Arterialbloodgases

Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

Arterial Blood Gases

Dr Russell W. Jamieson
July 2002

Aim
To produce an evidence-based guideline of best practice in the use of arterial blood
gas measurement in the A&E department

Reason
During my post I became aware that several common condition in A&E were
investigated by arterial blood gas measurements. This procedure is invasive, painful
for the patient and not without risks.

I felt that our use of arterial sampling was not evidence-based and varied from doctor
to doctor with disagreement over conditions in which venous gas sampling is
adequate, appropriate arterial sites and the use of local anaesthetic prior to sampling.

Points Considered
1. Does pulse oximetry provide a reliable screening tool for respiratory failure?
2. Can venous blood gas samples replace arterial samples in some cases?
3. Can the brachial artery be sampled safely?
4. Should local anaesthetic be used prior to sampling and which type of local
anaesthetic is effective?
5. Are capillary samples an alternative in the A&E setting?

Research Method
A Medline search was performed for papers relevant to the above topics – all papers
selected were published from 1990 onwards. Particular focus was applied to large
randomised trials in major journals.

Evidence
Pulse Oximetry As A Screening Tool

Researchers from the University of Maryland, Baltimore, USA performed


retrospective comparison of 513 arterial blood gas measurements with recorded pulse
oximetry saturations taken in the Emergency department. They established that a cut
off of 96% saturation should be used a screen for hypoxia (defined as a pO2 <9.1kPa)
or moderate hypercarbia (defined as pCO2 >6.5kPa). They then performed a
prospective study on 213 patients on which arterial samples were ordered and
confirmed that no cases of hypoxia or moderate hypercarbia would have been missed
if a cut off 96% saturation were to be used in performing arterial gases. (Reference 1)
This cut off is high and would lead to a large number of arterial samples being taken.
Type I respiratory failure is defined and as a pO2 <8kPa and thus a study looking at
pulse oximetry as a screening tool for detecting Type I respiratory failure would be
more relevant and practical.

In Norwich, England a prospective study of 89 patients admitted to a specialist


respiratory medical unit with acute severe asthma demonstrated that a saturation of
greater than or equal to 92% safely predicted the absence of respiratory failure
defined as a pO2<8kPa or a pCO2>6kPa. Arterial gases showed only 4.2% of cases of
respiratory failure would have been missed if arterial samples had only been taken
when the saturations were below 92%. The percentage missed rose to 7.3% if the cut-
off was dropped to 90%. (Reference 2)

Can Venous Sampling Replace Arterial Sampling In Some Cases?

Diabetic ketoacidosis – this is of particular interest as this patient group are often
young, normally healthy individuals, in whom respiratory pathology is less likely and
the primary reason for performing a blood gas measurement is to establish the pH.

In Oklahoma, USA there was a prospective study in which 44 episodes of diabetic


ketoacidosis were analysed by paired arterial and venous gas samples. The mean
venous pH was 0.03 lower than the arterial samples and the venous and arterial pH
had a correlation coefficient of 0.9689. A similar picture emerged for the HCO3
measurements with a mean difference of 1.8 and a correlation coefficient of 0.95. This
paper suggests that in diabetic ketoacidosis the degree of acidosis may be accurately
gauged by performing venous blood gas sampling. (Reference 3)

The Emergency Department team in Melbourne, Australia took this idea one step
further and asked if, for a range of patients admitted to the emergency department,
venous pH could safely replace arterial pH for gauging the degree of acidosis. Their
prospective study looked at 246 patients admitted to an emergency department, 196
with respiratory problems and 50 with suspected metabolic derangement. Paired
arterial and venous samples were taken with a correlation coefficient of 0.92 and 95%
confidence intervals of –0.11 to +0.04. Their conclusion was that for establishing if a
patient is acidotic venous sampling is adequate. (Reference 4)

The Melbourne team then considered the 196 respiratory cases separately and
compared the arterial and venous pH values in this group. The mean difference was -
0.034 with a good correlation between the samples. In their sample of 196, 56 patients
(29%) had significant hypercarbia defined as an arterial pCO2 >6.5. The group
showed that unfortunately venous pCO2 is not closely similar to the arterial pCO2
despite the similarity in the pH with the average venous pCO2 being 0.754 higher
than the arterial and the range being –1.14 and +2.665. However they were able to
show that a venous pCO2 useful in that a venous pCO2 of less than 5.85 was 100%
accurate in predicting the absence of significant hypercarbia. The limitation of this cut
off is that the reverse is not true with only 57% of patients with a venous pCO2 >5.85
having significant hypercarbia. (Reference 5)
Can The Brachial Artery Be Sampled Safely?

A group at the Texas A&M University Health Science Center, Temple, USA
performed a prospective analysis of 6185 arterial blood gas samples taken in a multi-
speciality hospital. The overall complication rate was 2% with this comprising 1.1%
reporting immediate pain and 0.9% reporting pain developing more than 24 hours
after the procedure. No patients suffered limb-threatening ischaemia as a result of the
sampling, only one required analgesia for the discomfort and the overall haematoma
rate was reported at 0.06%. This paper would suggest that the brachial artery is a safe
site to sample from. (Reference 6)

The literature is however scattered with isolated case reports of complications


following brachial artery blood gas sampling – median nerve palsy following false
aneurysm formation in a woman who had an INR of 2.13, median nerve palsy through
direct trauma and compartment syndrome in a uremic patient with a bleeding
diathesis. The most salient point to be gained by reading these case reports is a
recommendation to always use the non-dominant arm where possible. (References 7,
8 &9)

Should Local Anaesthetic Be Used Prior To Sampling And If So Which Type?

A group at Killingbeck Hospital, Leeds performed an intriguing telephone survey of


100 junior hospital doctors that established that 84% never used localanaesthetic
before arterial blood gas sampling, the reason for this cited by 47% of doctors was
that they considered the injection to be as painful as arterial puncture itself. The group
undertook a randomised double blind placebo controlled trial to establish whether
local anaesthetic is justified. The patients undergoing arterial puncture were randomly
allocated to one of three groups before arterial puncture: A--Infiltration with 2%
lignocaine; B--Infiltration with normal saline; C--No infiltration. Patient and doctor
then rated the discomfort of the procedure. Both patients and doctors rated the pain of
the procedure as a whole less when local anaesthetic was used. The paper suggested
infiltration with local anesthetic was no more painful than with placebo and that
arterial puncture was no more difficult following infiltration as assessed by passes
made, times skin broken and the doctor's rating of the procedure. The group therefore
felt that best practice would be to use local anaesthetic . (Reference 10)

Their opinion is supported by the work of a Spanish group who studied 270
respiratory patients. 210 patients were divided into three groups receiving either local
anaesthetic infiltration, placebo infiltration or nothing prior to arterial puncture. The
pain scores reported using a visual analogue scale were over 50% lower in the group
receiving local anaesthetic when compare to the group receiving nothing. In this trial
the placebo effect was minimal with patients receiving placebo rating their pain
similar to that of the group receiving nothing. The remaining 60 patients had to
compare the pain of arterial and venous punctures. The pain score for venous puncture
was more in this group that that of the pain score for the group receiving arterial
puncture with local anaesthetic first. (Reference 11)
Taking their investigation one stage further the Spanish team then investigated if
EMLA cream could have similar benefits in reducing the pain of arterial puncture.
Although less practical in the acute A&E setting, were the time constrains of waiting
for EMLA to numb the area makes it impractical in unstable patients were results are
required rapidly, it is an attractive idea as it removes the need for a stinging local
anaesthetic injection. Unfortunately a prospective, random double-blind study of 153
patients in three groups: group A, 51 patients who had 1 g of EMLA cream applied;
group B, 52 patients who had 1 g of placebo cream applied; and group C, 50 patients
who received infiltration of 0.2 ml of 1% mepivacaine showed no value in EMLA
use. Pain was assessed on a visual analog scale and was 2.6 +/- 1.8 in group A, 2.9 +/-
1.8 in group B and 1.6 +/- 1.8 in group C. The results for group C were statistically
different from those for groups A and B. The difference between groups A (EMLA)
and B (placebo), however, was not statistically significant. The study did, however,
support the idea that local anaesthetic in the form of a subcutaneous lignocaine-type
agent does reduce the pain of arterial puncture. (Reference 12)

Are Capillary Samples Suitable For A&E Use?

Capillary sampling of arterialised blood from an earlobe has become an accepted


means of obtaining arterial blood gases in chronic respiratory patients but there are
few studies of the use in an emergency department environment.
One paper from a group in Kettering, England 55 emergency medical admissions
agreed to compare arterial blood gas sampling with earlobe capillary sampling. The
first 29 had the arterial puncture performed without local anaesthetic and the
remaining 26 had 1% lignocaine infiltrated before arterial puncture. The pain score
reported ranked capillary sampling as the least painful with arterial sampling with no
local anaesthetic the most painful. Again local anaesthetic infiltration and arterial
puncture was a rated better than the arterial puncture alone although it was scored
significantly more painful than capillary sampling. This paper allowed just three
minutes for the rubefacient to arterialise the earlobe before the capillary sample was
taken and importantly the results showed only trivial differences between the arterial
and arterialised blood samples. The paper does not record the time required to obtain
the two sample types but the preparatory time of three minutes suggests that capillary
sampling technique may not be too protracted for it to be of use in acutely ill patients.
(Reference 13)
Summary
• Pulse oximetry should be used as a guide to likely respiratory failure and
saturations of 92% or more should be reassuring.

• Venous gas samples can give an accurate pH measurement and can be


used to predict the likelihood of significant hypercarbia.

• Brachial arteries can be used if required but the non-dominant arm


should be used in preference.

• Local anaesthetic has evidence to support its use but needs to be in the
form of subcutaneous infiltration with lignocaine or similar.

• Capillary samples may be of use within an Accident and Emergency


setting and give comparable results with a short preparation time.
Guidance for the need to perform ABGs in the Emergency Department

Parameter of interest

Hypoxia pH Hypercarbia

Sats Sats VENOUS VENOUS


> > (ph will be (pH will be
92% or = ~0.03 lower ~0.03 lower
92% than true) than true)

UNLIKELY YES Venous Venous


HYPOXIC ABGs required pCO2 pCO2
(~4%) to quantify < > or =
5.85kPa 5.85kPa

Special True 60% chance


cases pCO2 true
may need < pCO2
ABGs 6.5kPa >6.5kPa

ABGs
to
Quantify
References
Witting MD, Lueck CH. The ability of pulse oximetry to screen for hypoxemia and
hypercapnia in patients breathing room air. J Emerg Med 2001:20(4):341-8

Carruthers DM, Harrison BD. Arterial blood gas analysis or oxygen saturation in
the assessment of acute asthma? Thorax 1995:50(2):186-8

Brandenburg MA, Dire DJ. Comparison of arterial and venous blood gas values in
the initial emergency department evaluation of patients with diabetic ketoacidosis.
Ann Emerg Med 1998:31(4):459-65

Kelly AM, McAlpine R, Kyle E. Venous pH can safely replace arterial pH in the
initial evaluation of patients in the emergency department. Emerg Med J
2001:18(5):340-2

Kelly AM, Kyle E, McAlpine R. Venous pCO(2) and pH can be used to screen for
significant hypercarbia in emergency patients with acute respiratory disease. J Emerg
Med 2002:22(1):15-9

Okeson GC, Wulbrecht PH. The safety of brachial artery puncture for arterial blood
sampling.Chest 1998:114(3):748-51

Yip KM, Yurianto H, Lin J. False aneurysm with median nerve palsy after
iatrogenic brachial artery puncture. Postgrad Med J 1997:73(855):43-4

Watson ME. Median nerve damage from brachial artery puncture: a case
report.Respir Care 1995:40(11):1141-3

Safran MR, Bernstein A, Lesavoy MA. Forearm compartment syndrome following


brachial arterial puncture in uremia. Ann Plast Surg 1994:32(5):535-8

Lightowler JV, Elliott MW. Local anaesthetic infiltration prior to arterial puncture
for blood gas analysis: a survey of current practice and a randomised double blind
placebo controlled trial. J R Coll Physicians Lond 1997:31(6):645-6

Giner J, Casan P, Belda J, Gonzalez M, Miralda RM, Sanchis J. Pain during


arterial puncture. Chest 1996:110(6):1443-5

Giner J, Casan P, Belda J, Litvan H, Sanchis J. Use of the anesthetic cream EMLA
in arterial punction Rev Esp Anestesiol Reanim 2000:47(2):63-6

Dar K, Wlliams T, Aitken R, Woods KL, Fletcher S. Arterial versus capillary


sampling for analyzing blood gas pressures. BMJ 1995:3101:24-25

You might also like