Sources of preanalytical errors in blood gas and electrolyte testing
There are numerous sources of preanalytical errors in laboratory testing. Among these are the following:
Wrong test ordered
Wrong patient collected
Error in sample collection, including improper patient preparation
Errors during specimen transport
Errors during specimen processing
This chapter will address only the last three of these. As noted by Karon ( ) and Azman et al. ( ) , the top five reasons for redrawing specimens are shown in Table 10.1 .
|Reason for redraw from ED||% of ED redraws|
|Some contamination of specimen suspected||4|
|Sample never received in lab||2|
Clearly, hemolysis and cell leakage or rupture are major causes of preanalytical errors. Contamination of a sample could include improper anticoagulant, air bubbles in the sample and contamination with IV fluids, such as saline, anticoagulants, or drugs.
In vitro hemolysis is by far the most common preanalytical problem in clinical laboratory testing. It causes problems in (a) preventing hemolysis during collection, (b) detecting hemolysis in blood samples, and (c) both interpreting results and deciding whether to report the results for many tests, most notably potassium.
Preventing in vitro hemolysis . To prevent hemolysis, several guidelines may be followed:
Use a 23 gauge or larger needle. Smaller gauge needles can damage RBCs passing through the needle.
Collecting blood through an IV infusion line often causes a higher rate of hemolysis.
When mixing blood with anticoagulant, gently invert the tube. Never shake the tube.
For several reasons, blood collection tubes should be filled to their stated capacity.
Never forcefully expel blood from a syringe into a tube, especially an evacuated tube.
Detecting hemolysis: plasma/serum or whole blood . In former years, hemolyzed serum or plasma in evacuated tubes was usually detected by the watchful eye of skilled clinical laboratory scientists who manually handled centrifuged blood specimens. They would then grade the level of hemolysis, such as slight, moderate, or gross, then follow the laboratory’s protocol for whether to report none, some, or all of the results, or request a new specimen be collected. However, visual detection of hemolysis is subject to individual interpretation and variability. Furthermore, elevated bilirubin, a common occurrence in neonates, can further complicate the visual interpretation of hemolysis ( ) .
As automation has become common in central laboratories, the samples are centrifuged, analyzed, and often reported without visual inspection of the plasma or serum. This issue has largely been solved by modern analyzers that use spectrophotometry to detect hemolysis and assign a numerical grade to its severity to alert the laboratory if the level of hemolysis significantly affects any test results. A highly debated issue is whether to report a result that is urgently requested by the physician ( ) . This is also a significant concern for any testing done at point-of-care where whole blood is analyzed and hemolysis would not likely be detected by the user. There is one commercial system that advertises a separate device that is able to detect hemolyzed blood at the point of care ( www.hemcheck.com ).
Detecting hemolysis is a significant concern for blood gas samples that are collected in syringes and must be analyzed immediately and without centrifugation. Thus, visual detection of hemolysis is not a practical option before analyzing the sample. A less than ideal option is to rapidly centrifuge the specimen after the analysis if either the results (i.e., an elevated K) or the patient’s previous samples suggest hemolysis is present.
As yet, there are no blood gas analyzers that have built-in systems to detect hemolysis in uncentrifuged blood during analysis, although companies are likely working on systems that detect hemolysis during the testing process. Much of the technology is still proprietary, but the basic technology used for detection is likely based on either: (1) isolating plasma from the flow system during or before the measurement process and then using optical methods to detect levels of hemolysis; or (2) using an algorithm based on multiple analyte inputs to predict the degree of hemolysis in whole blood. While the first is similar to that used in automated clinical chemistry analyzers, the novel approach for blood gas analyzers would focus on in-line, real-time plasma isolation to determine a sample’s hemolysis index. The second may have limited reliability in its reliance on whole blood alone, given the known pathological variations and nonlinear relationship between analyte changes in the presence of in vitro hemolysis. As blood gas testing continues to shift to the point of care, whatever real-time detection system of hemolysis is used, it must be both reliable and simple and have flexibility in how to alert clinicians to the impact on critical results such as potassium.
In vivo hemolysis occurs when erythrocytes are ruptured in circulation. Causes include immune reactions with cells, hemolytic anemias, and mechanical rupture from cardiac bypass, ECMO, or heart valve devices. While in vitro hemolysis that falsely affects test results is much more common and accounts for 98%–99% of hemolyzed specimens, results on samples with in vivo hemolysis will likely give a true physiologic increase of analytes in the blood. Thus, laboratory results such as potassium are appropriate and should be reported. Detecting in vivo hemolysis presents other challenges as detection often requires either a patient’s history of a hemolytic process or one or more prior hemolyzed samples from the patient.
Reporting results on hemolyzed samples . Usually, the laboratory does not report any results that are significantly affected by in vitro hemolysis and requests that another specimen be collected, preferably by another person collecting the blood. The more difficult issue is when a suitable specimen cannot be recollected and the physician calls the laboratory with an urgent need for the result, which may provide some assurance in difficult clinical situations. Legal wisdom says to not report the results under any circumstances, while accommodating medical urgency says to report the result as a comment with disclaimers such as noting the result is significantly affected by hemolysis and is provided only at the request of the physician and after discussion with the laboratory director. An approach that addresses both concerns is to require one or two additional redraws, then if both specimens are still unacceptable, the laboratory provides the results if the physician requests them.
Proper collection and handling of samples
Blood collected for blood gas analysis is susceptible to changes, especially to p O 2 . Anaerobic conditions during collection and handling are essential because room air has a p CO 2 of nearly 0 and a p O 2 of ∼150 mmHg. The factors that must be controlled are:
Removal of all air bubbles
Use of the proper anticoagulant
Appropriate use of plastic syringes (glass syringes rarely used anymore)
The temperature of storage before analysis
The length of delay between collection and analysis of blood
Any agitation of the specimen
The complete removal of all air bubbles is especially important before sending blood in a syringe by pneumatic transport, which will agitate the sample and markedly affect p O 2 , with p CO 2 much less affected by air bubbles ( , ) .
The effect of liquid heparin at <10% (vol./vol.) of the volume of blood has variable effects on blood gas and other analytes ( , ) . There appears to be little effect on pH and a relatively small effect (∼3%) on p O 2 , although the p O 2 also appears to be affected by the p O 2 in the heparin diluent ( , ) . p CO 2 , bicarbonate, and base excess are affected proportionately by dilution, about a 10% decrease. As expected, liquid heparin will dilute other constituents in blood, such as electrolytes, lactate, and glucose, which are often analyzed simultaneously in many current blood gas and electrolyte analyzers ( ) . Therefore, only dry heparin should be used as an anticoagulant.
Although plastic syringes are used for nearly all blood gas measurements, they have a potential disadvantage because they are permeable to oxygen and absorb oxygen in polyethylene ( ) . When stored in ice, because of the increased affinity of Hb for O 2 at cold temperatures, blood can absorb oxygen within the wall of the syringe that has diffused through the plastic. This effect is most pronounced in samples with a p O 2 of ∼100 mmHg and above; that is, when Hb is already nearly fully saturated with oxygen and is unable to buffer any added O 2 . A p O 2 of 100 mmHg may increase by 8 mmHg during 30 min of storage on ice. When Hb is less saturated (e.g., at a p O 2 of 60 mmHg), it is better able to buffer the additional oxygen, causing a relatively small change in p O 2 .
Collection and transport of blood for oxygen measurements
As the p O 2 in the blood is easily changed, there are several important preanalytical cautions necessary to prevent errors in clinical measurements of oxygen and cooximetry results.
Delay in sample processing . In general, storage of blood in plastic syringes at room temperature is acceptable if the analysis is within 15 min, with average changes in pH of <0.01 unit, p CO 2 of less than 1 mmHg, and p O 2 of less than 2 mmHg ( ) . For room temperature storage up to 60 min, the rates of change for pH and p CO 2 are small. However, for p O 2 , storage at ambient temperature for >30 min decreased p O 2 by an average of about 5 mmHg, with a wide variation of changes ( , ) . In a study of samples with original p O 2 values of 50 and 250 mmHg, storage for 15 min at ambient temperature (22–24°C; 72–75°F) decreases p O 2 by an average of ∼5 mmHg after 15 min, and by ∼8 mmHg after 30 min ( ) .
Samples from patients with extremely high leukocyte and/or platelet counts must be analyzed as soon as possible because pH and p CO 2 can change to some degree, while p O 2 and glucose and lactate can change dramatically in samples stored at room temperature ( ) . Even storage on ice did not prevent significant changes in p O 2 in such samples. For such samples, point-of-care testing using either blood gas analyzers or pulse oximeters offer blood gas and oximetry results that are much less affected by cell metabolism.
Storage time of specimens after collection: Room temperature versus icing . Although storage of blood at room temperature can lower p O 2 , pH, and glucose, and increase p CO 2 and lactate, prolonged storage of plastic syringes in ice can also increase p O 2 . Because polyethylene is slowly permeable to oxygen (polypropylene is the membrane used in p O 2 electrodes) and because cold significantly increases the affinity of Hb for oxygen, prolonged storage of blood in plastic syringes can cause a slow diffusion of O 2 into the blood within the syringe. More of this oxygen binds to the Hb at the cold temperatures, then is released when the sample is analyzed at 37°C ( ) . As an example, a p O 2 of 100 mmHg may increase by 8 mmHg if stored on ice for 30 min. As a general guide, if a specimen will be analyzed in 30 min or less, do not ice the specimen. If analysis will be delayed longer than 30 min, the syringe should be placed in ice. Because of increased leakage from cells, storage on ice can also increase the potassium concentration in the stored blood ( ) .
Insufficient line waste draw . Arterial and venous catheters must be adequately flushed before drawing the blood specimen for blood gas testing. As a general rule, the waste draw volume should be 2-times the catheter dead volume ( ) .
Inadequate mixing of heparinized blood . There are at least two important reasons to properly mix blood specimens after they are collected in a syringe: to dissolve and mix the heparin anticoagulant, and to maintain an even distribution of the blood cells.
There are several different types of dry preparations of heparin used in syringes that may have distinctly different mixing requirements. If heparin is not distributed and dissolved rapidly, clots and/or microclots may form in the specimen.
Erythrocytes sediment rapidly in undisturbed specimens. If erythrocytes are not evenly distributed by mixing, significant errors can occur in the hemoglobin measurements. As a rule, specimens that have sat for over 5 min may require 2 min of mixing ( ) . Note that sedimentation is even higher in inflammatory conditions, such as infection or autoimmune disease, as this is the basis of the erythrocyte sedimentation rate (Sed Rate) test.
Dilution by liquid heparin . While the use of liquid heparin in syringes has been almost totally replaced by dry preparations of heparin, if any liquid heparin remains in a syringe before blood collection, it will dilute out the analytes to be tested, especially hemoglobin, electrolytes, glucose, lactate, and p CO 2 .
Trapped air in syringes . Trapped air bubbles in syringes can significantly increase or lower the p O 2 , %O 2 Hb and s O 2 , and O 2 content of the blood. Because atmospheric air has a p O 2 of approximately 150 mmHg, exposure to air will increase the p O 2 of blood from patients breathing room air. However, for patients on O 2 -enriched air, the p O 2 of blood can rapidly decrease upon exposure. The variables on this effect are as follows:
The volume of air trapped in the syringe.
The number and size of the air bubbles.
Any agitation of the specimen after collection, especially from pneumatic transport.
The original p O 2 of the specimen.
The original Hb concentration and %O 2 Hb of the specimen.
Pneumatic transport of specimens . As noted earlier, because pneumatic transport of blood gas specimens can agitate blood specimens, any air bubbles trapped in the syringe will be equilibrated with the blood and cause significant changes in the p O 2 result ( , ) . It is especially important for those collecting blood gas specimens to remove all air bubbles from the syringe before sending by pneumatic transport. As noted, the p O 2 result will be increased in samples from patients breathing atmospheric air at p O 2 ∼150 mmHg. However, for blood specimens from patients on supplemental oxygen with a p O 2 such as 300 mmHg, the p O 2 result will be decreased ( ) .
Collection of capillary samples . Although properly arterialized capillary blood can yield pH and p CO 2 results that are close to arterial results, the agreement of p O 2 results between capillary and arterial blood is more variable ( , ) . p O 2 results from capillary blood will always be lower than arterial blood, possibly only slightly lower, so that a normal or low-normal capillary p O 2 assures that the patient is not hypoxemic. The puncture site (finger or heel) should be prewarmed up to 42°C to increase blood flow several fold. After puncture, the blood should be allowed to flow freely, without “milking,” which can introduce venous blood and interstitial fluid. The capillary tube should be filled completely without air bubbles and properly sealed ( ) .
CLSI has recently published a document on collecting capillary blood specimens ( ) . It describes optimal procedures in the collection process for capillary blood, such as how to approach, greet, and identify the patient, and how to select the proper collection site including which sites to avoid, such as sites of infection or inflammation, fingers of newborns, and patients with either edema or severe dehydration. Proper positioning of the patient is described with collection techniques, along with labeling and transporting the specimen. Analytical variation between capillary and venous specimens is discussed, such as noting that total protein, calcium, bilirubin, sodium, and chloride are lower by 5% or more in capillary blood compared to venous blood. There are several topics covered related to quality assurance, such as safety, equipment management, handling unexpected events, and continual improvement ( ) .
Collection of umbilical cord (UC) blood specimens . There are two general techniques for collecting UC blood specimens. In Method 1, the cord blood is collected while the UC is still attached, and within 1 min after delivery of the neonate and before the placenta is released. This has the advantage of maximizing UC blood returning to the neonate.
In Method 2, a segment of the UC is clamped off within 1 min after delivery of the neonate. After clamping, blood may be collected from the clamped segment of the UC.
Blood from the UC artery should be collected first, then blood from the UC vein. The UC artery has tougher walls and is a bit more challenging to collect ( ) . Because of this, the needle should be inserted into the UC artery at an angle (see Fig. 10.1 ) to avoid accidently collecting venous blood instead of arterial blood. Collecting from the correct UC artery or vein is usually not a problem unless the blood volume in the UC is low ( ) .