W H I TE PA P E R

Addressing common pre-analytical errors associated with

arterial blood gas analysis using the Proxima System
Dr Jess Fox & Dr Gavin Troughton, Sphere Medical Limited

Up to 37% of pre-analytical
Up to 60% of all errors in errors in laboratory STAT (short
blood gas testing occur in the turnaround time) testing
pre‑analytical phase are caused by provision of a
deficient sample

The Proxima System revolutionises blood gas analysis sampling

Sources of pre-analytical error The Proxima System

Immediate sample analysis

Haemolysed samples
and no sample mixing required

Errors associated with anti‑coagulation Anti-coagulant not required

Repeatable high integrity

Low integrity blood samples blood samples present in the
Sensor for analysis

Arterial line tube fitted with

Sample storage and transport captive syringe, no further
manipulation required

Immediate sample analysis

Sample sedimentation
precludes sedimentation

Results displayed at
Errors in patient identification
patient’s bedside

Blood sample kept in a closed

Sample contamination
air-free system at all times

The Proxima System mitigates:

Patient misdiagnosis

Incorrect patient treatment

Delays in diagnosis

Need for resampling

Addressing common pre-analytical errors associated with
arterial blood gas analysis using the Proxima System
Dr Jess Fox & Dr Gavin Troughton, Sphere Medical Limited

Introduction
Common pre-analytical errors
Arterial blood gas analysis is common in critical care
settings, where the results almost always have the Haemolysed samples
potential to dictate an immediate or urgent response. One of the most frequent pre-analytical errors to occur
The pre-analytical phase of analysing a clinical sample during the process of sample handling is haemolysis9, the
refers to all of the activities that occur prior to the sample’s result of which is the release of intracellular components
insertion into the analytical instrument. These steps must into the plasma. Haemolysis can result when samples are
be carefully performed to ensure that the patient receives handled vigorously e.g. high sample fill rate, vigorous
an accurate test result and consequently appropriate and sample mixing or dropping the sample container on the
timely treatment. floor.
Many of the pre-analytical steps in arterial blood gas Sample haemolysis most affects the potassium
analysis are analogous to other laboratory tests, e.g. ion measurement accuracy as the concentration is
accurate sample labelling. However, some pre-analytical approximately 30 times greater in the intracellular
steps and potential sources of error are unique to blood compartment than in the plasma phase10. However,
gas analysis due to the physicochemical properties of sodium, calcium and haemoglobin (Hgb) measurements
the analytes being measured. This document explores can be affected, as summarised in table 1.
how and why the pre-analytical errors occur along with
the impact of the error on the analyte levels in the blood Analyte 0% 0.5% 5% 10%
sample. We also consider the Proxima System, an on- haemolysis haemolysis haemolysis haemolysis
demand arterial blood gas analyser that is designed to K+ (mM) 4 4.5 7.0 10.0
address many of the errors that can occur in the pre-
analytical phase. Na+ (mM) 140 140 136 133

Ca2+ (mM) 1.2 1.2 1.1 1.0

Incidence of pre-analytical errors Hgb (g/dl) N/A 0.08 0.8 1.5
Pre-analytical processes tend to be more error prone
than those later in the testing process as the specimen Table 1: The consequence of haemolysis on the observed concentrations
of K+, Na+, Ca2+ and Hgb9
collection is a manual procedure, while analytical
and post-analytical phases are often automated and
consequently subject to reliable computer checks. Errors associated with anti-coagulation
Pre-analytical errors in clinical chemistry occur in around Blood samples analysed on standard blood gas analysers
1-1.5% of all samples analysed1,2. They can result in require treatment with an anti-coagulant to avoid sample
patient misdiagnosis and incorrect patient treatment or a clotting, which can block the sample pathway of the blood
need for resampling3. Up to 60% of all errors in blood gas gas analyser and affect the current and future samples11.
testing occur in the pre-analytical phase4. The incidence The most common anti-coagulation errors are:
of commonly reported pre-analytical errors is shown in • Delayed or inadequate mixing: This can result in
figure 1. clotting of the sample. The associated processes will
release potassium ions and can positively bias the
potassium results4.
60%

• Wrong anti-coagulant: Blood gas syringes use a dry

balanced heparin to minimise the impact on the sample.
50%

40%
Liquid heparin causes negative bias to all parameters by
Incidence (%)

30% dilution and by binding the positive electrolytes4. The

20%
dilution effect varies depending on the volume of liquid
heparin added vs the volume of blood gas sample, e.g.
10%
0.05 ml liquid heparin added to a 1 ml blood sample
0%
(Hct 45%) will dilute the plasma phase by approx. 10%9.
Other anti-coagulants, such as EDTA, can change the pH
ge
is

st

or
y

er
nc
io

tio

io
ys

ue

rr

h
ra
at

at

Ot
ie
ol

re
ca

q
o
ul

in
fic
em

re
St

ifi

de
am
ag

De

ng
nt

of the sample as it is a weak acid.

Ha

Or
co

nt
e

si

Bonini et al. 2002

Id
ti-

is

Co
M
An

Carraro et al. 2007

Pre-analytical error Carraro et al. 2012
Grecu et al. 2014
Deficient samples
Up to 37% of pre-analytical errors in laboratory STAT
Figure 1: Incidence of pre-analytical errors in STAT laboratories reported in
(short turnaround time) testing are caused by provision
four studies:
of a deficient sample; this could be because the sample
A one year study by Bonini et al.5; a three month study by Carraro et al.6; a six
tube is empty, missing or inadequately filled or because
month study by Carraro et al.7 and a one year study by Grecu et al.8
the test request is missing6. Such errors require additional tend towards atmospheric concentrations, i.e. there
samples to be drawn and cause delays in diagnosis. will be a decreased ability to detect profound hypoxia
or hyperoxia. pCO2 within the sample will fall (since
Sample storage and transport atmosphere pCO2 is very low), but the effect is less
A blood gas sample which has been correctly collected marked than the change in pO2. As a direct consequence
and capped will continuously change during any delay of the decrease in the pCO2, pH will increase when
between collection and analysis e.g. due to transport arterial blood is exposed to air.
from patient to analyser or to availability of immediate •
Sampling container effects: The gold standard
analysis. container for collection of arterial blood gas samples
• Metabolic effects: Blood is a living tissue and is a glass syringe as the low rate of diffusion of gases
metabolism continues within the sample after the though the material preserves the oxygen level in
blood is drawn. As a result oxygen, carbon dioxide, pH, a blood sample17. However, for reasons of cost and
glucose and lactate levels are all affected. practicality plastic syringes are widely used for drawing
• Chilling: If a prolonged (>30 min) time delay is blood gas samples. The consequence of gas diffusion
anticipated before the sample is analysed, the from the plastic sampling container will be greatest if a
use of glass syringes and storage in ice water is small sample is drawn to support blood conservation.
recommended12. This slows down the metabolic effects,
but can result in haemolysis and changing potassium The Proxima System and pre-analytical errors
levels. Furthermore, if a plastic syringe is used this
Proxima is a patient dedicated arterial blood gas
can have a profound effect on the oxygen levels due to
analyser that allows critical care staff to obtain blood gas
transport through the syringe.
measurements without leaving the patient’s bedside –
• Sample sedimentation: Red blood cells separate from refer to figure 2.
the plasma during storage and a heterogeneous sample
The Proxima Sensor is integrated into the patient’s
will result if stored for even a short period of time.
arterial line operating as a closed system to minimise
The effect of a heterogeneous sample on the bias will
blood handling and infection risk. During the sample
depend on which portion of the sample is measured.
measurement procedure, the user draws blood into the
Apparently homogeneous samples which have been
Proxima Sensor using a closed sampler syringe. Once the
briefly mixed after just 10 minutes storage have been
sample has been analysed, all blood is returned to the
reported to under read by 28%4.
patient. Results are displayed on the bedside monitor.
• Transport: Many hospitals use pneumatic tube systems
to minimise the transfer time of samples from the
patient to the laboratory. This can have a profound
effect on the oxygen concentrations if there are any
bubbles present in the sample13.

Errors in patient identification

Incorrect or missing patient and/or sample identification
are common pre-analytical errors that occur in blood gas
testing14. Errors can be caused by3:
• Lack of patient identification and/or sample labelling
• Transcription errors due to manual data entry
• Lack of a dedicated procedure for identifying patient
samples
Missing/incorrect information causes delays in sample
analysis and can consequently lead to erroneous results,
Figure 2: The Proxima System designed for reliable, on-demand bedside
diagnosis and inappropriate treatment.
blood gas analysis

Sample contamination
The blood gas sample can become contaminated during The Proxima System has been designed to address many
the collection process: of the common pre-analytic and pre-analytical errors to
• Contamination from flush: When sampling from an deliver more reliable blood gas measurements. Table 2
arterial line it is important to remove an adequate presents the design features of the Proxima System and
‘discard volume’ to ensure the sample obtained for how common pre-analytical errors are addressed within
analysis is 100% blood. If an insufficient discard volume this new medical device.
is removed, the sample will be contaminated with flush
solution and the results affected.
• Contamination with air: Any air bubble in an arterial
blood can alter the pO2, pCO2 and pH of the sample
so that it no longer represents the patient’s status15.
Air bubbles with a relative volume of 0.5-1.0% of the
sample may cause significant errors16. Oxygen levels will
 Sources of
The Proxima System is designed to address pre-analytical errors
pre-analytical error

Haemolysed samples Sample analysis with the Proxima System is carried out immediately, avoiding risk factors for
sample haemolysis such as storage on ice or vigorous mixing of the sample.

Errors associated with Blood samples analysed on Proxima do not require treatment with an anti-coagulant, thereby
anti-coagulation avoiding poor incorporation, the risk of dilution or inappropriate reagent choice.

Deficient samples The Proxima System is designed so the user withdraws a fixed volume of blood which ensures a
high integrity arterial blood sample is present in the Sensor for analysis.
Since Proxima analyses the samples in situ, samples cannot be lost or test requests missed.

Sample storage and The Proxima System is incorporated into an arterial line tube fitted with a captive syringe. When
transport an analysis is performed, blood is drawn into the captive syringe until it is completely filled.
The analysis then occurs without further manipulation of the sample which ensures errors
associated with sample container choice are avoided.
The Proxima System analyses samples within a few tens of seconds, during which there is no
significant metabolism effect.

Sample sedimentation The Proxima System measures the blood sample immediately after flowing through the Sensor
and within a few tens of seconds, precluding the risk of sample sedimentation.

Errors in patient The Proxima System is a patient dedicated in-line blood gas analyser displaying results directly
identification at the patient bedside. This eliminates accidental misidentification or exchange of samples.
The patient’s ID is captured in the Proxima System at the start of the episode of care and is
automatically saved with all electronic records.

Sample contamination The Proxima System is designed to accommodate a suitable discard volume within the line
reservoir to ensure that a high integrity arterial blood sample is present in the Sensor for
analysis. By returning the contents of the reservoir to the patient there is no reason to minimise
the discard volume and risk a sample contaminated with flush.
The Proxima System keeps the blood sample in a closed, air-free system at all times, thereby
allowing the user to analyse blood samples at the patient’s bedside removing the impact of
transport on sample quality.

Table 2: Design features of the Proxima System that address common pre-analytical errors

References
1 Chawla R, Goswami B, Tayal D, et al. Identification of the types of pre- 11 Higgins C, The use of heparin in preparing samples for blood gas
analytical errors in the clinical chemistry laboratory: 1-year study at analysis, Apr 2007.
G.B. Pant Hospital. Lab Medicine. 2010;41:89–92. http://acutecaretesting.org/en/articles/the-use-of-heparin-in-
2 Upreti S, Bansal R, Jeelani N, Bharat V, Types and Frequency of Pre- preparing-samples-for-blood-gas-analysis
analytical Errors in Haematology Lab, Journal of Clinical and Diagnostic 12 Procedures for the collection of arterial blood specimens; approved
Research. 2013 Nov, Vol-7(11): 2491-2493. standard 4th edition. NCCLS, H11-A4, Vol. 24, No. 28.
3 Radiometer website: 13 Higgins C, Pneumatic tube transport of samples for blood gas analysis,
http://www.avoidpreanalyticalerrors.com/#en/handbook/ Jan 2005.
arterial+line/patient+&+sample+id http://acutecaretesting.org/en/articles/pneumatic-tube-transport-
4 Wennecke G, Juel G, Avoiding pre-analytical errors – in blood gas of-samples-for-blood-gas-analysis
testing, 2008, ISBN87-88138-94-1. 14 Kahn SE, Specimen mislabeling: A significant and costly cause of
5 Bonini P, Plebani M, Certiotti F, Rubboli F. Errors in Laboratory Medicine. potentially serious medical errors, Apr 2005.
2002; 48(5):691-98. http://acutecaretesting.org/en/articles/specimen-mislabeling-a-
6 Carraro P, Plebani M. Errors in a STAT laboratory: types and frequency 10 significant-and-costly-cause-of-potentially-serious-medical-errors
years later. Clin Chem 2007; 53: 1338-42. 15 Toffaletti J, Effect of small air bubbles on changes in blood pO2 and
7 Carraro P, Zago T, Plebani M, Exploring the initial steps of the testing blood gas parameters: calculated vs. measured effects, Jul 2012.
process: frequency and nature of pre-preanalytic errors. Clin Chem http://acutecaretesting.org/en/articles/effect-of-small-air-bubbles-
2012;58:638-42. on-changes-in-blood-po2-and-blood-gas-parameters
8 Grecu DS, Vlad DC, Dumitrascu V, Quality indicators in the preanalytical 16 Mueller RG, Lang GE, Beam JM, Bubbles in samples for blood gas
phase of testing in a Stat laboratory, Lab Medicine. 2014; 45(1):74-80. determinations – a potential source of error. Am J Clin Pathol 1976; 65:
9 Wennecke G, Useful tips to avoid preanalytical errors in blood gas 242-49.
testing: electrolytes, Oct 2003. 17 Wiwanitkit V, Glass syringes are better than plastic for preserving
http://acutecaretesting.org/en/articles/useful-tips-to-avoid- arterial blood gas for oxygen partial pressure determination: an
preanalytical-errors-in-blood-gas-testing-electrolytes explanation based on nanomaterial composition, Int J Nanomedicine
10 Radiometer website: 2006:1(2) 223–224.

http://www.avoidpreanalyticalerrors.com/#en/handbook/
arterial+line/hemolysis