-
Chapter 24 Pulse Oximetry P.776
Introduction Pulse oximetry, sometimes called the fi fth vi tal
sign , is a noninvasive method of
measuring hemoglobin saturation (SpO2) by using a light s ignal
transmitted through
t issue. A low SpO2 can prov ide warning of hypoxemia before
other s igns such as
cyanosis or a change in heart rate are observed.
Unt il the 1980s , non invasive ox imeters, known as ear
oximeters , were large,
expensive, and cumbersome. They required arte rial ization by
heat or chemical
trea tment, and the ir u ti l i ty was l imi ted by dif f icul
ties in different ia ting l igh t
absorbance of arterial blood f rom that of venous blood and
tissues.
Technical advances, inc luding l igh t-emitt ing d iodes (LEDs),
min ia turized
photodetectors, and microprocessors, allowed the creat ion of a
new generat ion of
ox imeters, which were smal ler, less expensive, and easier to
use. These
differentia te the absorp tion of l ight by the pulsa ti le
arterial component f rom the
stat ic components, so they are cal led pulse oximeters .
A pulse oximeter may be a stand-alone device or incorpora ted
into another device
such as a mult iparameter moni toring system. A relatively new
development is a
combined pulse oximetry and transcutaneous carbon dioxide
tension ear sensor
(1,2 ,3,4).
The American Soc ie ty of Anesthesiologis ts (ASA) and American
Associat ion of
Nurse Anesthetis ts have made assessment of oxygenation a s
tandard for
intraopera tive and postoperative moni toring. In 2005, an
audible alarm for the pulse
ox imeter was added to the ASA monitoring s tandard (5). In
ternat ional s tandards for
safe practice endorsed by the World Federation of Societ ies o f
Anes thesiologists
highly recommend cont inuous use of a quanti tative monitor of
oxygenation such as
pulse oximetry (6). In some states, the use of pulse oximetry is
mandatory. A study
of c losed claims of anesthet ic -rela ted malpract ice cases
determined tha t a
combination of pulse oximetry and capnography could have
prevented 93% of
avoidable mishaps (7). One study determined tha t pulse ox
imetry prov ided the f irs t
warning of an inc ident in 27% of si tua tions (8). The number
of unant ic ipated
-
intensive care uni t admissions dec reased afte r the introduct
ion of the pulse
ox imetry (9).
Operating Principles The pulse oximeter estimates SpO2 f rom the
dif ferent ial absorpt ion of red and
infrared ligh t in t issue (10,11,12,13,14,15,16). The two wave
lengths al low
differentia tion of reduced hemoglobin and oxyhemoglob in .
Reduced hemoglobin
absorbs more l ight in the red band than oxyhemoglobin (Fig. 24
.1). Oxyhemoglobin
absorbs more l ight in the inf rared band. The pulse oximeter
computes the ratio
between these two s ignals and relates this rat io to the arte
rial oxygen saturat ion,
us ing an empirical a lgori thm.
Pulse oximeters discriminate between arterial b lood and other
components by
determin ing the change in transmitted l ight caused by the f
low of a rterial blood. The
ox imeter pulses the red and inf ra red LEDs ON and OFF several
hundred times per
second. The rapid sampl ing rate allows recogni tion of the peak
and trough of each
pulse wave. At the trough, the l igh t is transmitted through a
vascular bed that
contains mainly capil lary and venous blood as wel l as
intervening t issue. At the
peak, i t sh ines through all this plus arterial blood. A
photodiode collec ts the
transmitted light and converts i t into elec trical signals .
The emitted s ignals are then
amplif ied, processed, and displayed on the moni tor. Oximeters
have a phase wi th
both LEDs OFF to allow detect ion of and compensat ion for
extraneous l ight. Light
readings during the OFF period are subtracted from the next
sequence.
Fract ional oxygen satura tion (% HbO2 ) is the ra tio of
oxyhemoglobin to the sum of
al l hemoglobin spec ies present, whether avai lable for revers
ible binding to oxygen
or not (17). Func tiona l oxygen saturat ion (SaO2 ) is def ined
as the ratio of
oxyhemoglobin to al l functional hemoglobins . These must be
determined by using
an
P.777
in v itro oximeter. For pat ients wi th low dyshemoglobin levels
, the d if fe rence
between f ract ional and functional sa turation is very smal l .
However, when
dyshemoglobin levels a re elevated, the two values can vary
great ly, and pulse
ox imeter readings may not agree wi th e ither the true f rac
tional or func tiona l
sa turat ion values (18).
-
View Figure
Figure 24.1 Absorbance of light as a function of wavelength. The
extinction coefficient is a measure of the tendency of a substance
to absorb light. At the red wavelengths (650 to 750 nm), reduced
hemoglobin absorbs more light than does oxyhemoglobin. In the
infrared region (900 to 1000 nm), the reverse is true.
Transmission Pulse Oximetry The mos t common type of pulse
oximeter is the transmission ox imeter. With th is
technology, a l ight beam is transmitted through a vascular bed
and is detected on
the opposi te side of that bed.
Reflectance Pulse Oximetry Ref lectance oximetry rel ies on l
igh t that is ref lected (backscattered) to determine
oxygen saturat ion. The probe has both an LED and a photodiode
(Fig. 24.16).
Transmission pulse oximetry probes are not accurate when used in
the manner of
ref lec tance oximetry (19). Ref lec tion orig inates f rom
nonhomogene ity in the opt ical
path, tha t is , at the in terfaces between materials with
different reflec tive indices.
The t issue mus t be wel l perfused to obtain a strong signal .
Heating the
measurement s ite and app ly ing pressure may be helpful
(20,21).
There are a number of l imi tat ions of ref lectance ox imetry.
The probe design must
el iminate l igh t that is passed di rec tly to the probe or is
sca ttered in the outer
surface of the sk in . The signa ls are weaker than those found
in transmission
ox imetry, so the photodiode area needs to be as la rge as
possible. I f the probe is
located over an arte ry or a vein , the read ing may be art
ifactua lly low (22,23).
Vasoconstric t ion can cause overestimation of the oxygen
saturation (23).
Physiology
-
Eff ic ient oxygen transport rel ies on the abil ity of
hemoglobin to revers ib ly load and
unload oxygen. The relat ionship between oxygen tens ion and
oxygen binding is
seen in the oxyhemoglob in dissociat ion curve (Fig. 24 .2),
which plots the
hemoglobin oxygen saturation against the oxygen tension. The
sigmoid shape of
the curve is essential for physiologic transport. As oxygen is
taken up in the lungs,
the blood is nearly ful ly sa turated over a large range of
tensions. During passage
through the
P.778
systemic capi llaries, a la rge amount of oxygen is released
with a relat ively smal l
drop in tens ion. This al lows oxygen to be released a t suf f
ic ien tly high
concentrations to provide an adequate gradient for di f fusion
into the cel ls .
View Figure
Figure 24.2 The oxyhemoglobin dissociation curve. Hemoglobin
saturation is plotted as a function of oxygen tension.
The shape of the oxyhemoglobin d issociat ion curve l imi ts the
degree of
desatura tion that can be tolerated. Between 90% and 100%
saturat ion , the part ial
pressure of oxygen in arterial b lood (PaO2 ) wi l l be 60 torr
or above. Below 90%
saturat ion, the curve becomes steeper, and smal l drops in
satura tion correspond to
large drops in oxygen part ial pressure. If a problem develops,
there may not be
much warning before the oxygen level becomes dangerous ly low.
Normal satura tion
wi l l decrease as alti tude above sea level inc reases
(24).
Equipment
Probes
-
The probe (sensor, transducer) is the part tha t comes in
contact wi th the patient. It
contains one or more LEDs (photodiodes) that emit l ight at
specif ic wavelengths
and a photodetector (photocell , t ransducer). The LEDs provide
monochromatic
l ight. This means that they emit a constant wavelength th
roughout their l i fe , so they
never need recal ib rat ion. LEDs cause relat ively l i t t le
heating and are so
inexpensive that they may be used in a d isposable probe. The
light, partial ly
absorbed and modula ted as i t passes through the t issue, is
converted into an
electronic signal by the photodetector.
Figures 24.3 to 24 .11 show several types of p robes. Probes may
be reusab le or
disposable. They have the same accuracy (25 ,26,27,28). A
disposab le probe is
usual ly attached by using adhesive. Reusable probes either cl
ip on or a re attached
by using adhesive or Velcro. Disposable probes may be easier to
use, but reusable
probes are more economical as long as personne l are careful not
to damage the
reusable probe (26,29,30,31). Self -adhesive (band, wrap) probes
are less
susceptible to motion art ifact and are less likely to come off
if the pat ient moves
than those that c lip on. However, they are usua lly not as wel
l shielded f rom ambient
l ight as c l ip-on probes. A ttaching reusable probes by using
an adhes ive or Velco
wrap may improve their s tab il i ty. P robes lined wi th soft
material may be associa ted
wi th fewer motion art ifac ts (32).
Some probes are availab le in dif ferent s izes. If a p robe is
too large for the patient,
some of the l ight ou tput from the LED can reach the photoce ll
without pass ing
through tissue, and falsely high SpO2 readings wil l be produced
(33,34). The
photocel l may not al ign wi th the probe, and readings wi l l
not be possible.
Loss of reusable probes can be reduced by making i t diff icul t
to separate the probe
f rom the cable (31). Attaching the probe to the oximeter case
when not in use wil l
reduce damage and make it easy to f ind (35).
P.779
-
View Figure
Figure 24.3 Disposable flexible probe. This can be placed on a
variety of sites, including the finger, ear, cheek, tongue, toe,
penis, hypothenar or thenar eminence, palm, foot, and wrist.
(Picture courtesy of Masimo Corporation, Irvine, CA.)
To reduce contaminat ion, a glove, the f inger of a glove, or
other covering may be
used either over the appl ica tion si te o r over the probe
(36,37,38). Mi tts are
availab le to shield ambient l ight (F ig. 24.12).
Cable The probe is connected to the oximeter by an electrical
cable. Cables from d if fe rent
manufac turers are not inte rchangeable.
Console Many d if fe rent consoles are avai lable (Figs.
24.13,24.14,24.15). Most oximeters
that are used in the operat ing room are part o f a phys io
logic mon itor. Most s tand-
alone un its are l ine operated but wil l work on batte ries,
making them useful during
transport. Some oximeters are hand-held (Figs. 24.14,
24.15).
A mic rocomputer mon itors and controls s ignal levels , perfo
rms the calculat ions,
implements s ignal val id ity schemes, activates ala rms and
messages, and monitors
i ts own c ircui try to warn of mal funct ions. A variety of
messages may be provided to
info rm the operator of i ts
P.780
functional s ta tus (13). The panel usual ly displays percent sa
turat ion, pu lse rate,
and a larm l imits . Most uni ts have a bright display, al
lowing them to be seen in a
wel l -l ighted room.
-
View Figure
Figure 24.4 Disposable flexible probe in place on a finger.
(Reprinted by permission of Nellcor Puritan Bennett Inc.,
Pleasanton, CA.)
View Figure
Figure 24.5 Reusable probe. This is most commonly used on a
finger or toe. In infants, this type of probe can be placed on part
of the hand or foot. These probes offer good shielding from ambient
light. (Picture courtesy of Masimo Corporation, Irvine, CA.)
-
View Figure
Figure 24.6 Reusable finger probe in place. (Reprinted by
permission of Nellcor Puritan Bennett Inc., Pleasanton, CA.)
The disp layed values for SpO2 and pu lse rate are usual ly we
ighted averages . Some
ox imeters allow the averaging period to be adjusted. A mode
that averages over a
longer period of time may work better if there is much probe
motion (39). Changes
in pulse rate or sa turat ion wil l be ref lec ted more rapidly
if the averaging is done
over a shorter period of t ime.
Pulse ampli tude may be represented by a signa l indicator.
Other uni ts use a
graphic that indica tes pu lse amplitude and may prov ide a
plethysmographic
waveform.
Most instruments provide an audib le tone whose pi tch changes
with the satura tion.
In this way, the opera tor can be made aware of changes in SpO2
without looking at
the oximeter. By using a variab le tone pu lse oximeter,
anesthes ia providers
recognized an episode of oxygen desaturation more quickly than
those using one
wi th a f ixed tone (40). There is usually a means to control
the volume of the audible
s igna l.
Alarms are commonly prov ided for low and h igh pu lse rates and
low and high
sa turat ion. Many units
P.781
generate an alarm when the probe is not properly applied to the
patient o r if for
some o ther reason the signal is inadequate. ASA s tandards fo r
Basic Anesthetic
Moni toring require that the variable pi tch pulse tone and low
thresho ld alarm be
audible (41).
-
View Figure
Figure 24.7 Disposable nasal probe in place. The clip from a
disposable oxygen mask may be used to improve contact and to hold
the probe in place. (Reprinted by permission of Nellcor Puritan
Bennett, Inc., Pleasanton, CA.)
View Figure
Figure 24.8 Reusable probe designed for use on the ear. This may
be used on other locations, including the cheek. (Picture courtesy
of Masimo Corporation, Irvine, CA.)
-
View Figure
Figure 24.9 Reusable probe on the ear. (Reprinted by permission
of Nellcor Puritan Bennett, Inc., Pleasanton, CA.)
Most pulse oximeters offer trend data. Inte rfaces for hard copy
recording and data
management sys tems are usually available.
View Figure
Figure 24.10 Disposable wraparound probe on the foot. (Reprinted
by permission of Nellcor Puritan Bennett, Inc., Pleasanton,
CA.)
-
View Figure
Figure 24.11 Disposable wraparound probe on the toe. (Reprinted
by permission of Nellcor Puritan Bennett, Inc., Pleasanton,
CA.)
Oximeter Standards The internat ional and U.S. s tandards are
quite s imilar (42,43). Among the
provisions are the following:
There must be a means to l imit the durat ion of continuous
operation at
temperatures above 41C.
The accuracy mus t be stated over the range o f 70% to 100%
SpO2. If the
manufac turer c la ims accuracy be low 65%, the accuracy mus t
be stated over
the addi t ional range.
I f the manufacturer c laims accuracy during motion, this and
the tes t methods
used to es tab lish i t must be disclosed in the ins tructions
fo r use.
I f the manufacturer c laims accuracy during condi tions of low
perfusion , th is
and the test methods used to establish it must be disc losed in
the
instructions for use.
There must be an indicat ion when the SpO2 or pulse rate data is
not curren t.
I f the pulse oximeter is p rov ided with any physiologic alarm,
i t mus t be
provided wi th an a la rm system that moni tors for equipment
faul ts , and there
must be an alarm for low SpO2 tha t is not less than 85% SpO2 in
the
manufac turer-conf igured a la rm preset. An alarm for high SpO2
is optional .
An indication of s igna l inadequacy must be provided if the
SpO2 or pulse rate
value displayed is potent ially incorrec t.
-
P.782
View Figure
Figure 24.12 Special mitts are available to shield pulse
oximeter probes from ambient light.
I f a variable pi tch audi tory s ignal is provided to ind ica
te the pulse signa l, the
pi tch change shall follow the SpO2 read ing (i.e., as the SpO2
reading lowers,
the pi tch shal l also be lowered).
Use
Sites Finger The probe is most commonly attached over the f
ingert ip (Figs. 24 .4, 24.6). The
failure rate is less, and accuracy is better when the probe is p
laced on the f inger
than on the earlobe (44 ,45,46).
The f inger is re lat ively sensit ive to sympathet ic system
vasoconstric tion (47,48). If
there is poor ci rcu la tion, a f inger b lock, digital pulp
space infi l tration , or a
vasodi lato r may improve performance (48,49,50,51,52,53,54).
Vigorously rubbing
the f ingert ip may temporarily improve c irculat ion to the
area (50).
-
I f there is dark fingernai l polish or synthet ic f ingernails
, the probe should be
oriented so tha t it transmits light f rom one side of the
finger to the other (55). Some
c lear acryl ic na ils do not affect pulse oximeter readings
(56).
A disadvantage of p lacing a probe on an extremity is that de
tection of desaturat ion
and resaturat ion is s lower than when probes are placed more
centrally
(48,57,58,59,60,61,62,63,64). Response time may be quicker when
the probe is
placed on the thumb (62).
Motion art ifac ts are less frequent when the probe is placed on
one of the larger
f ingers (32). The l it t le f inger may be useful i f the pat
ien t is part icularly large (65).
The probe may be placed over a finger tha t has a burn (66).
The probe should not be on the index f inger during recovery. As
a pat ient awakens,
the pat ient of ten wil l want to rub his o r her eye, usual ly
wi th the index f inger. I f the
ox imeter probe is on that f inger, the cornea can be sc
ratched.
In general , the arm opposi te f rom that on which the blood
pressure cuff is applied
or in which an arterial ca theter has been inserted should be
used. The pulse
ox imeter is sometimes in tegrated wi th the noninvasive blood
pressure monitor so
that the pulse ox imeter wi l l not ala rm during the inf lat
ion cycle if placed on the
same arm as the blood pressure cuff . Insert ion of a radia l
artery catheter is
commonly fol lowed by a trans ient
P.783
decrease in b lood f low and loss of an adequate s ignal for a
pulse ox imeter if the
probe is on a f inger of that hand (67). However, performance is
unaffected if pulse
ox imeter readings are made on the arm in which an arterial
cannula is present (68).
Occasiona lly, poor function may occur wi th probe attachment to
the same ex tremity
as the in travenous infusion, due to loca l hypothermia and
vasoconstric t ion.
-
View Figure
Figure 24.13 The console may be a freestanding unit. (Reprinted
by permission of Nellcor Puritan Bennett, Inc., Pleasanton,
CA.)
View Figure
Figure 24.14 Small handheld, battery-operated pulse oximeters
are often used, especially during patient transport. This unit is
in its recharger. (Picture courtesy of Masimo Corporation, Irvine,
CA.)
The posi tion of the arm may a ffect the reading. In most patien
ts , the SpO2 fa lls
af te r the monitored arm is raised (69). I t may also fall when
the arm is lowered
(70).
Toe The toe is an alternate s ite when the f inger is not
available or the s igna l from the
f inger is unsatisfactory. Detection of desatura tion or
resaturat ion wil l not be as
rapid as with more centrally placed probes (63). The de lay in
detec tion of
hypoxemia may be up to 1 to 2 minutes (64,71). The toe may
provide a more
-
reliable signa l in patients who have had an epidural block
(72). An inc rease in
pulse ampli tude from the toe may be a s ign of a successful
block (73).
View Figure
Figure 24.15 Combined pulse oximeter and carbon dioxide monitor.
(Reprinted by permission of Nellcor Puritan Bennett, Inc.,
Pleasanton, CA.)
Nose The nose is usual ly a convenient loca tion. Nasal probes
respond more rapidly to
changes in saturation than probes placed on ex tremities. The
bridge (Fig. 24 .7), the
wings of the nostri ls , and the nasa l septum have been used
(74,75,76). The nose
c lip f rom a disposab le oxygen face mask can be attached to
the outer surface of a
f lex ib le probe to make it f i t snugly over the bridge
(77).
Accuracy at this s i te is controversial . It has been
recommended under condi tions
such as hypothermia, hypotension, and infusion of vasoconstric
to r drugs. In
hypothermic pa tients, the nasal septum was a more re liable si
te than the f inger
(74,76). However, some studies have found tha t nasal probes
often give grossly
erroneous results and have a higher fai lu re ra te than other s
ites under condi tions of
poor perfusion (78 ,79).
I f the patien t is placed in the Trendelenberg posi tion ,
venous congestion may occur
around the nose, causing the pulse oximeter to display art i f
icial ly low saturations
(80,81).
Ear An ear probe (Figs. 24.8, 24 .9) may be held in place by a
plastic semic i rcula r
device hung around the ear.
-
P.784
Stabi l izing devices such as headbands or around-the-ear loops
may be useful .
Using a c lip may improve the qual i ty of the s ignal (82). The
earlobe should be
massaged for 30 to 45 seconds wi th a lcohol o r vasodi la tor
or EMLA cream can be
applied for 30 minutes prior to probe app lica tion to increase
perfusion (83).
An ear probe can be part icula rly useful when there is finger
motion . Response t ime
is fas ter with an ear probe than wi th a f inger probe
(48,58,62,64,84,85). Under
condi tions of poor perfus ion , some ear probes perform better
than finger probes
(78). The ear is rela tively immune to vasoconstric t ive effec
ts of the sympathetic
system (47). The ampli tude of ear plethysmography wi l l
respond mainly to changes
in pulse pressure. Ear probes may g ive more erroneous readings
than f inger p robes
in patien ts wi th tricuspid incompetence (86). A steep head
down posit ion may resul t
in e rroneous readings (87). An ear oximeter may be combined wi
th a
transcutaneous carbon dioxide sensor (4,88).
Tongue A tongue probe can be made by placing a mal leable
aluminum strip behind the
probe to allow i t to bend around the tongue (89,90). A
disposable probe wrapped
around the t ip of the tongue in the sagi ttal p lane may also
be used (91).
Ref lectance pulse oximetry has been used on the superior
surface of the tongue.
The mouth should be c losed.
Glossal pulse ox imetry has been shown to be accurate (91). This
s ite may be
especial ly useful in pat ients who have burns over a large
percentage of thei r body
surface (90,91). Desaturat ion and resatura tion is detected by
a probe at the tongue
quicker than one on the f inger or toe (63).
A l ingual probe is more resistant to s igna l in terfe rence
from electrosurgery than
probes p laced on peripheral s ites but may be dif f icult to
maintain in place during
emergence (89,90). Tongue quivering may mimic tachycardia. Other
problems are
venous conges tion f rom a head down posi t ion and excess oral
secre tions . The
probe mus t be posit ioned af te r tracheal intubation or insert
ion of a supraglottic
ai rway. I t can be easi ly dislodged.
Cheek A probe with a meta l s trip backing can be used to hold a
disposable probe around
the cheek or l ips (92,93). A cl ip -on probe wi th a cover over
the part of the probe on
the buccal surface can also be used (94,95,96,97). Th is method
of use is not
-
recommended by the manufacturer (98,99). Probes special ly
designed for this s ite
are commercial ly avai lab le .
Buccal pulse oximetry is more accurate than f inger pu lse
oximetry (93,100). Probes
at this location detect increases and decreases in saturat ion
more quickly than
f inger or toe probes (63). Buccal oximetry has been found to be
ef fective during
hypothermia, decreased cardiac output, increased sys temic
vascular res istance,
and o ther low pulse pressure states. This site is useful in pa
tients who have burns
(101). Disadvantages inc lude dif f icult placement, poor
acceptance by awake
patients, and art i fac ts during ai rway maneuvers .
Esophagus This probe uses ref lectance oximetry. The esophagus,
a core organ, is bette r
perfused than the extremit ies during states of poor peripheral
perfus ion and may
therefore provide a more consistent, reliable source for pulse
oximetry wi th
hemodynamic instabi l ity (102,103,104,105,106,107,108). It ref
lec ts changes in
arte rial sa turat ion more qu ickly than peripheral sites such
as the f inger. Th is s i te
may be useful for pat ients who have ex tensive burns, on whom
conventional p robes
wou ld be dif f icul t to place (109,110).
Accurate placement of the esophagea l probe requires prac tice
by the user (107).
Ach ievement of a re liable s ignal may be a problem (104).
Forehead A f lat ref lectance pulse oximeter sensor can be used
on the forehead (82,111,112)
(Figs. 24.16 , 24.17). I t should be placed just above the
eyebrow so that i t is
centered s light ly lateral of the i ris (113). The sensor s ite
shou ld be c leaned wi th
alcohol before app lying the sensor to help secure the adhesive.
Pressure on the
probe f rom a headband or pressure dressing may improve the s
ignal (21,82).
This si te is usually easily accessible. The fo rehead is less
affec ted by
vasoconstric tion from cold or poor
P.785
perfusion than the ear or f inger (114,115). Changes in sa turat
ion can be detected
more rapidly at the fo rehead than at the finger (48,116).
However, pool ing of
venous blood due to compromised return to the heart may cause
low saturation
readings in sup ine patients (81). It should not be used if the
patien t is in the
Trende lenburg posi t ion. There are usually few motion art i
fac ts when the forehead is
used (117,118).
-
View Figure
Figure 24.16 Reflectance pulse oximeter probe. The light source
and sensor are situated next to each other. (Picture courtesy of
Masimo Corporation, Irvine, CA.)
View Figure
Figure 24.17 Reflectance pulse oximeter on the forehead. Note
the headband securing it in place. (Picture courtesy of Masimo
Corporation, Irvine, CA.)
Other Pharyngeal pulse oximetry by using a pulse oximeter
attached to a laryngeal mask
may be useful in patien ts wi th poor peripheral perfusion
(119,120).
Flexible probes (Fig . 24.3) may work th rough the palm, foo t,
penis, ankle, lower
ca lf , or even the arm in infants (27 ,121,122) (Fig.
24.10).
Pulse oximetry may be used to moni tor fetal oxygenation during
labor by attaching
a ref lectance pulse oximetry probe to the presenting part
(20,23,123,124). A
-
disadvantage is that the probe has to be placed bl indly and may
be positioned over
a subcutaneous vein or artery, which wi l l affect the reliab il
i ty of the readings (23).
Fixation Proper probe placement is c rucial fo r good
performance. A malposi tioned probe can
resul t in false-posi tive and false-negative alarms. Probes can
be total ly o r partially
dislodged without this be ing noticed.
Adhesive probes may s tay on better than c lip-on probes. It may
be benef icial to
tape probes in place when they wi l l be inaccessible during
surgery, but it is
importan t to avoid compression of the f inger or other part .
Wrapp ing the l imb l ightly
wi th gauze may help to f ix the probe in posit ion . Another
method is to s lip the cut
f inger of a glove over the probe (125). The probe should be
protected from bright
l ight (F ig. 24.12).
Stabilizing the Signal The search that the pulse oximeter goes
through when a probe is ini tia l ly appl ied
(or dis lodged) includes sequential tr ia ls of various intensit
ies of l ight in an effort to
f ind a signal s trong enough to transmit th rough the t issue
but not so s trong that the
detec tion system is saturated (12). Once a pulse is found,
there is usua lly a delay
of a few more seconds whi le SpO2 values for several pulses are
averaged.
Appearance of a satis fac tory waveform is an indicat ion that
the readings are
reliable. Comparison of the pulse rate shown by the oximeter and
that by an
electrocardiograph moni tor is a lso an ind ication that
saturation readings are
reliable. A disc repancy between the rates is f requently an
indication of probe
malposit ion or malfunction . A discrepancy can also occur
during certain
dysrhythmias.
Reusing Disposable Probes Because d isposable probes are costly
, many inst itut ions reuse them
(126,127,128,129,130,131,132,133,134). Although concerns about
this have been
raised (135,136), several s tudies show tha t the fai lure rate
of reprocessed probes
is equal to or less than that of new probes, and the accuracy is
not affected
(132,133,134,137).
Testing Dev ices that can be used to test pulse oximeters are
available (138). Some al low
the tes ter to se t the p lethysmograph ic waveform at dif
ferent ampl itudes and to test
the accuracy of the heart rate as we ll as the SpO2.
-
Applications
Monitoring Oxygenation Anesthetizing Areas Oxygen desatura tion
can occur a t anytime during anesthesia , regardless of the ski l
l
and experience of the anesthes ia provider. Desatura tion
greater than 10%
P.786
occurs in up to 53% of anes thetized patien ts
(139,140,141,142). Pediatric pat ients
are especially at r isk (121,143,144,145,146,147). Mos t severe
desaturations occur
during induct ion or emergence. During maintenance,
desaturations are milder but
more frequent (148). Studies have shown that a reduc tion in the
number of
hypoxemic events occurred when pulse oximetry was used
(144,149). The inc idence
of myocardial ischemia was also decreased (150).
Pulse oximetry may help to detect inadvertent bronchial
intubation (8,151,152,153).
This method is not always reliable, part icula rly if an
elevated inspired oxygen
concentration is being used (12 ,154). The absence of desaturat
ion does not rule
out bronchial intubation (155). Methods to detect b ronchial
intubat ion are d iscussed
in Chapter 19 .
Oximetry is useful in managing one-lung anes thesia to help
assess the
ef fectiveness of measures taken to increase the oxygen sa turat
ion (156) Chapter
20).
Oximetry is useful for patients undergoing regional and moni
tored care anesthesia
(157,158). Of ten, the signs of hypox ia are confused with
restlessness from an
inadequate block. Ins tead of supplying oxygen and assist ing
resp irat ion , add it ional
sedation is provided, which compounds the problem. With
oximetry, the pat ien t's
oxygenation status can be assessed and measures taken to improve
SpO2.
Pulse oximetry may be useful to conf i rm correct tracheal tube
placement when a
functional carbon dioxide moni tor is not avai lab le (159). I f
oxygen saturation rises
af te r in tubation , correct tube placement is l ikely.
However, a pu lse oximeter should
not be rel ied on fo r th is purpose because preoxygenation may
delay the onset of
desatura tion beyond the time when esophageal intubation would
be considered
l ikely.
Other problems that can cause a drop in oxygen saturat ion
include fat embolism,
amniot ic f luid embolism, pulmonary edema, breathing system
disconnec tions and
-
leaks , aspirat ion, tracheal tube obstruction, hypoxic gas mix
ture, oxygen del ivery
failure, hypoventilat ion, anaphylaxis , bronchospasm,
pneumothorax, malignant
hyperthermia, and pulmonary embol ism
(8,160,161,162,163,164,165,166,167,168).
Causes of hypox ia related to equ ipment are discussed in
Chapter 14 .
Postanesthesia Care Unit The recovery room is another location
where desatura tion is common
(141,169,170,171,172,173,174,175,176,177). Routine oxygen
administration to
recovering patients may not be necessary when patients are
monitored with pulse
ox imetry (178). Before leaving the recovery room, a tr ial of
breathing room a ir wh ile
moni toring oxygen saturat ion may provide an indicat ion of the
need to cont inue
oxygen beyond the pos tanesthesia care un it (PACU) or to re ta
in the patient in the
unit for a longer time (179).
Transport Unrecognized oxygen desaturation may occur while the
patien t is being transported
between the opera ting room and the PACU and between that unit
and other areas
(180,181,182,183,184,185,186,187,188,189). Pulse oximetry is
included on mos t
transport moni tors , and portable pulse oximeters are
available.
Other Intrahospital Areas Patients frequently experience hypoxic
episodes in the pos topera tive period after
leaving the PACU (190,191,192,193,194). Pulse oximetry can
detect these episodes
and a id in deciding when oxygen therapy should be discontinued.
Telemetric pulse
ox imetry monitoring may be a cost-effect ive method of
maximizing qual i ty of care
when used to monitor patien ts on a genera l care f loor
(195).
Pulse oximetry is useful fo r moni to ring pat ients in the
intens ive care unit (196). It
may be he lpful during weaning from art if ic ial venti lat ion
(197,198).
Pulse oximetry has been used during cardiopulmonary resusci
tation (199,200,201).
Because of art ifacts and lag t imes, i t is more useful in
primary respiratory arres t
than in card iac arrest. I t is useful in assessing oxygenation
during newborn
resusci tation (203).
Another area where pulse oximetry has proved useful is the
emergency department
(204,205,206).
Ref lectance pulse oximeters can be useful for assessing fetal s
ta tus during labor
and delivery by applying a forehead probe (207).
-
Pulse oximetry is useful in identify ing wh ich patien ts wi th
tonic-c lonic seizures are
at r isk of hypoxic cerebral brain damage (208).
Out-of-hospital Use Pulse oximetry is useful in the prehospi tal
setting, inc luding when transporting
patients by hel icopter or ambulance
(209,210,211,212,213,214,215,216).
Controlling Oxygen Administration Pulse oximetry allows the
lowest safe oxygen flow and concentration compatible
wi th safe levels of a rterial oxygenation to be administered.
Keeping the oxygen
concentration and f low low wi l l help to dec rease the risk of
a f ire (Chapter 32).
Monitoring Peripheral Circulation Pulse oximetry can detect arm
posit ions that compromise c i rculat ion (217). The
pulse oximeter tha t is attached to a toe can help to warn of
decreased perfusion at
the foo t in patients in the l ithotomy posit ion (218).
However, i t cannot rel iably
detec t inadequate perfusion (219,220).
Moni toring oxygen saturat ion during shoulder arthroscopy has
been recommended
as a test for
P.787
brach ial artery compress ion (221). However, an adequate pulse
s ignal may be
present wi th brachial plexus compression (222).
Pat ients with l imb fractures may have compromised c i rcula
tion distal to the
f rac ture. Pulse oximetry may serve as a usefu l guide to b
lood flow to that area
(223,224,225). However, i t may not be helpful in warn ing that
a compartment
syndrome is developing, because diminution of the arterial pulse
dista l to the
compartment is a late s ign (226).
Pat ients who undergo mediastinoscopy are at r isk fo r
brachiocephal ic artery and
aort ic arch compression between the mediast inoscope and the
sternum. Arte rial
compression may be detec ted by measuring pulse wave ampli tude
on a pulse
ox imeter (227,228).
Pulse oximetry may be used to evaluate the effect of a
sympathetic block as
indicated by an increase in peripheral blood f low (229,230). It
may be useful during
and after angiography to detect inadequate blood f low
(231).
-
Pulse oximetry may be used to determine the best s i te of
amputation or a rterial
bypass surgery (232). I t has been used to moni to r reimplanted
or revascularized
digi ts (233,234,235).
Pulse oximetry can be used to measure pa lmar col la teral c
irculat ion
(236,237,238,239,240,241,242,243,244). However, i ts usefulness
for this has been
disputed (245,246). There is a report of radial artery occlusion
that was detected by
pulse oximetry (247). It has proved usefu l in evaluat ing a
pain ful hand af ter
creat ion of an arte riovenous f is tula (248). A s imilar test
of the collateral ci rculat ion
may be performed on the foot by using pulse oximetry (237).
Determining Systolic Blood Pressure A pulse oximeter can be used
to determine the sys tol ic blood pressure
(249,250,251,252,253,254,255,256). The blood pressure cuff is
applied to the same
arm as the pulse oximeter. The cuff is inf lated slowly, and the
pressure at the point
at wh ich the waveform is lost is no ted. It also can be
determined by inf lat ing the
cuff we ll past the systolic pressure and looking for the onset
of a s ignal as the cuff
is def lated . One study found that the bes t agreement wi th
Korotkoff sounds and
noninvasive blood pressure equipment occurred when the average
of blood
pressures est imated at the disappearance and reappearance of
the waveform was
taken as the systolic pressure (257). In pediatric pat ien ts ,
blood pressure
determined by th is method was found to be more accurate than
tha t determined by
an automatic noninvasive blood pressure monitor (258).
Pulse oximetry has been used for patients with pulseless d
iseases of the
ex tremit ies to monitor sa turation and sys to lic blood
pressure (259).
Locating Arteries When the axil lary artery cannot be palpated,
i t may be located by plac ing a pu lse
ox imeter on a f inger on that s ide and pressing in the axil la
unti l the pulse wave
disappears (260,261,262). Pu lse oximetry has also been used to
locate the femoral
and dorsalis pedis arteries by using a pulse oximeter app lied
to a toe
(231,263,264).
Avoiding Hyperoxemia In p remature neonates, administration of
oxygen may be associated with
development of retinopathy and other pathologic condit ions.
Pulse oximetry can aid
in t i trat ing insp ired oxygen by detecting hyperoxemia
(265,266,267,268,269). I t is
-
recommended that the high SpO2 alarm be set at 95% or lower fo r
th is purpose
(269,270,271).
Monitoring Vascular Volume and Sympathetic Tone I f the pulse
oximeter beg ins skipping beats or performing intermit ten tly, the
cause
could be hypovolemia (272). A corre lat ion between pu lse
waveform ampli tude
variation during pos itive-pressure venti la tion and
hypovolemia has been reported
(272,273,274,275,276). I f brief interruption of vent ilation
causes the waveform to
return to normal or more cons tant funct ion, a tr ial of f lu
id therapy may be
warranted.
One of the mos t useful and commonly overlooked plethysmographic
features is
waveform ampl itude (277). Ampl i tude changes can be concealed
by the auto-gain
function found on most pulse oximeters . When the auto-gain is
tu rned OFF, certain
observat ions can be made. The plethysmograph s igna l ampl i
tude is di rectly
proport ional to the vascular dis tensibil i ty over a wide
range of card iac output.
During anes thesia, the unga ined pulse oximeter s ignal may be
used to de termine
the extent of attenuat ion of the sympathetic response to st
imuli .
Another important fea ture of the waveform is the dic ro tic
notch (277). The notch
tends to descend toward the basel ine during increas ing vasod
ilation and gets
higher with vasoconstric t ion .
Other Uses Other s i tuat ions where oximetry may be usefu l inc
lude high-frequency jet
venti lation and determining the effect iveness of therapeutic
bronchoscopy. I t can be
combined with measurement of mixed venous oxyhemoglobin saturat
ion to estimate
oxygen consumption (278,279).
Pulse oximetry has been used to gauge pulmonary b lood flow in
infants and
ch ildren with cyanot ic congen ital heart lesions
(280,281).
P.788
Pulse oximetry can give warning of f luid extravasation
(282).
Advantages
Accuracy Pulse oximetry is accurate , and accuracy does not
change with time. Numerous
studies have shown that the dif fe rence between sa turat ion
determined by pulse
-
ox imetry and arterial blood gas analys is is c linical ly ins
ignif icant above an SpO2 of
70%
(13,14,26,28,46,57,60,121,156,267,283,284,285,286,287,288,289,290,291,292,293,
294,295,296,297,298,299,300). Mos t manufac turers claim that
errors are less than
3% at saturat ions above 70% (243). This accuracy shou ld be
suff icient for mos t
c linical purposes , except possibly neonatal hyperoxia. Changes
in accuracy are
negligible over temperatures encountered in c linical use
(301).
Pulse oximetry is accurate in pat ients wi th dysrhythmias, p
rov ided tha t the SpO2 is
s table and the plethysmogram is noise-free and has reasonable
ampli tude (302).
The SpO2 may be correct even if the pulse rate is not.
Independence from Gases and Vapors Pulse oximetry readings are
not affected by anes thetic gases or vapors.
Fast Response Time Pulse oximetry has a fas t response time,
especia lly compared wi th transcutaneous
measurements (156).
Noninvasive Pulse oximetry is non invasive , which al lows it to
be used as a routine moni tor. I t is
readi ly accepted by awake patients, so it can be applied before
induct ion of
anesthesia. The bleeding, arterial insuffic iency, emboliza
tion, and in fection
sometimes seen after a rterial punc ture are avoided. Temporary
elevat ion of the
PaO2 induced by pain and apprehension is avoided.
Continuous Measurements Saturation , pulse rate , and blood f
low are continuously moni to red. Develop ing
trends can be detec ted and remedial act ion taken before severe
hypoxia ensues.
Separate Respiratory and Circulatory Variables Continuously moni
toring the quality of the peripheral pulse may be helpful in
determin ing whether a hypotensive patient has good cardiac
output. If blood
pressure is low and pulse signal s trength is high, the pat ient
is probably
vasodi lated but perfusing adequately. If , however, both blood
pressure and pu lse
strength are low, perfusion may be inadequate.
Perfusion is indica ted by the pulse s ignal s trength, and
oxygenation is indicated by
sa turat ion. Unl ike trans -cu taneous monitoring, the values
disp layed do not require
inte rpretat ion . Most oximeters wi l l signal if the flow is
not adequate to prov ide a
-
saturat ion value. Th is is helpful in determining a truly low
sa turation value as
opposed to one caused by low f low.
Convenience The probe is simple and fast to apply. Si te prepara
tion is minimal . Arte rial ization
of the sk in is not usually not necessary, except when the
earlobe is the moni toring
s i te. No cal ibra tion or changing of electrolyte or membrane
is required . A variety of
different p robes are available for d if fe rent si te app lica
tions .
Fast Start Time There is minimal delay in ob taining the oxygen
saturation . Readout typically begins
wi th in a few beats after applica tion of the probe. This is a
distinct advantage over
transcutaneous monitoring , which requires a prolonged warm-up
time.
Tone Modulation Changes in pulse tone wi th varying saturat ion
a llow the user to be continuously
updated on SpO2 wi thout tak ing his or her eyes off the pat
ient. Tone modulation
al lows a much quicker recogn ition of hypoxic episodes than
does a f ixed tone (40).
Most anes thesia providers can detect the direction (but not the
magni tude) of a
change in satura tion by l is tening to the change in pi tch of
a pulse ox imeter tone
(303).
User-friendliness Most instruments are user-f riendly. Min imal
training is required to learn to operate
the instrument.
Light Weight and Compactness The conso le can be made
lightweight and compact. This fac i li ta tes use during
transport. Hand-held pulse oximeters are avai lable . Oxygen
saturation monitoring is
availab le in nearly al l physiologic moni tors .
P.789
Probe Variety The wide variety of probe configurat ions confers
broad cl in ical applicabil i ty to all
types of patients , inc luding preterm infan ts. The abil ity to
use various vascular
-
beds offers advantages f rom the standpoin t of access during
surgery and avoids
disturbing the surgica l field.
No Heating Required Heating the sk in is no t required . The
probe can usual ly be lef t in place for extended
periods wi thout r isk of thermal inju ry.
Battery Operated Most stand-alone uni ts and those incorporated
in to transport mon itors can be
operated on batteries .
Economy The use of pulse oximetry can save money by l imiting
oxygen administrat ion to
s i tuat ions where i t is really needed and by decreas ing the
number of blood gas
analyses (178,194,304,305,306). I t may be cost-effect ive to
moni to r certa in pat ients
at high risk for transfer to the intens ive care un it (307).
The use of monitors with
superior art ifac t f i l tering abi li ty may result in cost
savings (308,309,310).
Limitations and Disadvantages
Failure to Determine the Oxygen Saturation There is a small but
defini te incidence of failu re wi th pu lse oximetry
(140,311,312,313,314,315). Factors that are reported to
contribute to higher fai lure
rates inc lude ASA phys ical s tatus 3, 4, or 5 patients; young
and elderly pat ien ts ;
orthoped ic, vascular, and cardiac surgery; electrosurgery use;
hypothermia ;
hypotension ; hypertension; durat ion of in traoperative
procedure ; chronic renal
failure; low hematoc ri t; and motion
(140,312,313,314,316,317,318,319). The actual
failure rate varies with the monitor (320).
A pulse oximeter may zero out, meaning that i t displays 00 for
the SpO2 and pulse
rates values when i t fai ls to produce a measurement or it
might display ______ fo r
the values (321). Some pu lse oximeters blank the display or g
ive a message such
as Low Qual ity Signal or Inadequate Signal . Others freeze the
disp lay.
Poor Function with Poor Perfusion Pulse oximeters require
adequate pulsat ions to dis tingu ish l ight absorbed f rom
arte rial blood f rom venous b lood and tissue l ight (322).
Readings may be unreliable
or unavailable if there is loss or diminution of the peripheral
pulse (proximal blood
pressure cuff inf lation, external pressure, improper posi
tioning, hypotension,
-
hypothermia, Raynaud's phenomenon, cardiopu lmonary bypass, low
cardiac output,
hypovolemia, peripheral vascular disease, a Valsalva maneuver
such as seen in
laboring pat ients or in those wi th infusion of vasoactive
drugs
(78,140,251,287,319,322,323,324,325,326,327,328,329,330,331,332,333).
Methods to improve the s ignal inc lude applying vasodilating
cream, performing
sympathet ic and dig ital nerve blocks, administe ring
intra-arteria l vasodi la tors, and
warming cool extremities (49,51,52,83,334,335,336,337). The use
of a probe on a
better perfused si te such as the cheek, tongue, nasal septum,
or esophagus may be
helpful . Improved signa l techno logy by newer pulse oximeters
can improve
performance during low-perfusion condit ions
(320,330,338,339,340,341).
Difficulty in Detecting High Oxygen Partial Pressures At PaO2
values above 90 mm Hg, small changes in saturat ion are assoc iated
wi th
rela tively large changes in PaO2. Thus, i t has l imited abil
ity to dist inguish high but
safe levels of a rterial oxygen from excessively elevated levels
(342).
Delayed Hypoxic Event Detection While the pulse oximeter
response t ime is general ly fas t, there may be a significant
delay between a change in alveolar oxygen tension and a change
in the oximeter
reading. It is possible for arte rial oxygen to reach dangerous
levels before the
pulse oximeter ala rm is ac tivated (343). Sett ing the low SpO2
alarm threshold
higher wil l decrease the delay.
Delayed response can be re la ted to probe locat ion (62,85).
Desaturation is
detec ted earl ier when the probe is p laced more centrally. Lag
time wil l be increased
wi th poor perfusion (251,344,345). Venous obstruc tion,
periphera l vasoconstric tion,
co ld, and motion art ifacts wi l l cause increases in the t ime
to detect hypoxemia
(32,325,327).
The algori thms that are used to prevent false alarms may
increase the delay in
detec ting hypoxic events (346). A pulse ox imeter may respond
to a noisy or weak
s igna l by s imply holding on to an old value (321). Increasing
the time over which
the pulse s ignals are averaged also inc reases the delay t
ime.
Erratic Performance with Dysrhythmias I rregular heart rhythms
can cause the pulse oximeter to perform errat ical ly (347).
During aortic balloon pulsat ion, the augmentation of diasto lic
pressure exceeds
P.790
-
that of sys tol ic pressure. This leads to a double- or
triple-peaked arterial pressure
waveform that confuses the pulse oximeter, so it may not prov
ide a reading
(348,349). Pulse oximetry works in pa tients who have had an
aortomyoplas ty (105).
Inaccuracy Different Hemoglobins Most pulse oximeters are
designed to detect only two species of hemogobin:
reduced and oxygenated. Whole blood of ten contains other moiet
ies such as
carboxyhemoglobin, sulfhemoglobin, and methemoglobin. Th is dis
turbs the
absorbance rat io of the wavelengths used to de termine oxygen
saturation (350).
Methemoglobin Normally less than 1% of the to tal hemoglob in ,
methemoglobin (metHb) is an
ox ida tion produc t of hemoglobin that forms a reversible
complex wi th oxygen and
impairs the unloading of oxygen to t issues (351,352).
Methemoglobinemia can be
congen ital (353) or acquired . Drugs causing methemoglobinemia
inc lude
ni trobenzene (354), benzocaine (355,356), pri locaine
(357,358,359), and dapsone
(352,360,361). Methemoglobin absorbs l ight equal ly a t the red
and inf ra red
wavelengths that are used by most pu lse oximeters. When
compared wi th func tional
sa turat ion, most pulse oximeters give falsely low read ings fo
r saturations above
85% and falsely high values for saturations below 85%
(351,362,363,364,365,366).
The disc repancy between SpO2 and functional sa turation inc
reases as the level of
metHb increases and functional hemoglobin satura tion dec reases
(351). With
trea tment of the methemoglobinemia, the SpO2 readings become
more accurate
(357,358,361,365).
I f there are conf l ic ting resul ts between the pulse oximeter
and arte rial blood gas
analys is, methemoglobinemia should be suspec ted, and the
diagnosis should be
confi rmed by mult iwavelength co-oximetry. The s tandard blood
gas analys is is no t
capable of de tecting and measuring metHb (367).
A new pulse ox imeter capab le of measuring metHb as we ll as
carboxyhemoglobin is
now avai lable (Fig . 24.18).
Carboxyhemoglobin Carboxyhemoglobin (HbCO, COHb), formed when
hemoglobin is exposed to carbon
monoxide (CO), has an absorpt ion spectrum s imila r to that of
oxyhemoglob in , so
most pulse oximeters wil l over-read SpO2 by the percentage of
carboxyhemoglobin
-
present (18,368,369,370,371,372,373,374,375). In one study, the
pulse oximeter
reading did not go below 96% with carboxyhemoglob in levels as
high as 44% (376).
In v itro CO-oximetry can measure the percentages o f other
moiet ies by using more
than two wavelengths .
View Figure
Figure 24.18 Pulse oximetercarbon monoxide monitor.
An increase in HbCO may occur during laser surgery in the ai
rway, bu t the levels
are not high enough to keep pu lse oximetry from rel iably est
imating saturat ion
(377). Carbon monoxide production in assoc ia tion wi th dry
carbon dioxide
absorbent is discussed in Chapter 9.
Pulse oximeters that different ia te between oxyhemoglobin and
carboxyhemoglobin
and that can measure carboxyhemoglobin are now avai lable (Fig.
24.18).
Fetal Hemoglobin Most studies show that fetal hemoglobin (Hb F)
does not appear to affect the
accuracy of pulse oximetry to a c linically important degree
(378,379,380,381,383),
al though very high levels may cause i t to read sl ight ly low
(382).
Hemoglobin S The use of pulse oximetry in the pat ient wi th s
ick le cell disease is controvers ial .
Severa l investigators have concluded that pulse oximetry is
inaccurate in these
patients, wh ich makes i t unreliable for de tec ting serious
hypoxemia (383,384,385).
-
Other studies have found i t to be suff icient ly accurate to be
useful
(386,387,388,389,390).
Sulfhemoglobin Sulfhemoglob inemia may be caused by drugs such
as metoclopramide, phenacetin,
dapsone,
P.791
and sulfonamides (391). Sulfhemoglobin causes the pulse oximeter
to d isplay
arti factual ly low oxygen sa turation.
Other Hemoglobinopathies Hemoglobin Koln is associated with art
ifac tually low oxygen satura tion as measured
by the pulse oximeter (392,393).
Hemoglobin Hammersmith and Hemoglob in M (Mi lwaukee) affec t
pulse oximeter
reading so much that ox imetry is not usefu l (394,395,396).
Hemoglobin-H disease wil l cause the pulse ox imeter to indicate
a higher saturat ion
than is actual ly present (397).
A pat ient wi th hemoglobin Constant Springs and
alpha-thalassaemia 2 in which
pulse oximetry readings were cons isten tly low has been
reported (398).
Heinz Body hemolyt ic anemia causes the pulse oximeter to read
low (399).
Bilirubinemia Severe hyperbil i rubinemia can cause an
artifactual elevation of metHb and
carboxyhemoglobin when using in v itro oximetry but does not
affec t pulse ox imetry
readings (283,378,400,401,402,403).
Low Saturations Pulse oximetry becomes less accurate at low
oxygen saturat ions
(10,26,57,58,61,84,284,285,288,291,293,294,297,298,306,404,405,406,407,408,40
9,410,411,412,413,414). This inaccuracy is greater in patien ts
wi th dark sk in (415).
I t should be used wi th caution in pat ien ts wi th cyanotic
heart disease
(410,411,416). Measuring PaO2 or SaO2 at low saturations is
recommended fo r
importan t c linical decis ions.
Malpositioned Probe Oximeters wi th probes that are not applied
we ll vary greatly in their behavior,
depending on both the actual SpO2 and the manufac turer and
model of the oximeter
(70,417,418,419,420,421,422). If the probe is not properly posi
t ioned, i t may al low
-
the l ight from the emitter to the detector to only graze the t
issue instead of passing
through i t. Th is penumbra effec t reduces the s ignal-to -no
ise rat io and may result in
spurious SpO2 values in the low 90s in normal pa tients. I f the
pat ient is hypoxic ,
the oximeter may overestimate the true value (418,423). In one
case, the probe was
comple tely unattached but cont inued to prov ide apparent ly
accurate readings.
Closer examinat ion of the waveform revealed an unusual pattern
(424).
To avoid the problems of probe posi tion, the posi tion should
be checked frequently
and inaccessible loca tions avoided. The use of too large or too
small a probe may
resul t in inaccurate readings (33,34,300). Long f ingernai ls
can cause inaccura te
posit ioning (425).
Venous Pulsations Pulse oximeter design assumes that the pulsa
ti le components of l igh t absorbance
are due to arterial b lood. Prominent pulsations of venous b
lood may lead to
underest imation of the SpO2 (70,426,427,428,429,429A). Pulse
rate determination
may be correct. The error may be worse when probes are used on
the head (86,87)
but less when the probe is placed on the f inger. In pat ien ts
wi th low sys temic
vascular res istance, the pulse ox imeter may under-read the
saturat ion , possibly
because the ox imeter is sensing pulsat i le venous f low
(430).
High ai rway pressures during art if icial venti lat ion may
cause phasic venous
conges tion, which may be in terpreted by the oximeter as a
pulse wave (431). In
some cases , it may be necessary to turn the vent ilator OFF to
obta in a correct
reading.
Mixing Probes SpO2 measurements may not be accura te if one
manufacturer's p robe is used with
a dif ferent manufacturer's instrument (432,433).
Severe Anemia The pulse oximeter may overes timate SpO2,
especially at low saturations, in
patients with severe anemia (434,435,436,437). However, i t is
accurate for non-
hypoxic SaO2 values in these patien ts (438).
Skin Pigmentation Although some earl ier s tudies have shown
that pulse oximeter readings were
s light ly high in pat ients wi th dark sk in (286,439,440),
newer studies have shown
that pigmenta tion does not make a s ign if icant dif fe rence
in pulse oximeter accuracy
-
(26,441,442). Spurious readings were reported in a pat ient wi
th oculocutaneous
albinism who was taking a herbal remedy (443).
Dyes Certain dyes inc luding methylene blue, indocyanine green,
lymphazurin ( isosulfan
blue), indigo carmine, ni trobenzene, and patent blue when injec
ted intravenously,
intra-arteria lly, into the lymphatics , intradermal ly, or into
the uterine cavi ty can
resul t in decreases in SpO2 without actual decreased
saturation
(354,360,444,445,446,447,448,449,450,451,452,453,454). In v i
tro oximetry may
also be affected by dyes (360,445,455,456,457). Usual ly, the in
te rference lasts only
a few minutes but may persist much longer, even hours, when
lymphatics are
injected (451,452).
The reac tion of the pulse oximeter to exogenous dyes has been
used as a means of
confi rming intravascular catheter placement. The dye is
injected into the catheter,
and the pulse oximeter is observed (458). The pu lse oximeter
may be useful to
es timate card iac output by the dye di lut ion method
(459).
P.792
Fingerprin ting ink wil l cause a low satura tion reading (460).
Henna, a stain used by
some Middle Eastern women on the fingers and toes , can cause a
low satura tion
reading (461). Chi ld ren who have been f inger paint ing wi th
b lue pa in ts may exhibit
low SpO2 readings (462).
Optical Interference Stray l ight or l ight f l ickering a t
frequencies s imilar to the frequencies of the LEDs,
including sunligh t, f luorescent l ights , operat ing room
lights , infrared heat ing lamps,
infrared radiant warmers, l ight sources for various scopes,
xenon lamps , bi l i rubin
l ights , phototherapy, or surgical imaging instruments, can
enter the photodetector
and resu lt in inaccura te or e rratic readings
(14,332,463,464,465,466,467,468,469,470,471,472,473,474). Having
probes on two
adjacent f ingers can cause an abnormal trace (475).
One clue that optica l in terference is occurring is inconsis
tency between the pulse
rate on the pulse oximeter and tha t on other mon itors (350). A
lthough excessive
ambient l ight usua lly prevents the oximeter f rom track ing
the pulse , it can resul t in
apparent ly normal but inaccurate measurements in some ins
tances (470).
-
Oximeters vary s ignif icantly in their susceptibil i ty to opt
ical inte rference (326,466).
Some manufacturers try to minimize the effect of s tray l igh t
by tak ing in termittent
readings when both of the LEDs in the probe are turned OFF and
then subtrac ting
these background readings from measurements taken by the
photodetector when
ei ther LED is turned ON. Sensi tiv i ty to l igh t may be
increased with reduced pu lse
ampli tude.
There are a number of ways to minimize the e ffects of op tical
interfe rence. These
include selection of the correct p robe for the patient and use,
applying the probe so
that the detector is across f rom the LEDs , making certain the
probe remains
properly posi t ioned, and shielding the probe f rom l ight and
o ther nearby probes
(Fig . 24.12). Extraneous l ight can be eliminated by covering
the probe wi th an
opaque material such as a surg ical towel , gauze, f inger cot,
blanket, alcohol wipe
packet, or o ther foil shield (59,127,476). This may also help
to stabil ize the probe.
All l igh t may not be adequately shielded by a s imple covering
(463,470).
Nail Polish and Coverings Some shades o f brown, black, blue,
and green (but not red or purple ) nai l po lish
may cause s ignif icant ly lower satura tion readings
(477,478,479). Synthet ic nai ls
may interfere wi th pulse oximetry readings . The presence of
onychomycosis, a
ye llowish gray color caused by fungus, can cause fa lsely low
SpO2 read ings (480).
Dirt under the na il can also cause diff icul ty in obta in ing
rel iable readings (481).
Al though there is one report of dried blood on a f inger that
caused erroneous low
sa turat ion read ings (482), other authors have found tha t
dried b lood does not affec t
pulse oximeter accuracy (483,484). Pat ients rece iv ing
docetaxe l may have
discolored f ingernai ls that cause lower saturation readings
(485).
In most cases, this problem can be overcome (wi thout removing
the pol ish or the
synthetic na il ) by turn ing the probe 90 degrees so that i t
transmits l ight from one
s ide of the f inger to the other s ide (55,481).
Electrical Interference Electrica l in terference from an
electrosurgical unit can cause the oximeter to give
an incorrect pulse count (usual ly by counting extra beats) or
to falsely register a
decrease in oxygen satura tion (486). This problem may be inc
reased in pat ients
wi th weak pulse s ignals (13). Manufacturers have made s ign if
icant p rogress in
reduc ing their instruments ' sensi tiv i ty to electrica l in
terference (13,207,243,326).
Some monitors d isplay a notice when s ignif icant in terference
is p resent. Some
f reeze the SpO2 display during such interference, which may
give a false sense of
-
security. Addi tional s teps to minimize e lectrical
interference inc lude locating the
electrosurgery grounding plate as close to, and the ox imeter
probe as far f rom, the
surgical field as possible ; routing the cable from the probe to
the oximeter away
f rom the electrosurgery apparatus; keeping the pulse oximeter
probe and console
as far as possible from the surgical s i te and the
electrosurgery grounding plate and
table; and operating the uni t in a rapid response mode. The
electrosurgical
apparatus and pulse ox imeter should not be p lugged into the
same power c ircui t
(486).
Motion Artifacts Motion o f the probe can cause an arti fact
that the pulse ox imeter is unable to
differentia te f rom normal a rterial pulsat ions. Motion
artifact creates both fa lse-
posit ive (false alarm) and fa lse-negative (missed hypoxemia)
errors (32,487).
Changing alarm thresholds to reduce one of these errors wil l of
ten inc rease the
incidence of the other type of e rror (488).
Motion is usual ly no t a problem during genera l anesthesia,
but i f the patien t is
sh ivering, has a condit ion such as Park inson's disease, o r
is mov ing about o r being
transported, motion art ifac ts can be s ignif icant (489,490).
Evoked potential
moni tors and nerve stimulators can produce motion artifacts if
the pulse oximeter
probe is on the same extremity (491,492,493,494,495). Motion
arti facts have been
caused by patients tapping their fingers whi le under regional
anesthesia (496).
The oximeter's abi l i ty to deal with motion art ifact depends
on the correlat ion with
the onset of the motion and the s tart of monitoring. If the
motion precedes the
onset of moni toring, there is a greater decrement in
performance (497,498).
In the 1990s, pulse oximeter manufac turers began to make design
improvements,
and the newer genera tion
P.793
instruments have improved abi l ity to f i l ter motion art
ifacts
(320,321,338,341,488,497,498,499,500,501,502,503,504,505). In
addi tion to
reduc ing alarms, the use of these instruments has resul ted in
less need for arte rial
blood gas measurements and fas ter weaning f rom high concentra
tions of oxygen
(309). They are also associated wi th a shorter loss of s ignal
when placed d istal to a
blood pressure cuff or tourniquet (506).
Lengthening the averaging t ime wi l l increase the l ikel ihood
that enough true pulses
wi l l be detected to reject motion artifacts (15,507) but may
delay detection of
-
hypoxemia. Most pulse oximeters al low the user to select one of
several t ime-
averaging modes.
Motion art ifac ts can usual ly be recognized by false or e rrat
ic pu lse rate d isplays or
dis torted plethysmographic waveforms. Inc reased pulse ampli
tude indicates
movement but not necessari ly arti factual SpO2 readings
(487,508).
Motion art ifac ts may be decreased by applying the probe to a
less active si te.
Flexible probes that a re taped in p lace are less suscept ible
to motion art i facts than
are c lip-on probes (13,15). Larger f ingers may be less
susceptible to motion art ifac t
(32).
Pressure on the Probe Pressure on the probe may resul t in
inaccurate SpO2 readings wi thout affect ing
pulse rate determinat ion (429).
Hyperemia I f a l imb becomes hyperemic af te r blood f low is
interrup ted, the oxygen satura tion
shown by the pulse oximetry may be arti f ic ia lly low (509). A
pulse ox imeter placed
near the s ite of blood transfusion may show transien t
decreases in oxygen
sa turat ion wi th rapid blood infusion (510).
Probe Damage A damaged pulse oximeter probe can cause the oxygen
saturat ion to be higher than
the actual value (511,512). The use of a cleaning agent that is
not recommended by
the manufacturer on a reusable probe can resul t in damage to
the probe, prevent ing
i ts reuse (513).
False Alarms A high percentage of pulse oximetry alarms are
spurious or triv ial
(489,503,505,514,515,516,517). Art ifac t-induced ala rms occur
in two ways. When
an artifact is mis taken fo r a pulse, i t can corrup t the
measurement and resul t in an
alarm. When an artifact obscures the pulse, it can resul t in a
loss -of -pulse alarm.
False alarms are mos t commonly caused by motion artifact bu t
are also associa ted
wi th poor s ignal qua li ty , probe displacement, external p
ressure, and interfe rence.
False alarms are a more s ignif icant problem outs ide the opera
ting room because
patients are commonly moving, they are of ten poorly perfused,
and there are many
sources of elec tronic and opt ical interference.
False alarms do not represent a di rec t danger to the patient
but may encourage the
care provider to take inappropriate ac tions such as disabling
the alarm, setting the
-
l imi ts to inappropria te values, or lowering the ala rm
volume. Misinterpretat ion of
alarms can resul t in failu re to treat hypoxemia or unnecessary
treatment.
Some false alarms can be avoided by s imple measures such as
putt ing the probe
on a different extremity than the automated blood pressure cuff
and in a locat ion
where i t is un likely to be affected by ex terna l
pressure.
Newer pulse oximeters tha t are designed to reduce
motion-related art i fac ts can
s ignif icant ly reduce the inc idence of false alarms
(320,497,498,499,503,518,519,520,521,522,523,524,525,526).
However, some of
these have shown less rel iabi l i ty in ident ifying hypoxic
episodes and bradycardia
than older models (346).
Delaying the time between detecting low SpO2 and alarm activa
tion, us ing a longer
averaging t ime, and sett ing the low SpO2 alarm l imit lower
can reduce the number
of false alarms (311,517,527,528,529) but may increase the lag t
ime before
detec ting hypoxemia. With some pulse oximeters , tu rning OFF
the low pu lse rate
alarm prevents a la rming when the blood pressure cuff is inf
lated (13).
Synchronizing the pulse oximeter wi th the electrocardiogram
(ECG) moni tor can
lessen art ifac ts (32,39). However, the oximeter may
synchronize with ECG art ifac ts
generated by motion or shivering , resul ting in e rroneous
readings (13).
Furthermore, with this system, the pulse rate displayed by the
oximeter wil l
necessari ly be equa l to the pulse rate shown by the ECG
monitor, so equal ity of the
pulse rates cannot be used as an indicat ion tha t the d
isplayed sa turation data are
valid.
Failure to Detect Impaired Circulation The presence of a pulse
oximeter signa l and a normal reading does not necessari ly
imply that tissue perfus ion is adequate. Some pulse oximeters
show pulses despi te
inadequate tissue perfusion (251,327) or even when no pulse is
present
(200,530,531,532). Ambient l ight may produce a fa lse s ignal
(471).
Pulse oximetry is not rel iable in diagnosing impaired perfusion
wi th increased
intracompartmental pressures (533,534).
Discrepancies between Readings from Different
Monitors A discrepancy in readings between different b rands of
ox imeters on the same
patient a t the same time is no t uncommon
(17,270,535,536,537,538). There is a lso
variation in
-
P.794
the time tha t it takes various moni tors to detect resaturation
(539).
Failure to Detect Hypoventilation Hypoventi lation and
hypercarbia may occur wi thout a dec rease in hemoglobin
oxygen saturat ion, espec ial ly if the pat ient is receiv ing
supplemental oxygen
(540,541,542). Pulse oximetry cannot be relied on to detect
leaks, disconnect ions
(543), or esophagea l in tubation (544). Methods to detect
esophageal intubat ion are
discussed in Chapter 19.
Problems with Sound Recognition There is cons iderable variat
ion in the volume and aud io spectrum of available pulse
ox imeters between models and even wi th in mode ls (545). Some
c linic ians have
trouble detecting changes in the pi tch o f the sound emitted by
pulse oximeters as
the saturation changes (303,546).
There is wide variat ion in the pitch f requency of di fferent
pulse oximeters (547).
Thus, in locat ions where d if ferent pulse ox imeters are
encountered, the potential
for confusion exists.
Lack of User Knowledge Pulse oximetry is of ten used by
personnel whose knowledge of it is l imi ted.
Physicians, nurses, and others who use the instrument of ten do
not know the bas ic
principles and make serious errors in interpreting readings
(548,549,550,551).
Interference with Other Monitors Electromagnetic in terfe rence
from the pu lse oximeter power supp ly may cause
arti facts and fa lse readings on certa in thoracic impedance
moni tors (552).
I f a pulse oximeter probe is placed in front of certain plasma
display touch screens ,
a normal-appearing waveform and 100% saturation is displayed
(553).
Patient Complications
Corneal Abrasions Patients recovering from general anesthes ia f
requently rub their eyes. If there is a
pulse oximeter on the index f inger, a cornea l abrasion may
resul t (65,554,555). A
f inger other than the index f inger may be a more appropriate
loca tion fo r the probe
during recovery (556).
-
Pressure and Ischemic Injuries Inju ries ranging from pers
istent numbness to ischemic injury at the s ite on which a
probe was placed have been reported
(557,558,559,560,561,562,563,564,565,566,567,568). Loss of the s
ignal may occur
(569). These risks are increased by pro longed probe applica
tion, compromised
perfusion of the extremity, and tight appl icat ion of the
probe. Frequent examination
of the s i te and moving the probe to different s ites wi l l
reduce the l ikel ihood of
injury. Pat ients wi th large f ingers should not have a c i
rcumferent ial probe placed on
the f inger. I f the pulse oximeter reading appears to be weak,
the s i te should be
checked for inc reased pressure .
Burns Inju ries ranging from reddened areas to third-degree
burns under pulse oximeter
probes have been reported
(562,570,571,572,573,574,575,576,577,578,579,580,581).
Considering the mil l ions
of long-term applica tions , the inc idence of these burns is
quite low (43).
Burns can result f rom incompatib il i ty between the probe f
rom one manufacturer with
the pulse oximeter of another (575,578). A number o f pulse
oximeter probes have
connec tors that f it d if fe ren t pulse ox imeters, bu t the
probes are not compatible. The
use of a damaged probe can resul t in a burn (570,577). A pu lse
oximeter probe may
provide an al ternate pa thway for electrosurg ical currents
(582).
Burns that are associated with hypothermia have been reported
(579). Burns have
been reported when a pu lse oximeter was used during
photodynamic therapy
(583,584).
To avoid these injuries, f requent inspect ion of the probe s i
te and s ite ro tat ion are
recommended (585). When a probe is placed on a finger or toe,
the l igh t source
should be placed on the nai l rather than on the pulp (576). A
glove can be placed
on the finger to protect i t f rom thermal inju ry wi thout
affect ing the accuracy of the
instrument (36). If the pu lse oximeter display freezes, the
cause should be
investigated. Only the probes recommended by the ox imeter
manufacturer should
be used (578).
Burns that are associated with pulse oximetry during magnetic
resonance imaging
(MRI) as a resul t of induced skin current beneath looped cables
ac ting as antennae
have been reported (586,587,588). The MRI envi ronment is
discussed in Chapter
30. During MRI, the danger of burns can be reduced by the fol
lowing measures:
-
All potential conduc tors should be checked before use to ensure
that there is
no f rayed insulat ion, exposed wires, or other hazards.
All unnecessary conduct ive materials such as unused surface
coils should be
removed f rom the MRI system bore before pat ient moni toring is
ini t iated.
P.795
The probe should be placed as far f rom the imaging s i te as
possible.
Cables , leads , or wires f rom monitoring devices should be
posit ioned so that
no loops are fo rmed. A braid should be made of the slack port
ion of wires .
I f possible , no potential conduc tors should touch the patient
a t more than
one location.
A thick layer of thermal insu lat ion should be p laced between
any wires or
cables and the patien t's sk in.
Monitoring devices that do not appear to be operat ing properly
should be
removed f rom the pat ient.
Electric Shock An electrica l shock related to dia thermy has
been reported (589). In this case,
there were bare wires in the pu lse oximeter probe.
Carbon Monoxide Monitoring Carbon monoxide can accumulate in the
breathing system from various sources,
including the react ion between anes thetic agents and
desiccated absorbent. This is
discussed in detai l in Chapter 9. It is diff icul t to de
termine if CO is present in the
inspired gases when the pat ient is anesthetized.
In 2005, a combinat ion pulse CO-oximeter capable o f
measuring
carboxyhemoglobin (SpCO) became avai lable (Fig. 24 .18). The
same sensor is
used for measuring both SpCO and SpO2 . It ala rms wi th CO
concentra tions
between 5% and 50%. The instrument uti l izes an e ight-wave
length sensor to
dis t inguish between oxygenated blood, deoxygenated blood, and
blood containing
CO. I t can a lso measure metHb.
References
-
1. Tschupp A, Fanconi S. A combined ear probe for pulse ox
imetry and carbon
dioxide tens ion monitoring : accuracy in c ri tica lly i l l
children. Anesth Analg
2003;96:8284.
[Fu ll text Link]
[CrossRef]
[Med line Link]
2. Rohl ing R, Biro P. Clinica l investigation of a new combined
pulse oximetry and
carbon d ioxide tension probe in adult anaesthesia. J Cl in
Monit Comput
1999;15:2327.
[CrossRef]
[Med line Link]
3. Hayoz J, Rohl ing S , Tschupp A. World's f i rs t combined
digi ta l pulse oximetry
pulse oximetry and carbon d iox ide tens ion ear sensor. Anesth
Analg
2002;94:S103.
4. Bendjel id K, Schutz M, Stotz M, e t al . Transcutaneous PCO2
moni to ring in
crit ically i l l adu lts : c linical evaluat ion of a new
sensor. Crit Care Med
2005;33:22032206.
[Fu ll text Link]
[CrossRef]
[Med line Link]
5. Stoe lt ing RK. APSF president presents state of foundation
report. APSF Newslett
20052006;20:62,67.
6. Gravenstein JS. Inte rnational s tandards for safe practice
endorsed by WFSA.
APSF News le tt 1992;7:2931.
7. T inker JH, Dul l DL, Cap lan RA, et al . Role of mon itoring
devices in prevention of
anesthet ic mishaps. A c losed c la ims analysis. Anesthes
iology 1989;71:541546.
[Fu ll text Link]
[CrossRef]
[Med line Link]
8. Runciman WB, Webb RK, Barker L, et al. The pulse ox imeter:
app licat ions and
l imi ta tions . An analys is of 2000 inc ident reports .
Anaesth Intens Care 1993;21:543
550.
[Med line Link]
9. Scuderi PE, Bowton DL, Anderson RL, e t al . Pulse oximetry:
would fu rther
technical al te rations improve pat ient ou tcome? Anes th Analg
1992;74:177180.
10. Tremper KK, Barker SJ. Pulse oximetry. Anesthesio logy
1989;70:98108.
-
[Fu ll text Link]
[CrossRef]
[Med line Link]
11. Wuki tsch MW, Petterson MT, Tobler DR, et al. Pulse
oximetry. Analys is of
theory, technology, and pract ice. J Cl in Mon it
1988;4:290300.
[CrossRef]
[Med line Link]
12. Kel leher JF. Pulse oximetry. J Clin Monit 1989;5:3762.
[CrossRef]
[Med line Link]
13. Alexander CM, Teller LE, Gross JB. Princ iples of pulse
oximetry. Theoretica l
and prac tical considera tions . Anes th Ana lg
1989;68:368376.
[Fu ll text Link]
[CrossRef]
[Med line Link]
14. Anonymous . Pulse oximeters. Health Devices
1989;18:185230.
[Med line Link]
15. Hamlin MD. A guide to pulse oximetry moni toring and
troubleshooting . J Clin
Eng 1995;20:476483.
16. Schnapp LM, Cohen NH. Pulse oximetry. Uses and abuses.
Chest
1990;98:12441250.
[CrossRef]
[Med line Link]
17. Pologe JA. Funct ional saturation versus fract ional
saturation. What does pulse
ox imetry read. J Clin Moni t 1989;5 :298299.
18. Reynolds KJ, Palayiwa E, Moyle JTB, et a l. The effec t of
dyshemoglobins on
pulse oximetry. Part I : theoretical approach. Part I I:
experimental results using an
in v itro test system. J Clin Monit 1993;9:8190.
[CrossRef]
[Med line Link]
19. Ingal ls L, Fogarty C, Fasnacht R, et al . Digi tal
pass-through pu lse oximeter
probes are not accurate or precise when used as a ref lec tance
dev ice. Respir Care
2000;25:986.
20. Konig V, Huch R, Huch A. Ref lectance pulse
oximetryprinciples and obstetric
applica tion in the Zurich sys tem. J Clin Moni t
1998;14:403412.
-
21. Dassel ACM, Graaff R, Sikkema M, et al. Ref lectance pu lse
oximetry at the
forehead improves by pressure on the probe. J Clin Moni t
1995;11:237244.
[CrossRef]
[Med line Link]
22. Nij land R, Jongsma HW, van den Berg PP, et a l. The effec t
of pu lsating arte ries
on reflec tance pulse ox imetry: measurements in adu lts and
neonates. J C lin Moni t
1995;11:118122.
[CrossRef]
[Med line Link]
23. Nij land R, Jongsma HW, Verbruggen IM, et al. Ref lec tance
pulse ox imetry in
fetal lambs: subcutaneous vessels and vasoconstric tion affect i
ts rel iabil i ty . J Clin
Moni t 1996;12:225230.
[CrossRef]
[Med line Link]
24. Thi lo EH, Park -Moore B, Berman ER, e t al . Oxygen
saturation by pu lse oximetry
in heal thy infants at an a lt itude of 1610 m (5280 f t). AJDC
1991;145:11371140.
[Med line Link]
25. Bel l C, Lu ther MA, Nicholson JJ, et al. Effect of probe
design on accuracy and
reliabil i ty of pulse oximetry in pediatric patien ts . J Clin
Anes th 1999;11:323327.
[CrossRef]
[Med line Link]
26. Avant MG, Lowe N, Torres A. Comparison of accuracy and
signal consistency of
two reusable pulse ox imeter probes in cri tical ly i l l
children. Respir Care
1997;42:698704.
27. Mikawa K, Maekawa N. A s imple al ternate technique fo r the
appl ication of the
pulse oximeter probe to infants . Anaesthesia 1992;77:400.
28. Hess D, Kochansky M, Hassett L, et al . An evaluation of the
Nel lcor N-10
portable pulse oximeter. Respir Care 1986;31:796802.
29. Joshi GP. Disposable versus reusable oximeter probes : a cos
t-benef i t ana lys is.
Anesth Analg 1998;86:S38.
30. Anonymous . Hospitals look for savings in pulse ox imetry
probes. Technol
Anesth 1992;13(5):12.
31. Maruschak GF, Johnson RM. Pulse oximeter per usesecuring
savings .
Anesthesiology 1989;71:167168.
[Fu ll text Link]
[CrossRef]
-
[Med line Link]
32. Langton JA, Hann ing CD. Effect of motion artefact on pulse
oximeters.
Evaluation of four ins truments and f inger p robes. Br J
Anaesth 1990;65:564570.
[CrossRef]
[Med line Link]
33. Anonymous . Oximeter probes. Technol Anesth 1994;15:14.
34. Zahka KG, Dean MJ. Failure of pu lse oximetry to de tec t
severe hypox ia:
importance of probe select ion . Clin Ped iatr
1988;27:403404.
35. Yoder RD. Preservation of pulse oximetry probes. Anesthesio
logy
1988;68:308.
[Fu ll text Link]
[CrossRef]
36. Ackerman WE, Juneja MM, Baumann RC, et al . The use of a v
inyl glove does
not affec t pulse ox imeter monitoring. Anesthesiology
1989;70:558559.
[Fu ll text Link]
[CrossRef]
[Med line Link]
37. Bardoczky GI, Engelman E, d'Hol lander AA. One more use of a
v inyl glove:
ready-made protection fo r the reusable pulse oximeter.
Anesthesiology
1989;71:999.
[Fu ll text Link]
[CrossRef]
[Med line Link]
38. Cheung P, Hardman JG, Whiteside R. The effect of a
disposable probe cover on
pulse oximetry. Anaesth In tens Care 2002;30:211214.
[Med line Link]
39. Barrington KJ, F iner NN, Ryan CA. Evalua tion of pu lse
oximetry as a
continuous moni toring techn ique in the neonata l in tensive
care uni t. Cri t Care Med
1988;16(11):11471153.
[CrossRef]
[Med line Link]
40. Craven RM, McIndoe AK. Continuous audi tory moni to ringhow
much
information do we regis ter? Br J Anaesth 1999;83:747749.
41. American Society of Anes thesiologists. Standards for basic
anes thetic
moni toring (http://www.ASAhq.org). Park Ridge, IL: Author,
1999.
-
42. American Society for Testing and Materials . Med ical
electrica l equipment
part icula r requirements for the basic sa fety and essential
performance of pulse
ox imeter equipment for medical use (ASTM 9919). West Con