ETCOc hal/I). - BOb. 0. IS. 20. 25. 30. 35 ..... RS, Hopper AO, Cowan. BE, Stevens. GB,. Stevenson. DK. Paired ... 9. Smith. DW, Hopper. AO, Shahin. SM, Cohen.
Clinical
Chemistry
50-56
42:1
(1996)
Evaluation of a fully automated end-tidal carbon monoxide instrument for breath analysis
We
HENDRIK
J.
evaluated
a novel,
VREMAN,I,*
portable
fying
instrument
[Baby’s
lyzer
(BB);
Medical],
or with
Natus
gas
onstrated and
samples
that
precise
tests
excellent
correlation
surements
(ETCO)
determined 96%.
ETCOc
and
of total
(COHbc)
correlated
1.tLIL
CO
30).
The
0.97,
imprecision,
ments
with
that
ETCOc reliable
and
in neonates
mean
term
>3
correlated
can
be used
and
0.25
that
[7-91
be measured
=
hemne degradatian inhaled nous
known
CO
is
ever,
measure
severely
.
hemolysis
#{149} hyperhiliru-
ment
neonates
.
application
hampered
by
including Physiologic
jaundice
Furthermore, blood,
and
neonatal hilirubin The
production
spleen
and
and
the relatively is
liver
can
and
and
and
the neonatal
accumulations
liver premnaturity
natural
1.14.99.3)
during
disease,
hyperbiliriabinemnia
the
sources,
is common
hemolytic
exacerbate
the
period.
try
of extravascular the
delicate
been
degree
balance
of
results
smnall amount
degraded
by
the
to biliverdin
and
carbon
caf erythrocytes
release
derived
of heme. from
enzyme
This
heme
oxygenase (CO)
Because
of Pediatrics,
Stanford
University
School
in heme,
portable
Author
Medical, Inc., San Carlos, CA 944)71). for correspondence. Fax 415-725-7724;
e-mail
henk
(EC
March
27,
1995;
accepted
September
28,
more breath In
ETCO of the
when
can rate
of
corrected
for
of endoge-
a novel
of envi-
concept.
measurements
of
obtaining
Howhas
obtaining
the
been
necessary
representative
breath
of analytical
equip-
contamination,
or improper
operation.
been
with
techniques,
measured (GC)
of its specificity
and electrochemis-
and
measurements. and
several
[6-11]
sensitivity,
GC
However,
this
fairly complicated
has
meth-
and thus is
of breath, especially from neonates, isdifficult
automated
breath
we
initially developed
sampler
and
a semi-
electrochemical
for use at the bedside to help identify neonates
mnight be at risk for hyperbilirubinemia mnent (af this concept
[15/.
Further
CO who
develop-
has led to the availability for clinical
of a portable, fully integrated breath sampler/
Stanfird,
Nonstandard abbreviations: lIb, hemoglobin; GO! lb carhoxyhenioglohin; ETCO, end-tidal breath GO; ETCOc, end-tidal breath Cd) corrected for inhaled Cd); CC, gas chromatography; BB, Baby’s Breath’ Carbon Monoxide Aaalyzer; BBc, BBh, Baby’s Breath instrument in catheter mode or l,alloon mode, respec-
vrenian
. Received
or,
equilibrium. bilirubin,
susceptibility
to) error. Therefore,
analyzer /14]
(:A 94305. 2 Natus
of
is time consuming
and prone
in equimolar
of Medicine,
costs
the
for these
research purposes Department
[6/
not suitable for routine use in a clinical setting. Furthermore,
nonerythropoietic
monoxide
by
concentrations
is hartily
chromatography used
of
estimated
determination
to toxic
the
and has
obtaining samples degradation
in the
13].
widely
odology
between
elimination.
pathological
/12,
gas
the
of the
degradation be
an index
of these
difficulty
for analysis, in breath
/IJ
in end-tidal
and
for the
smoking
to interferences, CO
the
excretion
provides
exposure
and
the
in
(80%
/9J.
clinical
equipment,
CO
to the
dissociates
production
measurements
production,
the
and
binds
to formri car-
physiological
and biliruhin
CO.
and
can
of COHh
(ETCOc) CO
ronmental
instrument
at
as an
most
concentration
subjects
noninvasively
CO
Breath
adults.
carboxyhemoglobin
TERMS:
binemia
of
of
from
body
the CO
is
index
COHb
Because production
of total
which
skin
erythrocytes
originates
to measurements
samiiples INDEXING
rate
-
showed
to noninvasively
the
body
to bilirubin,
of the cells
[5, 6J.3
biliruhin
by measuring
contrast
popula-
with
this
n
of
out
is exhaled.
in the rate
mneasuring
Measure-
newborns
the
(ETCO)
0.01,
11%.
produced
simply,
CO
ETCOc =
of the
was
pilL
as
the CO
and
circulating
(COHb)
where
heme,
was
for inhaled
hemolytic conclude
(ETCOc)
SE slope
by the
We
conditions.
clinically ETCO
of
CO
mea-
efficiency
=
0.47,
=
showed CO
carboxyhemoglobin
determinations,
values
hemolytic
blood
COHbc
S
and
CO
corrected
assessed
healthy
adults
sampling
lungs,
STEVENSON’
is reduced serumn,
diffuses
of the
boxyhemoglohin
(GC)
biliverdin CO
(Hb)
K.
DAVID
in plasma,
/3, 4]. The
hemncaglobin
accurate,
breath
inhaled
Breath and
CV for triplicate
lion’s
for
GC.
hemoglobin)
=
with
end-tidal
strongly:
(r2
linear,
measured
jaundice
dem-
chromatography
performed
between corrected
0.01
gas
/1, 2/. The
usually
bedside
tests
and
STONE,2
quantities
at the
Bench
were
with
measurements
(%
bags.
T.
ROBERT
Ana-
Monoxide
for use
into
(n = 30)
by BB
fractions
developed
compared
BAXTER,’
sampler/CO-quantiCarbon
CO measurements
the
results. In vivo
breath
Breath’
collected
when
M.
LOUISE
1995.
tively;
50
and (;ollhc,
carhoxyhemoglobin
corrected
for
inhaled
CO.
electrochemical Breath
instrument
(BB;
Natus
incorporates
several
a disposable
activated
contaminants,
such
settings;
(b)
CO
anti
or
unconscious we
the
as an index
human
adults phases.
of anal-
to assess
testing
with
the
evaluate
and
(e)
a smnall
imnprecision
first
and
of newborns
was
oaf the
and
BBc
GC.
anti
three
of
collected1
was done
by the
assess
analysis
the
study
versity
protocols
blood
Declaration
the
obtained
Healthy
Stanford was
in accordance
smoking
(n
employee SD
±
pool. 38
Subjects
ranged
different ethnic origins: Caucasian panic
(12%),
The
smnokers
and
Native
usually
neonates
American smoked
(n
University Medical
in age fromn 20-72
13 years, median
±
35 years) and
(60%), (4%);
Asian
years were
(24%),
13(52%)
of
His-
were
female.
and
noses of suspected 1), Rh
isoimmunization
drogenase
deficiency (n
=
was important
were
with retested
(n =
=
BB
the
following
because (n
instrument day,
of diag-
2), anemia
=
I), glucose-6-phospbate
(n
(n
=
1). It
the sampling
were
made
in triplicate
(BBc).
Most
of the neonates
a total
of
via
17 ETCOc
measurements.
‘
tional
The device
Breath
for
research
‘
instrument purposes
is commercially only,
pending
FDA
blood
for
Triplicate
tubes
and
were
X
samples
the
BBc
ccallected
room
then
COHb
(76
Bloaod
air into
sealed
with
determinatians
was
mmn (i.ti.),
capacity
stored
O(
1.7
were
air samples
CO,),
at 4
tmntil
were imnmnediately analyzed
by GC
ETCOc with
(H,,
(:0),
measurements
the
nasal
reproducibility
restudied
were
AND
Gases. the
Go
the
and
by infrared
were
catheter
made
with
to determine
anti
ETC()c
next
day.
the
ETCOc Most
concemitrations.
BB
instruments from
calil)rateti pure
Scott with
CO
were
INSTRUMENTATION
in air was
GG
prepared
(99.3%
filled 30
1-1,-free
air
.tL/L).
from
with
and
were
ob-
PA)
and
against
appropriate
FL-free
air.
was
obtained
Mixing
by
(:0-
air
filter
into
was tione
camnpressetl
combustion
(:0
Liquid
(Hamilton,
cycles.
by passing catalytic
of
syrimige
syringe
compression/decompression
of
with H,,
(100%;
gas-tight
l000-mL
and
mixtures
volume or H2
a calibrated acrylic
jzL/L)
and mixtures
Gases)
calilarated
Hopcalite-containing ical, Menlo
University the
(10.8 mixtures
(Plumsteadville,
mixtures,
Specialty
(:0-
GC gas
Gases
injecting IL)
the The
air.These
Scott
Chicago, NV)
to) calibrate
at Stanford
by
analyzed;
Carbonic,
used (20.5 Specialty
in CO-free
and
through
(Trace
a
Analyt-
Park, CA).
Sampling
bags.
The
collected
from
adults
Measure
Bag; National
chromatography.
The
measurements
was
Trace
of both
H,
molecular
on
available
as an investigafor
Samples
O(
a 13X with
a CO-free (100-1000
air samples
room
l000-mL
polyethylene
Draeger,
Pittsburgh,
comparison GG
with
Analytical),
quantification
approval
use.
into
Gus
85
clinical
breath and
CO (RGA2;
Baby’s
Finally,
with
subject
breath which
were
dehy-
at rest during
making
each
mL),
by
smoking)
in triplicate
intervals,
of end-tidal
and
instrument
performing
weighing
I), and factor D hemolysis
mneasurements
the
Unit
3), pneumnonia
to keep the neonates
ETCOc
procedure. catheter
Intensive Care
sepsis (n
after
(GO,).
Reno,
2.8 to 4.1 kg, were selected for study fromn the populatiomi of patients in the Neonatal
(CO
a septum-sealed,
cigarettes per day.
2-30
9), 1-7 days postpartum
=
15) atlult
=
measured
BBb
measurements
mm
(1000
breath and room
tained (n
to demiionstrate equiv-
COHbc
(10
I-olin
samnples
BBh
SUPPLIES
anti
with the Helsinki
10) and nonsmoking
=
to in the
Uniobtained
in the study. Breath
were recruited fromn the Stanford
Term
the
consent
BB instrumnents
vs BBc and toacompare
with
in custom-made
Neonates.
of 1975, as reviseti in 1983.
volunteers (mean
fromn
by BBb
ETCOc at
bags
measurement
Informned
subject before participation
samnples were
Center
approval
Board.
the exact
transferred
ncanin-
and Methods
received
Review
their
Then,
necanates
Internal
for each
had
valves.
The
to
instrument.
SUBJECTS
HUMAN
were
analysis.
BB
The
to determine
mixtures
by three
smokers
180 1iL] via fingerstick.
different
of possible excessive hemolysis.
Materials
of (:0 anti (or)
(BBh).
and
separate
consisteti
to
CO
measurements
Nonsmokers
three-way
for
for
ETCOc
in
with
studied
analyzed
by GC The
sensor system
mixtures
Tests with adults were perf(armed
hemoiysis
Subsequently,
instrument
them
and H,.
alency of the measurements
polyethylene
adults
we prepared
of CO
or bag mode
in
integrated
GO,
analyzed
then
CO
study
performance. of the
performance
vasive detection
Oui
nonsmoking
functions
anti evaluate
t(ameasure
bench-tested
in vitro
smoking
group
and
[8/.
were
sampling
and
instrumnent.
instrumnent,
pathological
newborns their
bags
Adults.
and
as neonates,
method,
and imprecision
test the ability of the instrument’s quantify
in air and
balloon
mneasure-
the
To
assur-
(H2);
of
relative to the GC
Instrumnents
gas mixtures
Finally,
H,
internal
principle
of nattmral and
[101
tests.
to accurately
seated
design
for accuracy
breath
1, 1996
concentrations
clinical
quality
subjects, such
No.
in
adults.
describe
its performnance
in vivo
(
42,
of organic
ratiometric for
hydrogen
in noncooperating
the instrument
three
(c) automated efficiency;
(a)
mneasuremiient
concentrations breath
Bench
fcaund and
fcar
PROTOCOLS
device
adsorption
adjustment
and
Baby’s
This
samnpling of end-tidal respiratory gas anti analysis for
Here report
for
[16-18J: interference:
frequently
(CO,)
efficacy
concentrations infants,
filter
Chemist-iy
CA).4
to minimize
alcohols, zero
CO
Carlos,
features
compensation
automated
San
carbon as
dioxide
of sampling
ment
quantifying
concentrations;
of carbon
ance
novel
automnated
background ysis
for
Medical,
Clinical
H,
and
technique alkawed
CC) [19/.
sieve air
carrier
jiL)
were
column
Co [80
at a flow introduced
(Alcotest
PA).
a Redtmction
which
I L) were
(-
bags
ft)r laboratory Gas
for
Detector
simultaneous
was separated X
0.22
fromii
cmn (i.d.)I
rate of 31 mL/min. with
a
gas-tight
at
52
Vreman
syringe
into
the
attached
to the
The
separated
reacting
with
ated The highly
CO
solid
mercuric
(Hg)
[20].
of GO
GC
(HgO),
Hg
was
and
H,
of Hg
was
limits
before
linear (50%
generwith
this
bed.
after
1.0 cm into
subject’s
upper
[2.0 mm
(o.d.) x 1.0 mm
to
breath
0.3
content.
The
GO)
fraction
earlier GC
in
[5, 19]. for
blood
Room
air
correction
(GOHb
corrected
was
determined
CO
concentrations
of GOHb
by
for
GC
as
were
to COHbc
inhaled
described
[6/.
CO
[30
Breath x
14
carbon X 22
measuring,
monoxide
cm
at the
neonatal
analyzer.
x
(w
d x
bedside,
diagnostic
This
h),
5.4
ETCOc
use with
portable
instrumnent
was
for
in hospitals,
rooms.
sampling
with a catheter (BBc), the instrument
used
in the
air taken bags
balloon
by classical
and
the
stantially single
two
with
management time),
printer, a RS-232
tubing
port
activated
for
charcoal
The air)
GO
the average during
of CO2.
is used GO
and
CO2
of 0.1 L/L number
and
display
measurement The
mLIL
and
and
a hard
period
can
and
20.5
L/L
in the these
and
1 mL/L, per copy
For
also
be obtained
were
calibrated
final
ETCOc
in-line
and
according
to
When
with
a
date,
recorded.
A
informawith
activated,
the
instrument
concentration
of
samples CO
in
room
While
1505;
trap
Servomex,
GO2,
of
into
a
Nor-
single-wavelength
component
the
2 mL
instru-
of the BB instrument,
with
a resolution
of 0.1 mL/L
million
are
unit
L/L.
used
presented
to between
the data
in these
Linear
regression
the
the
(S1,.) and
of variance
was with
was used to determine data
three BB instruments
sets
in terms
to the
SI concen-
equations
and
72,
To
describe
the
imprecision
we have
included
the
standard
the standard
whether
closely
(r), relaof error
error of the slope. In
(at 95%
addition, confidence
statisticaldifferences exist during
bench
and for single measurements
of various multiple measurements
linear
at P 3
5
.tL/I.
represent
having ETCOc
(j.LL/L)
sured
-
by
term The
BBc
7.3 ± 4.1 iL/L
values -J
0
For
nonhemolytic
neonates n
and
term
The
mieonates,
imprecision
neonates
in
5), respectively.
=
healthy
is 3.1 p.L/L.
as
as mea-
(0.9-3.0 .tL/L, n
tL/L, for
in these
(a,
were
pling.
diagnosed
ETCOc,
1.9 ± 0.6 iIJL
value
2 SD,
SD
is given
as
for the
in Table
1.
/9, 15/, we have used the highest value of triplicate
measurements
LU
+
previously ±
(3.7-14.4
ETCOc
determination
Previously
LU
was
been
mean
term
neonates,
by mean
ETCOc LU
The
in healthy
mnaximnumn
defined
whoa had
disease.
hemnolytic 12) and
=
patients
hemolytic
b, c) of
caused the
by
ETCOc,
BB instrument,
a single
measurement.
ET(:Oc
data
ET(:Oc
mneasurement
assuming
inaccurate
manual
that or
the mamiufacturer
Therefore,
the
lower
automated
sam-
suggests
we
separately
>3
pilL
by
(a),
the
mean
making
analyzed
the
U
F-
3
measurements
the
or
greatest
2
4
6
8
0
2
14
6
Fig. 2. Accuracy assessed
of the
indicated
with
the
ETCOc
measuring
of
(n
triplicate analysis
(B)
30) measured Linear regression
in
ability
of the
chemical also
Fig. 3. Validation of nonsmoking collected
samples
of the
20 (al/LI
25 -
ETCOc
and smoking analyzed
30
35
40
BBc
10
15
20
ETCOc hal/I)
method:
subjects in the
of variance
difference
between
any
and a single measurement.
Discussion results caf bench
comparison
measured
balloon
mode
method,
show
25 - BOb
sensors
sensors
parison
15
(a, b, c),
Analysis
no significant
first
of the first two
measurements
measurements.
measurements
technique
The
of nonsmoking and smoking adult subjects directly via the balloon mode )BBb) method. for a subset of the data in A, i.e., CO 16
measurements
methods
of the
indicates
concentrations
ETCOc
measurements
The ity
studied,
at ETCOc for
ETCO
30)
=
and
instrument
centrations, which and
with
relevant
in neonates
have
with the BB
the results
comparison
/11/
possible
population
in
of in-
cost-effective
introduced
decisions
the analyzer
of hemolysis
are equivalent. However,
1.tL/L,
clinically
for neonates,
from
24,