REVIEW

Arterial lines: an analysis

of good practice
SUMMARY
▪ The of arterial line monitoring in
use
Sharon Lee Clarke RGN, RN (Child Branch),
Staff Nurse.
critically ill children provides an Address for correspondence:
indispensable foundation of critical care. Burns Unit, City Hospital, Hucknall Road,
▪ The information provided by their use NG5 1PB
Nottingham
ensures accurate monitoring and blood gas
analysis and subsequent modifications to raised intra-cranial pressure. The current treatment for head
prescribed treatment.
injuries in use involved cooling the body to a temperature
▪ The invasive nature of intra-arterial of 32°C, thereby resulting in a state of hyperventilation
catheters may lead to complications of which maintains pC02 levels between 3.5 and 4 Kpa; re-
haemorrhage, thrombosis and infection. ducing alveolar carbon dioxide levels causing vasoconstric-
▪ The value of arterial cannulation in relation tion and a subsequent decrease in cerebral blood flow
to alternative less invasive methods of (Carter, 1993). Fluid input was restricted at two-thirds of
monitoring is discussed. normal requirements for five days, a practice that is said to
▪ The literature reveals that in many areas reduce the overall circulating volume of the body and cer-
nurses are failing to unite theory and
ebral blood volume and lower intra-cranial pressure

practice. This has implications for the safety (Carter, 1993).

of children and the accountability of nurses
INTRODUCTION
caring for them. The purpose of this paper is to analyse the nursing inter-
vention in relation to a particular aspect of technical care
KEY WORDS: Arterial cannulation, Blood gas and focuses specifically on illustration of the nurses role in
analysis, Critically ill child, Haemodynamic the care of an indwelling arterial catheter. Rationale un-
monitoring, Management of care, Professional derpinning current nursing practice will be discussed with
accountability. reference to evidence based literature and hospital proto-
col.
Consideration will also be given to the role of non-in-
CASE HISTORY vasive monitoring techniques in the critical care situation.
The nursing care outlined which follows will be related to The article will then further explore the importance of im-
the management and treatment of Jo, aged six. Her name proving professional knowledge to cover areas of legal ac-
has been changed to conceal her identity (UKCC, 1996). countability and a demonstration of a high standard of
Her treatment regime was planned as a direct result of ar- evidence based care. Finally, conclusions will be made in
terial catheter monitoring. relation to the way future paediatric nursing strategies may
Jo, aged six years, was admitted onto a critical care be improved in the best interests and safety of the child.
unit from accident and emergency following a road traffic
accident in which she sustained a head injury and devel- RATIONALE OF NURSING CARE
oped cerebral oedema. Initially, it was expected that she The nursing responsibility for the ongoing care of a child
would require only overnight ventilation. An intra-cranial with an indwelling arterial catheter appears to relate prin-
pressure bolt was inserted for purposes of monitoring, and cipally to the reduction and prevention of the complica-
an arterial line was inserted into her right radial artery as tions associated with its use. This responsibility therefore
part of her care. begins with the nurse’s own knowledge and an under-
During her first night on the unit Jo’s condition be- standing of the implications of the care to be delivered
came unstable when she developed hypertension and (Scales, 1996). All care given should be based upon

23
recommendations set down by hospital protocol and (Morton, 1997). The policy in use (Nottingham Health
underpinned by evidence wherever possible. Authority, 1986) suggests that the bag of fluid and giving
Jo’s condition had been assessed by the Consultant in set should be changed every forty-eight hours as heparin
charge of her care as critical. He requested the immediate becomes unstable after this time at room temperature
insertion of an arterial line for the purpose of continuously (Raffin, 1986). A transducer was connected to the giving
monitoring her changing blood pressure and for the ease set and this transforms the pressure from the catheter into
of obtaining arterial blood for frequent gas analysis (Sellden an electrical signal which is then transmitted to the moni-

et al.,1987). This treatment is supported by Oh (1990) who tor as a waveform (Hazinski, 1992). Thelan et al., (1994)
suggests that in critically ill patients, normal physiological suggest that in order for the nurse to successfully interpret
control mechanisms are severely changed, therefore con- and analyse physiological data and information given via
stant monitoring is required. the monitor, knowledge and understanding is required of
The arterial line was inserted into Jo’s right radial ar- the normal age-related readings.
tery. Sprung and Grenvik (1985) indicate this to be the saf-
est site of choice in children. The rationale for this is that it Blood Sampling via the Arterial Line
is close to the surface and is therefore easy to palpate and As Jo was ventilated it was important to monitor her blood
may be closely observed (Millam, 1988). The cannula was gas levels. Arterial blood gas sampling is an important part
covered only with a semi-permeable transparent dressing of this monitoring process, as it can help determine the ef-
in order to secure anchorage of the line, enable visibility of ficiency with which the lungs oxygenate the blood (Carter,
the cannula and avoid transport of bacteria into the can- 1995). It is suggested by Preusser et al., (1989) that in the
nula site (Pearson, 1996a,b). care of a ventilated patient, blood gas analysis serves as a

In line with current practice, Jo’s cannula was observed basis for management decisions concerning ventilation and
for any signs of disconnection, dislodgement, leakage and clinical treatment in critical care settings. The role of the
redness and swelling indicative of infection. Wilson (1995) paediatric nurse in this intervention involves obtaining
suggests that arterial cannulas are associated with a much appropriate blood samples and accurately understanding
higher rate of infection than intra-venous peripheral lines. and interpreting the results which are obtained within min-
Jo’s hand and fingers were also observed each hour in or- utes through the use of a gas sampling machine. Wearing
der to ensure adequate perfusion of blood was taking place unsterile gloves, approximately 2 millilitres (mls) of blood
as an arterial cannula is believed to impede the flow of would first be removed by the nurse in order to clear the
blood to the peripheries and increase the risk of tissue is- line of heparinised blood so that a non-contaminated sam-
chaemia (Allan, 1989). Zimmerman and Gildea (1985) also ple may be obtained (Hazinski, 1992). This volume of blood
suggest that attention should be paid to colour, warmth, would then be discarded, a practice said to avoid dilution
pain and motor and sensory nerve function. This process of the sample by the flushing solution (Alford et al., 1993).
was problematic as Jo was ventilated, sedated, paralysed A further 0.5mls would then be removed in a second sy-
and furthermore cooled as part of the local head injury ringe to use for gas sampling. The system would then be
protocol, and was therefore vasoconstricted. Her skin was flushed using the giving set flush device, to clear the line
pale and cold and she was unable to respond to outside of blood. This procedure is set out in the local policy (Not-
stimuli. It was therefore necessary to monitor for the pres- tingham Health Authority, 1987). As a result of obtaining
ence of a distal pulse and capillary refill. The transparent this information so promptly it was then possible for swift
dressing was not covered by a secondary dressing and Jo’s decisions to be made by the medical and nursing team con-
right arm was left exposed on top of the bed clothes at all cerning the management of Jo’s respiratory support.
times so that the line was always visible and any acciden-
tal disconnection of the line may be detected immediately. Associated Problems
This action ensured that any major arterial haemorrhage Jo’s radial cannula began to cause blanching of her fingers
was prevented (Hudak et al., 1997). The line was clearly on flushing and a dampened waveform after a period of
labelled and colour coded with red tape for clear identifi- two days indicating a possible occlusion at the catheter tip
cation in order to ensure that no inadvertent intra-arterial (Hazinski, 1992). Therefore, the catheter was re-sited into
injection of drugs could occur (Jones, 1995). her left radial artery. This line remained satisfactory for a
further three days during which time Jo’s body was
The Arterial Line warmed to a normal temperature and she no longer re-
The arterial line was connected via a giving set with a flush quired full ventilatory support. Once Jo’s condition had
mechanism to a 500ml bag of solution containing 0.9% sa- stabilised the arterial line was removed and she was trans-
line solution and 500 units of heparin surrounded by a pres- ferred to a ward. Blood and catheter cultures indicated that
sure bag. This was set at 2 mmHg and delivered into the she had suffered no systemic infection as a result of the
artery 1 unit of heparin per ml of fluid, at a rate of 2ml / invasive line.
hour. This maintains the infusion pressure at a greater pres-
sure than the vascular pressure being monitored (Hazinski, ANALYSIS OF NURSING CARE
1992). This system is believed to prevent the coagulation Analysis of the current nursing actions in the care of a child
of blood within the arterial catheter and prevents back-flow with an indwelling arterial line concentrates on the use of
of blood into the monitoring system, therefore avoiding heparin in the flush mechanism and safety issues relating
thrombus formation and prolonging the catheter life to infection and tissue ischaemia. It appears that many

24
practices are not evidence-based (Hazinski, 1992) and this often insufficient to prevent the spread of infection
raises questions of professional ethics and legal account- (Gorman et al., 1993). As such infections may become life-
ability (Pearson, 1996a,b). threatening in the seriously ill, it could therefore be con-
Many studies have been carried out into the use of cluded that sterility should be maintained by nurses when
heparin as part of the flush mechanism. In support of cur- handling arterial catheters and ports (Hudak et al., 1997).
rent practice, Butt et al., (1987) demonstrate how the use of Furthermore, Pearson (1996a,b) agrees that an aseptic
heparin in arterial infusions inhibits fibrin clot formation technique is appropriate whenever handling of the cath-
at the tip of the cannula, thereby maintaining its patency. eter hub is required. However, many guidelines for blood
However, it is important to note that the use of heparin in sampling from an arterial line do not recognise the neces-
the flush solution has been found to lead to numerous prob- sity of an aseptic procedure and do not appear to be based
lems. Preusser et al., (1989) suggests that heparin leads to on firm evidence. Larson (1994) suggests that infectious
contamination of blood gas samples and this directly re- complications of intra-arterial catheters may be reduced
sults in the current practice of withdrawing a discarded by thorough care of the site.
sample to remove the heparinised solution prior to obtain- Elliott and Faroqui (1992) illustrate that microbial colo-
ing blood samples. In children who require ventilatory nisation of intravascular catheters may be caused by mul-
support, frequent arterial blood gas sampling results in the tiple factors. However, Pearson (1996a,b) and Morton (1997)
cumulative loss of large volumes of blood which they can agree that the risk of infection relates directly to the dura-
not afford to lose and this may lead to anaemia and the tion of cannulation and their studies indicate that a cath-
need for a blood transfusion. eter in place for forty-eight hours or less carries a 0% risk
At the time the study was undertaken, the protocol of bacterial colonisation.
suggested that for children younger than two years, the
discard volume should be replaced following the removal Jo’s catheter remained patent for two days, at which
of the blood gas sample, a practice designed to minimise time blanching of her fingers was noted on flushing and
blood loss (Nottingham Health Authority, 1987). However, the monitor waveform became damp. This phenomenon
evidence has revealed that this process may dislodge small results from clots or air bubbles being flushed from the
clots that form around the catheter tip (Sprung and catheter site to the digits, compromising arterial circula-
Grenvik, 1985). Studies by Butt et al., (1985) further sug- tion and indicates poor ulnar collateral flow (Sprung and
gest that micro bubbles and fibrin platelet aggregate are Grenvik, 1985). This sign should result in the immediate
carried during flushing into the aortic arch and eventually removal of the line due to an increased risk of tissue is-
to the brain. Pressure exerted by volumes larger than 0.5ml chaemia (Zimmerman and Gildea, 1985). However, Pearson
may result in consistent elevation of intra-cranial pressure (1996a,b) states that there is no consensus of opinion on
and potential ischaemic injury. Cicala et al., (1988) suggest the maximum complication-free duration of arterial can-
that, despite the presence of protocols to guide practice, nulation and suggests that the potential benefits of re-sit-
techniques aimed at obtaining an accurate sample are of- ing a cannula routinely should be assessed on an individual
ten performed incorrectly in clinical settings and such ac- patient basis. It is also important to note that, although
tion may be potentially fatal in a child with existing cer- regular checks were made to monitor adequate perfusion
ebral oedema and raised intra-cranial pressure. of blood to Jo’s fingers whilst her arterial line was insitu,
In order to minimise this risk or subsequent unneces- vasoconstriction may result in peripheral cyanosis and
sary blood loss caused by disposal of the discard volume, therefore the accuracy of capillary refill in a vasoconstricted
Hazinski (1992) suggests the disposal of 0.1- 0.5ml of blood child is questionable (Carter, 1995).
initially, depending on the length of the line, to remove
old blood and small emboli, and further removal of 2mls NON-INVASIVE MONITORING
to clear the line of heparin. The 2ml may then be re-infused The potential danger of complications relating to the use
through a central line in order to avoid potential arterial of arterial lines leads health professionals to question the
thrombo-embolitic complications. Gamby and Bennett necessity for their use. Oh (1990) suggests that the most
(1995) however, question the necessity for heparin to be effective monitoring systems for use in paediatrics are those
added to flush solutions at all. Their studies suggest that which are non-invasive and carry minimum risk for the
catheter patency is maintained adequately enough by the child. Furthermore, Zimmerman and Gildea (1985) suggest
flush under pressure rather than the heparinisation of the that intra-arterial monitoring devices do not always ensure
flush. Pearson (1996 a,b), further supports the practice of accuracy of blood pressure measurements. Similarly, Park
using only 0.9% saline flush solution to maintain catheter and Menard (1987) suggest, as a result of their study, that
patency and adds that the presence of heparin may in fact oscillometric dinamap blood pressures obtained in the
enhance the growth of certain bacteria and lead to further upper arm accurately reflect radial arterial pressures in
haemorrhagic complications. infants and children. Thelan et al., (1994) however, favour
It is suggested by Band and Maki (1979) that 12% of all intra-arterial monitoring for assessment of arterial
septicaemias originate from arterial lines. Wilson (1995) perfusion to the major organ systems of the body and evalu-
states that infections associated with intra-vascular devices ation of cardiac function, circulating blood volume and
commonly arise from microorganisms colonising the skin, physiological response to treatment.
usually originating from staff who handle the device. There On occasions, Jo’s blood pressure gave cause for con-
is evidence to suggest that handwashing techniques are cern.Whilst elevated prior to her treatment on the head

25
injury protocol, it occasionally decreased lower than the - 1: Normal Arterial Bloodo Gas Values in
&dquo;
w w a
acceptable reading for her age at rest yet would regularly Child (from Hazinski, ....
increase considerably during physiotherapy and nursing
intervention, therefore regular accurate monitoring was
imperative. However, due to the vasoconstriction caused
by Jo’s cooled body, non-invasive blood pressure readings
dependant on blood flow rather than pressure, would have
proved inaccurate and unreliable (Henneman and
Henneman, 1989). Hazinski (1992) indicates that fluctua- order to monitor the physiological response of her body to
tions in arterial pressures may precede a life-threatening the regular changes. In the ventilated child, respiratory
crisis, and therefore in Jo’s situation, accurate and reliable acidosis can be seen where there is inadequate mechanical
data was imperative in order for prompt treatment to be ventilation and results in:
carried out. However, once the waveform on the monitor
became dampened, the accuracy of Jo’s arterial pressure ~
pH decreased
recordings was compromised, as communication from the ~
pC02 increased
artery to the transducer was interrupted, possibly by a clot ~ HC03 normal or increased (Carter, 1993)
or air bubble at the catheter tip (Thelan et al., 1994). This

offered further indication for immediate re-siting of the line. The paC02 indicates the efficiency with which carbon di-
It is suggested that an Allen test (Hazinski, 1992) may have oxide is being eliminated (Allan, 1989). It is suggested by
been helpful prior to selection of the radial artery in estab- Pandit (1995) that in life threatening conditions arterial
lishing adequacy of ulnar arterial blood flow (Millam 1988). blood gas pressures should be monitored. Hazinski (1992)
further supports that data obtained via an indwelling ar-
Allen Test terial catheter provides the most reliable assessment of
This is a test designed to verify the adequacy of ulnar arte- partial pressure of oxygen and carbon dioxide in the body
rial flow to the hand prior to selection of the radial artery and offers much more important insight into effective ven-
for cannulation (Hazinski, 1992). The test begins by com- tilation.
pression of the wrist over the radial and ulnar arteries. The
patient clenches their hand into a fist and the position is CONCLUSION
held for a period of five seconds. The hand is opened, re- Following a review of the literature, it would appear that
leasing the pressure over the ulnar artery. The length of intra-arterial monitoring is widely used in intensive care
time which it takes for the hand to regain its colour is noted. monitoring (Ducharme et al., 1988). Sellden et al., (1987)
If flow through the ulnar artery is adequate, the hand will believes that in experienced hands, complications are few.
become re-perfused within five seconds. However, if longer However, in respect of the issues raised and the potential
time is needed, it can be concluded that a catheter placed severity of complications associated with the use of arte-
in the radial artery may seriously compromise arterial rial lines, it is perhaps prudent to question the readiness
blood flow to the hand. with which arterial cannulation is considered routine in
the critical care setting. Techniques suggested by the lit-
Arterial blood gas analysis is carried out for ventilated erature often appear to differ quite considerably from those
and critically ill children in order to evaluate oxygen and in current use, leading to conflict between the UKCC codes
carbon dioxide gas exchange and acid-base status (Preusser, statement of values (UKCC, 1996) and procedures under-
1989) (see table 1). Capillary blood gases are suggested as pinning nursing guidelines. Scales (1996) suggests that
an alternative non-invasive method of measuring oxygen nurses have a duty to act on research, yet in doing so may
saturation. Dar et al., (1995) suggest that capillary blood incur disciplinary action caused by practising outside hos-
offers data which is as accurate as that supplied by arterial pital policy. As members of a profession, paediatric nurses
blood and favour this method for avoidance of potential have an ethical responsibility to utilise research and estab-
complications of arterial cannulation. However, Hazinski lish innovative practices in order to ensure that optimum
(1992) argues that capillary P02 (partial pressure of oxy- nursing care is delivered to children and their families.
gen) does not accurately reflect pa02 readings obtained
from arterial blood in the presence of poor systemic
perfusion of the sampled capillary bed. REFERENCES
Jo’s body was cooled and vasoconstricted, and there- Allan D. (1989) Making sense of arterial catheterisation. Nursing Times; 85
(40):
fore values obtained from capillary blood may have been 45-47.
inaccurate (Tinker and Zapol, 1992). However, Reinhardt Alford A, Kirton B, Peacock J (1993) Evaluation of an alternative method of
blood sampling from arterial cannulae. Anaesthesia; 48: 530-532.
et al., (1987) suggest that studies have found no significant
Band J, Maki D. (1979) Infections caused by arterial catheters used for haemo-
difference between samples obtained from arterial blood dynamic monitoring. The American Journal of Medicine; 67: 735-741.
and those from venous blood. Jo’s ventilation settings were Butt W, Gow R, Whyte H, Smallhorn J, Koren G (1985) Complications result-
frequently altered during her second day in the unit in re- ing from use of arterial catheters, retrograde flow and rapid elevation
in blood pressure. Paediatrics, 76
(2): 250-253.
sponse to her raised paC02, indicating that she was hypo-
Butt W, Shann F, McDonnell G, Hudson I. (1987) Effect of heparin concentra-
ventilated and subsequently acidotic. Gas results were re- tion and infusion rate on the patency of arterial catheters. Critical Care
quired at regular intervals following ventilatory change in Medicine; 15
(3): 230-232.

26
Carter B. (ed ) (1993) Manual of Paediatric Intensive Care Nursing. London: Park M, Menard S. (1987) Accuracy of blood pressure measurement by the
Chapman and Hall. dinamap monitor in infants and children Paediatrics; 79(6) 907-913.
Carter B. (1995) Nursing support and care. meeting the needs of the child Pearson M. (1996a) Guidelines for prevention of intra-vascular device-related
and family with altered respiratory function In. Carter B, Dearmun A, infections. Part One: an overview. American Journal of Infection Con-
(eds.) Child Health Care Nursing: Concepts, Theory and Practice. 277-280. trol; 24
(4) 262-293.
Oxford: Blackwell Science Ltd. Pearson M. (1996b) Guidelines for prevention of intra-vascular device-related
Cicala R, Cannon K, Larson J, Fabian T. (1988) Evaluation of coagulation stud- infections. Part Two: recommendations for the prevention of non-
ies from heparinised arterial lines with the use of lab-site high pressure socomial intra-vascular device-related infections. American Journal of
tubing. Heart and Lung; 17(6) 662-666. Infection Control; 24
(4): 277-290.
Dar K, Williams T, Aitken R, Woods K, Fletcher S. (1995) Arterial versus capil- Preusser B, Lash J, Stone K, Winningham M, Gonyons D, Nicker J. (1989) Quan-
lary sampling for analysing blood gas pressures British Medical Journal; tifying the minimum discard sample required for accurate arterial blood
(6971):24-25.
310 gases Nursing Research; 38 (5): 276-279.
Ducharme F, Gauthier M, Lacriox J, LaFleur L (1988) Incidence of infection Raffin T. (1986) Indications for arterial blood gas analysis. Annals of Internal
related to arterial catheterisation in children : a prospective study. Cnti- Medicine; 105: 390-398.
cal Care Medicine; 16(3). 272-276. Reinhardt A, Tonneson AS, Bracey A, Goodnough SK. (1987) Minimum dis-
Elliott T, Faroqui M. (1992) Infections and intravascular devices. British Jour- card volume from arterial catheters to obtain coagulation studies free
nal of Hospital Medicine; 48
(8): 496-502. of heparin effect. Heart and Lung; 16 (6 pt1): 699-705.
Gamby A, Bennett J. (1995) A feasibility study of the use of non-heparinised Scales K. (1996) Legal and professional aspects of intravenous therapy. Nurs-
0 9% sodium chloride for transduced arterial and venous lines. Intensive ing Standard; 11
(3): 41-45.
and Critical Care Nursing; 11:
148-150. Sellden H, Krister M, Larsson L, Ekstrom-Jodal B. (1987) Radial arterial cath-
Gorman L, Sanai L, Notman W. (1993) Cross infection in an intensive care unit eters in children and neonates: a prospective study Critical Care Medi-
by Klebsiella Pneumoniae from ventilator condensate Journal of Hospi- cine; (12).
15 1106-1109.
tal Infection; 23:
17-26.
Sprung C, Grenvik A (eds ) (1985) Invasive Procedures in Critical Care. Lon-
Hazinski M. (1992) Nursing Care of the Critically III Child. (2nd ed.) London: don. Churchill Livingstone.
Mosby Year Book. Thelan L, Davie J, Urden L, Lough M. (1994) Critical Care Nursing, Diagnosis
Henneman E, Henneman P. (1989) Intricacies of blood pressure measurement: and Management (2nd ed.). London: Mosby Year Book Inc.
re-examining the rituals. Heart and Lung. 18(3): 263-271. Tinker J, Zapol W. (eds.) (1992) Care of the Critically III Patient (2ed.). London:
Hudak C, Gallo B, Morton P. (1997) Critical Care Nursing. An Holistic Approach. Springer-Verlag.
(7th ed.) Philadelphia J B Lippincott Company. UK Central Council for Nursing, Midwifery and Health Visiting. (1996) Guide-
Jones N. (1995) Inadvertent intra-arterial injection of drugs, why does it still lines for Professional Practice. London: UKCC.
occur? British Journal of Intensive Care; 5 (5) 166-168 WilsonJ (1995) Infection Control in Clinical Practice. London: Baillière Tindall.
Larson E. (1994) Issues in infection control: does antiseptic make a difference Zimmerman S, Gildea J. (1985) Critical Care Paediatrics. London: WB Saunders
in intravascular device-related complications ? Heart and Lung; 23
(1): 90-
92.
Company.
Millam D. (1988) Getting into an artery. American Journal of Nursing; 88
(9):
1213-1224 Acknowledgement
The author would like to acknowledge and thank Stuart Pedley,
Morton N (ed.) (1997) Paediatric Intensive Care. Oxford: Oxford University Clinical Tutor, School of Nursing and Midwifery, Queens Medi-
Press.
cal Centre, Nottingham for his support and encouragement in
Nottingham Health Authority. (1986) Guidelines for the care of a child with preparing this article for publication.
an arterial line. Nottingham.
Nottingham Health Authority. (1987) Blood Sampling from an Arterial Line. Editor’s Note
Nottingham. This article provides a
Oh T. (ed.) (1990) Intensive Care Manual. (3rd ed.) London: Butterworths.
really wonderful example of how re-
search can inform and develop clinical practice. It is particu-
Pandit JJ. (1995) Sampling for analysing blood gas pressures. Arterial samples larly pleasing to see Sharon, a Staff Nurse, taking on responsi-
are the best. British Medical Journal; 310
(6986):1071. bility for examining this particular issue. 11m

27