean +/- 2 SD.The data for PCO2 are summarized in Table 2 and Figure 2. The presence of air bubbles was associated with a significant decrease in the PCO2 measurement for both manual and pressure tube transport methods (paired t-test: P < 0.001 and two-sample t-test: P < 0.001). No statistical interaction was found between the effects of transport method and air bubbles.Table 2: PCO2: Effect of a 0.2-mL Air Bubble and Pressure Tube TransportationFigure 2: Bland-Altman difference plots showing effect of pressurized tube transportation and an air bubble on PCO2 determination. Horizontal lines in each panel represent the mean and mean +/- 2 SD.Discussion Others have previously documented that air bubbles in the syringe used to transport arterial blood gas specimens can affect the PaO2 and PaCO2 measurement [2-4]. In this study the air bubbles effect on the measurement PO (2) was much more pronounced than the PCO2. Pressure tube transport exaggerated the air bubble's effect on measured PO2 but not PCO2. Incorrect handling of samples for blood gas analysis can create errors in the measured PaO2 by several mechanisms. These include 1) sample dilution by excess heparinized saline [5,6], 2) oxygen consumption due to metabolism by leukocytes, platelets, and reticulocytes [7,8], 3) diffusion of gases through plastic syringes which are semipermeable to gases [9,10], and 4) the presence of air bubbles in contact with the sample during transport [2-4]. Gas diffusion through plastic and the presence of air bubbles can also potentially affect the PaCO (2) measurement [3]. The error induced by diffusion of oxygen through plastic and with air bubbles will be exaggerated by cooling of the specimen until analysis since this cooling is associated with increased solubility of oxygen in plasma and increased oxyhemoglobin affinity [4,11]. In addition to this "temperature cycling" effect [4], the size of the error introduced by air bubbles is dependent on the relative size of the bubble(s) to the sample, the time delay to analysis, and whether mixing of the bubble and blood occurs [2,4]. Most waste blood aliquots used in this study had a lower PO2 than commonly encountered in the arterial blood of anesthetized patients. From Figure 1 it would appear that the error created by air bubbles was most pronounced at PaO (2) levels normally associated with breathing room air. At higher PO2 this influence is less and possibly not clinically important. It has been noted that equilibration of oxygen tension between blood specimens with a high PO2 and air bubbles will, as expected, lower the PO2[12] of the blood. The short duration of transportation in the pressurized tube system (approximately 35 s), capping the syringe and placement of the latter in a dedicated container for transportation did not prevent this mode of specimen transport exaggerating the error created by the presence of air bubbles in the syringe. The specimens showed no evidence of frothing which is known to cause substantial changes in the PO2 and PCO2 of the blood [2]. We assume that pressurization alone was responsible for producing this error.