The clinical evaluation of the Datex Ohmeda GE S/5 Monitor’s Non
Invasive Blood pressure (E-PRESTN) module and the optimal placement
of measurement cuff on the anaesthetized dog.
Drs. R.C. Wienesen
Student-number: 0356972
October 2008 - March 2009
Supervisors Division Anaesthesiology,
Dept Clinical Sciences of Companion Animals, FVM, UU
R. Sap
Drs. L.C. Akkerdaas
Contents
Abstract
3
Introduction
4
Invasive Arterial Blood Pressure measurement
4
Non Invasive methods
5
Goals of this study
7
Materials and Methods
8
Results
10
Discussion
15
Conclusions
17
Clinical relevance
18
Acknowledgements
19
Appendix
20
References
25
2
Abstract
Objective 1 To evaluate the Datex Ohmeda GE S/5 Non Invasive Blood Pressure module in
comparison to the standard invasive arterial blood pressure (IABP) technique in anaesthetized
dogs.
Objective 2 To determine optimal cuff position and site for the noninvasive system on the patient.
Objective 3 To assess the effect of different anesthetic protocols (methadone-isoflurane,
dexmedetomidine-propofol and dexmedetomidine-alfaxalone) on the performance of the Datex
Ohmeda GE S/5.
Study design A prospective study.
Animals. Thirty-two client owned dogs undergoing a variety of surgical procedures.
Methods Three different anesthetic protocols were used. Invasive blood pressure was measured
using either the femoral or dorsal pedal artery. For oscillometric noninvasive blood pressure (ONIBP) measurement, the cuff was placed at the contra lateral limb during the surgical procedure
when possible, otherwise the tail base was used. Recordings were made pre-, per- and post
operative. The tail base and front limb O-NIBP measurements were continued post operatively.
Statistical analysis of the Invasive Arterial Blood Pressure (IABP) and O-NIBP data was performed
using a (modified, percentual) Bland Altman analysis. An ANOVA was used to compare the different
protocols.
Result 1 The Datex Ohmeda GE S/5 O-NIBP produced highly variable results for all blood pressure
measurements. The least variable results came from the Mean Arterial Pressure (MAP) for all
measurement-sites.
Result 2 The most reliable site for O-NIBP measurements is the tail base followed by the hind
limb. The front limb is less accurate.
Result 3 The number of failures in order to obtain a direct reading in the Datex O-NIBP were
higher in the dexmedetomidine protocols. Bias itself did not differ between the protocols.
Conclusion 1 & 2 The most accurate data produced by the Datex O-NIBP was the MAP. The
optimal site for cuff placement was the tail base. However, due to the high variability of all values
it is difficult to determine whether the machine is producing accurate (e.g. unbiased) results. It
should be noted that trend information (multiple consecutive measurements) gathered with the
Datex O-NIBP can be useful. However sudden drops or increases of O-NIBP do not necessarily
indicate that the anesthesia is insufficient, as it could also be accounted for by the variability of the
O-NIBP.
Conclusions 3 Anesthetic protocols do influences the number of failures to provide direct
measurement. This is most likely related to the use of dexmedetomidine (a2-agonist) which causes
vasoconstriction. This vasoconstriction is the likely culprit for the increased failure to provide direct
measurements with the Datex O-NIBP.
Clinical relevance The Datex Ohmeda GE S/5 can not be used on patients that need accurate
blood pressure monitoring because of high anesthetic risks (ASA-III or up). In cases where
placement of the arterial catheter-system for the IABP fails and patients in ASA-I and ASA-II
(where protocols usually don’t include measurement of the arterial blood pressure (ABP)) it can be
used to get information on the arterial blood pressure trend. In these before mentioned situations,
the MAP is shown to be the most reliable.
Keywords anesthesia, arterial blood pressure, dog, non invasive blood pressure, oscillometric,
measurement site.
3
Introduction
The Arterial Blood Pressure (ABP) is the most direct measurable cardiovascular parameter
to provide insight in the tissue perfusion. Directly dependant on the afterload of the heart, it is
commonly used as an indicator for the workload of the heart, and as such the cardiac performance.
At tissue level the ABP, the oscillations between systolic blood blood pressure (SAP) and diastolic
blood pressure (DAP) are dampened [Sharnvet, 2008][Deflandre, C.J.A., 2008][Wagner, A.E., 1997]. This
dampening means that tissue perfusion is largely dependant upon the average pressure in the
cardiovascular system. The determining factor is called the Mean Arterial Blood Pressure or MAP
and is relevant for virtually every tissue in the body.
One exception to the MAP being important in tissue perfusion is the heart itself. During the
systolic phase of the heart action the blood flow is limited because the small vessels are
compressed. As such the tissues get their share of oxygen and nutrients during the diastolic phase.
Because of this fluctuating availability of oxygen in the cardiac muscle, the DAP needs monitoring
as well [Sharnvet, 2008].
Blood pressure measurement (arterial and venous) is rarely preformed in the veterinary
practice. Equipment is expensive and measurement methods are usually time consuming. However
one of the common complications of anesthesia is hypotension (MAP < 70 mmHg), in both human
and veterinary patients [Wagner, A.E., 1997]. As early as in 1897, when Hill and Barnard experimented
with chloroform anesthesia, it was well known chloroform caused a significant drop in ABP. In
human medicine/anesthesia the ABP monitoring is part of basic monitoring and the American
College of Veterinary Anesthesiology published a similar set of guidelines in 1995 [Anonymous.,
(2005)][Birks S.H., (2007)]. These included the monitoring of the ABP for both canine and feline patients
with an ASA status (table 1) of III or higher and for equine patients undergoing general anesthesia.
Table 1: ASA-classification system modified for veterinary use
Classification
Implication for the patient’s status
ASA I
A normal healthy patient
ASA II
A patient with mild systemic disease but no expected functional
consequences
ASA III
A patient with severe systemic disease with clear functional consequence
ASA IV
A patient with severe systemic disease that is a constant threat to it’s life
ASA V
A moribund patient who is not expected to survive without the operation
Invasive Arterial Blood Pressure measurement
The golden standard for blood pressure measurement at the moment is the direct (or
invasive) arterial blood pressure measurement (IABP) and is accurate even in hypo- or
hypertensive patients [Billiet, E., 1998]. It is based upon a fluid filled catheter system which is
connected to a pressure transducer [Deflandre, C.J.A., 2008]. The catheter system is filled with a NaClLow dose Heparin Solution (2500 IE of Heparin in 500 mL NaCl 0.9%). The catheter is placed on
the femoral artery (central), or the dorsal metatarsal artery (peripheral) depending on the patient’s
size and hemorrhage risk. Currently this is the only method capable of providing a continuous
measurement of the arterial blood pressure.
The invasive method has, however, several disadvantages. For example, the placement of
arterial catheters is challenging in small dogs and in feline patients. There is a slight risk for
complications such as arterial hemorrhage, infection, septicaemia and partial or complete occlusion
of the artery. Occlusion (e.g. arterial embolisation) is potentially followed by ischemia of the tissue
distal of the catheter it theoretically could result in loss of a toe or other extremity [Deflandre, C.J.A.,
2008] [Gains, M.J., 1995].
Besides the risks to the patient’s health, the system is susceptible to outside influences.
The catheter is difficult to fixate to the patient, leading to dislocation especially when patients are
moved around. The IABP-system is also sensitive to gravity. If the transducer is not placed at the
correct height (equal to the height of the right atrium in any position) this can lead to under- or
4
overestimation of the ABP. Another factor resulting in over- or underestimation are air bubbles in
the fluid column. These air bubbles create a dampening effect because they are compressed with
the fluctuations in the blood pressure [Deflandre, C.J.A., 2008].
Non Invasive methods
Over the years several methods have been developed to assess the cardiovascular system
functioning and some are very useful in determining the MAP. Palpation of the peripheral pulse for
example does not yield any additional information on the arterial blood pressure other than that
there is a pressure wave traveling through the vascular bed. Similarly, determination of the heart
rate gives no information about the current arterial blood pressure [Chalifoux A., 1985][Wagner, A.E., 1997].
In human medicine the use of a stethoscope and sphygmomanometer with attached cuff
for determining the ABP, was common practice up until the introduction of digital equipment. The
latter method, which is the basis for many of the Non Invasive Blood Pressure (NIBP) measurement
methods nowadays, is largely dependant on the quality of hearing of the (general) practitioner
(G.P.) [Chalifoux A., 1985][Wagner, A.E., 1997].
First the cuff is placed and inflated far above the SAP for any patient. While the G.P. listens
with his stethoscope on the brachial artery, the cuff is slowly deflated. Eventually the pressure
drops below the SAP. When this happens a turbulent blood flow occurs. This turbulent blood flow
produces a distinctive sound know as Korotkoff-sounds, named after the person who discovered it.
As soon as the cuff pressure falls below the DAP the blood flow becomes unimpeded and the
Korotkoff-sounds disappear [Chalifoux A., 1985].
The Korotkoff -sounds are however virtually impossible to hear, even for the best trained
cardiologist, in canine and feline patients and as such the method is completely useless in small
animal medicine [Chalifoux A., 1985][Mandigers P., 2005][Wagner, A.E., 1997].
Figure 1 The principle behind the non invasive blood pressure measurement. Cuff is inflated above the 120
mmHg (B) and than lowered (line between B and C) as pressure decreases the pressure waves pass
underneath the cuff resulting in either Korotkoff sounds, Doppler-Audio output or pressure oscillations. These
disappear when the line B-C falls below the 80 mmHg. (humane) [Berne, R.M., 2004].
In veterinary medicine there are currently two non-invasive techniques in use. The Non
Invasive Doppler flow-ultrasonography (NDFU-NIBP) and the Non Invasive Oscillometric
Spygmomamometry (O-NIBP). Figure 1 shows the basic principle of all non invasive methods
[Deflandre, C.J.A., 2008][Mandigers, P., 2005][Nelson, R.W., 2003]
.
The NDFU-NIBP system is based on a comparable principle as the sphyngomanometric
method but instead of a stethoscope, an ultrasound transducer is used to determine when the
blood starts to flow again. It will create an audio signal as soon as the cuff pressure falls below the
SAP. DAP and the related MAP can not be determined reliably with this system because it is difficult
to determine when the blood flow becomes unimpeded by the cuff [Mandigers P., 2005][Nelson, R.W., 2003].
As a result the Doppler Ultrasonography is an unreliable system for determining
hypotension. The ultrasound-transducer needs to be moved underneath the cloths of the sterile
field, a region that is difficult to access for any anesthesiologist under normal circumstances. As
such the NDFU-NIBP has no place in the operating theater. Next to this surgical problem it has also
been determined that the NDFU-NIBP is more accurate at higher blood pressures The NDFU-NIBP
system can however be used reliable in animals suspected of systolic hypertension. Since cats are
5
known to suffer from idiopathic high arterial blood pressure this method is especially useful in
these cases [Mandigers, P., 2005].
Non Invasive Oscillometric Spygmomamometry (O-NIBP) once again uses the same
principle as the before mentioned auditory and Doppler-Ultrasonographic methods. It uses a cuff
that is inflated and deflated in cycles. The machine however detects small oscillations in the
pressure of the cuff during the phase that the cuff is below SAP and above DAP [Nelson, R.W., 2003].
The maximum reading is equal to the MAP and from that point an algorithm calculates the SAP and
DAP [Valerio, F., 2006].
Several O-NIBP devices have been tested over the years with variable results. Some were
capable of delivering accurate readings while others underestimated pressures [Deflandre, C.J.A.,
2008][Sawyer, D.C., 1994][Meurs K.M., 1996].
The cuff can be placed on three sites on the patient, the tail base (medial sacral artery –
medial caudal artery), the hind limb (dorsal pedal artery) or the front limb (palmar arches of the
median artery). Several authors point out that the most reliable place for O-NIBP measuring in
dogs is the tail base followed by the hind limb’s metatarsal region [Mandigers, P., 2005][Habermann, C.E.,
2006][Sawyer, D.C., 2004].
It needs to be noted that, although anesthesia should prevent any active movement, the
site should always be selected on basis of movement. It has been demonstrated (in conscious
dogs) that the cuff can best be placed on the tail base when the patient is awake, primarily
because cuff placement on the other measurement sites (metacarpal and/or hock) can lead a
preoccupation with removing the cuff from its extremity. This results in more movement and as
such less reliable data collection. Shivering should be avoided for the same reason [Grosenbaugh, D.A.,
1998].
For the O-NIBP, clipping the area of cuff-placement is unnecessary, although longhaired
breeds can prove to be a challenge. The long fur can obstruct the use of the Velcro used on most
cuffs [Mandigers, P., 2005][Bodey, A.R.,1994].
Another factor in reliability is the size of the cuff. In 1980 Geddes et al. demonstrated that
a cuff: circumference ratio of about 0,40 (or 40%) in dogs front limb produced the best results.
Selection of the wrong cuff can lead to under- or overestimation. The research of Geddes et al.
registered the MAP via an IABP system and the pressure in the balloon inside the cuffs. This in
conjunction with a registration of the oscillations in the balloon resulted in the advise to use a cuff:
circumference ratio of 0,40 or 40% [Geddes, L.A., 1980]. The manufacturer of the cuffs used in this
experiment also recommends such a ratio.
A major concern is the algorithm used by the machine especially when the O-NIBP is
determined via a machine originally designed for human medicine. The algorithm itself could be
unsuitable for veterinary use however this has yet to be determined [Valerio F., 2006].
6
Goals of this study
The objectives of this study are based upon the current use of the Datex Ohmeda GE S/5 O-NIBP
module in the surgical facilities of the department of Clinical Sciences of Companion Animals at the
faculty of Veterinary Medicine, University of Utrecht, Utrecht, the Netherlands. Since it is currently
used without any form of verification as to the reliability of the device, it remains difficult to
determine whether or not critical decisions in patient care can be made on basis of data generated
by this system. Hence the first objective of this study is to compare the Datex Ohmeda GE S/5 ONIBP and the golden standard (IABP) to determine its reliable and to determine whether it provides
comparable and accurate data in patients during general anesthesia, before and during surgery.
The corresponding hypotheses is “The oscillometric Arterial blood Pressure device is only reliable at
values equal or lower then normotension in comparison to the IABP”.
The second objective is to determine the optimal site for cuff placement for the Datex
Ohmeda GE S/5 on the patient. Corresponding hypothesis “The optimal site for cuff placement is
the tail base”.
After writing the research proposal it became clear that this study would be carried out in
conjunction with a study in to the anesthetic effect of dexmedetomidine and alfaxalone. As such it
became possible to look at the differences in the performance of the Datex Ohmeda GE S/5 in the
different anesthetic protocols (methadone-isoflurane, dexmedetomidine-propofol and
dexmedetomidine-alfaxalone). This became the third objective of the study.
7
Materials and methods
Thirty-one canine patients of various breeds, weighing 32.0 +/- 14.8 kg (mean, SD) were
used for this study. All the dogs were scheduled for general anaesthesia and would either need an
arterial catheter for monitoring there blood pressure or they took part in another study protocol (in
this case the recently registered anaesthetic agent alfaxalone (alfaxalone 10 mg/mL solution,
Alfaxan®, Vétoquinol Buckingham England) for which an arterial catheter was also required to
collect arterial blood pressure data.
The average age of the dogs was 4,5 years, with a variation of 1,5 months to over ten years of
age. The ASA classifications ranged from I to IV. Sixteen dogs were classified as ASA-I, six in ASAII, six in ASA-III and three in the ASA-IV.
There were no exclusion criteria based upon gender or breed. Size was limiting factor since
the smallest cuff could fit an extremity of 3 centimetres circumference or up and the largest cuff
could only fit an extremity of up to fifteen centimetres circumference.
No exclusion is made based up on the premedication and anaesthetic protocols used on the
patients. Commonly used protocols included dexmedetomide-alfaxalone (9), dexmedetomidinepropofol (9), methadone-midazolam-isoflurane (12). In one case sevoflurane was used in stead of
isoflurane in the methadone protocol. All of the dexmedetomidine patients were classified as ASA-I
and ASA-II.
All patients were intubated and the arterial catheter (Abbocath-T, 22 SWG, 32 mm, Abbot,
Sligo, Republic of Ireland), arterial cannula with flow switch, (20 SWG, 45 mm, Braun Dickinson,
Swindon, UK or Secalon T over the needle central venous catheter with flow switch, 18 SWG, 90
mm, Braun Dickinson Critical Care Systems Pte Ltd, Singapore) was placed either at the hock
(dorsal pedal artery, peripherally) or in the thigh (femoral artery, centrally) depending on the
preferences of the staff and on whether the animal participated in the alfaxalone-study (dorsal
pedal artery only).
The catheter is placed by an anesthesiologist or nurse anesthetist to ensure proper and
rapid placement. After placing the catheter is flushed using a 2500 IE heparin/0.9% and then fixed
to the patient.
In the operating room (OR) the arterial catheter is connected to a disposable transducer
(Gabarith TM single transducer set, Braun Dickinson Infusion Therapy Systems Inc, Sandy, UT,
USA) which is positioned at the height of the right atrium. The transducer is connected to the
Datex Ohmeda GE S/5 monitor and micro flushed continuously using a prepared 2500 IE
heparin/0.9% saline. To ensure proper micro flush function this solution is pressurized to
approximately 300 mmHg to prevent backflow of blood into the catheter-transducer system. The
IABP- system is then is zeroed to the ambient air pressure in the OR.
To determine the correct cuff-size for measuring the NIBP the circumference (in
centimeters) of the extremity is determined. Based on this measurement the appropriate cuff
(Critikon Neonatal soft cuffs, sizes 1 to 5 for DINAMAP available) size is chosen. Manufacturer’s
recommendations for human use were taken into account in selecting the appropriate cuff size
(e.g. the circumference of the measurement-site had to correspond with the manufacturer’s
circumference recommendations on the cuffs). Cuffs are placed on three different places. Before
and during the surgery the preferred location is the contra lateral hind limb just below the hock
(metatarsal region, 22 patients). If placement is impossible because due to the surgical procedure
undertaken in that region, the cuff is placed at the tail base (9 patients).
When the cuffs were placed they were connected to the Datex Ohmeda GE S/5 using the infant cuff
connector system. After placement, the O-NIBP system was tested to see if the selected location
delivered viable data, if not the cuff was repositioned on the same site (two times maximum) and
tested again. When it was possible to reach the tail base, the cuff was then moved to this position
and the procedure of placement and testing was repeated. If positioning of the cuff was no longer
possible due to lack of time (start of the surgery was imminent) the patient was considered lost to
this study.
A total of 5 consecutive measurements are recorded before the start of the surgical
procedure. This is repeated during surgery at intervals of 2,5 minutes minimum and a maximum of
10 minutes. After surgery the cuff is moved to the two other measuring locations to make three
consecutive measurements at each site.
8
Data is recorded by hand on a hard copy excel data-sheet and the IBP (continuous
measurement) in the following sequence: heart rate, invasive blood pressures (SAP, DAP, MAP)
and last O-NIBP pressures. This was done to get the best accuracy since the heart rate and
invasive pressures were continuous readings. After completion of the measurements a printer was
connected to the Datex monitor to print out the numerical trend sheet (displaying IBP, NIBP and
HR at any time, per minute, during surgery). The hand written data then function as a back-up.
Statistical analysis was performed using the Bland Altman method to compare two
machines for a parameter [Bland, J.M., 1986]. It is slightly modified to a percentual bias of the O-NIBP
measurement compared to the IBP.
Also the different protocols will be compared and the different measurement sites using an
ANOVA and scatter plots respectively [Petrie, A., 1999].
9
Results
Table 2 shows the different arterial blood pressures found in the two premedication-groups
measured during the pre-operative phase. The mean SAP, DAP and MAP in the dexmedetomidine
groups was significantly higher than in the methadone group.
Table 2: Average SAP, DAP and MAP, based upon data from the IABP (for all groups 0.02 < P < 0.05).
Premedication
In the protocol
Patients (N)
SAP (mmHg)
(mean +/- Sd)
DAP (mmHg)
(mean +/- Sd)
MAP (mmHg)
(mean +/- Sd)
Methadone
13
97 +/- 18
52 +/- 13
66+/- 14
Dexmedetomidine
18
149 +/- 19
88 +/- 11
104 +/- 18
The first time failure rate (FFR) (total and divided on basis of the anesthetic protocol) I
presented in Table 3. The first time failure rate is the calculated using the formula rate (%) =
(x/n)*100%. The FFR is technically the number of measurements in which the O-NIBP device failed
to provide an instant measurement. In the protocols were dexmedetomidine was used as
premedicant, the failure rates were higher compared to methadone (dexmedetomidine, average
FFR of 31,8% versus 8,5% in the methadone groups).
Table 3: First time failure rate divided on basis of the protocol and overall
Protocol
Patients (N)
Total number of
measurements (n)
Number of
failures (x)
FFR (%)
MethadoneIsoflurane
12
183
14
7
MethadoneSevoflurane
1
16
3
19
DexmedetomidinePropofol
9
132
49
37
DexmedetomidineAlfaxalone
9
132
35
26
Overall
31
463
103
22
In table 4 the standard deviations (Sd) and corresponding P-values are given for each
measuresite and bloodpressure parameter. The standard deviations of the NIBP values are all
higher, although the difference is only significant if the value of P is below 0.05. In other words
there is more variation. The tail base gives the least variable results (Sd NIBP is the lowest) in this
case and the variation (Sd) is higher during surgical stimulation. The MAP appears to be the best
(accurate) of the blood pressure determinants (P≤0,02).
Displayed in figure 2 a-c are the Bland-Altman, or difference plots for SAP, DAP and MAP
from the hind limb, taken pre surgically. These graphically display what can be found with the
Bland-Altmann method for comparing two different measurement-methods. The identity (thin, grey
line) is equal to zero bias (e.g. methods do not differ). The bias is the average difference (thick,
blue line) and the 95% limits of agreement are the limits in which 95% of the bias-point can be
found (thick, interrupted, light blue line). The 95% CI lines are the 95% confidence intervals (thin,
dotted, light blue lines). If a line would be drawn through the dot-clouds it would give a >-shaped
line with the point of the > in the direction of the higher blood pressures. This >- shape is visible in
the SAP and MAP scatter plots and means that the higher the ABP the smaller the Bias gets.
10
Difference Plot
Difference (gemSAP-NIBPpreAP - gemSAP-IBPpre AP) /
Mean of All
50%
Identity
40%
Bias (-2.5%)
30%
95% CI
20%
95% Limits of agreement
(-35.6% to 30.6%)
10%
95% CI
0%
-10%
-20%
-30%
-40%
-50%
40
60
80
100
120
140
160
180
200
Mean of All
Figure 2a: SAP difference plot for the O-NIBP (pre-surgical stimulation) at the hind limb.
Difference Plot
Difference (gemDAP-NIBPpreAP - gemDAP-IBPpreAP) /
Mean of All
60%
Identity
40%
Bias (-26.7%)
20%
95% CI
0%
95% Limits of agreement
(-78.6% to 25.2%)
95% CI
-20%
-40%
-60%
-80%
-100%
20
40
60
80
100
Mean of All
Figure 2b: DAP difference plot for the O-NIBP (pre-surgical stimulation) at the hind limb.
11
Difference Plot
Difference (gemMAP-NIBPpreAP - gemMAP-IBPpreAP) /
Mean of All
50%
Identity
40%
Bias (0.5%)
30%
95% CI
20%
95% Limits of agreement
(-35.4% to 36.3%)
10%
95% CI
0%
-10%
-20%
-30%
-40%
-50%
40
60
80
100
120
140
Mean of All
Figure 2c: MAP difference plot for the O-NIBP (pre-surgical stimulation) at the hind limb.
Scatter Plot with Fit
220
Linear fit (15.88 +0.8513x)
200
95% CI
180
95% Prediction interval
gemSAP-NIBPpreAP
160
140
120
100
80
60
40
20
60
80
100
120
140
160
180
200
gemSAP-IBPpre AP
Figure 3: SAP scatter plot of the IBP (x-axis) versus the NIBP (y-axis). Further data is presented in table 5.
12
As mentioned before, figures 2a-c only display the data for the hind limb under pre-surgical
conditions. In the appendix (Figures 1a-c to 3a-c) the other measurement sites can be found.
These figures are virtually the same as the figures displayed in the text (with the exception of the
front limb, which appears to be equally biased at any arterial blood pressure).
Table 4: Standard deviations for each measurement location
Location
sd SAP sd SAP PIBP
NIBP
value
sd DAP sd DAP PIBP
NIBP
value
sd MAP sd MAP PIBP
NIBP
value
Hind limb
Pre-anaest
2.35
1.55
0.044
2.11
2.89
0.717
1.24
1.94
0.001
Hind limb
during Sur.
3.94
9.80
0.071
2.89
19.99
0.118
3.13
3.86
0.010
Front Limb
Pre-anaest
2.19
5.73
0.032
1.59
3.15
<0.001
1.54
5.62
0.020
Tail base
Pre-anaest
1.22
5.73
<0.001
1.12
2.85
0.001
1.02
1.78
0.008
In Figure 3 the scatter plot displays the NIBP compared to the IBP for the Systolic Arterial
blood pressure. As can be seen in this figure the difference between the NIBP and the IBP
decreases as the pressure becomes higher. However it should also be noted that it is not an exact
fit. This is expressed by the fact that the identity of the line is not Y=X. In other words when the
IBP is 60 mmHg the NIBP gives a value of approximately 67 mmHg.
In conjunction with data presented in Appendix table 1, which containes the corresponding
regression lines for all the different measurement sites (hind limb, front limb and tail base) it is
possible to determine that the best correlation is found in the tail base measurements (R is high).
MAP once again gives the best correlations and the front limb is less accurate (based upon the
identity of the lines/equation).
Figure 4 a-c show information about the MAP trend over time for 6 randomly selected
patients without regard for the protocol or measurement site. Although some of the peaks and dips
that appear in the IABP/IBP are not found in the NIBP and the other way around the lines are quite
similar. SAP and DAP plots (not included) show a similar pattern however the variability is higher in
SAP and DAP so the NIBP doesn’t always follow the IABP so closely.
Summarizing all of the results above, the tail base O-NIBP provides the least biased results
compared to the IABP. The front limb appears to be the least reliable of the three measurementsites. Furthermore the MAP is the most accurate of the three measurement parameters for the
arterial blood pressure.
The cuff size itself does not seem to influence the accuracy if the manufacturer’s
recommendations are taken in to account.
The anesthetic protocols influence the O-NIBP to some extent. The dexmedetomidine
protocols tend to make the Datex Ohmeda GE S/5 less capable to produce instant results, which
means that it will take more time to get a single measurement since the inflation-deflation
sequence of the cuff has to be done over again. The anesthetic protocols do not appear to differ in
the amount of bias found (ANOVA yields no significant differences).
Analysis of the trend-lines of all patients suggests, that the Datex Ohmeda can provide
information about the SAP, DAP and MAP over time.
13
MAP Trend (patient 1)
MAP Trend (patient 2)
200
mmHg
mmHg
150
NIBP MAP
100
IBP MAP
50
0
160
140
120
100
80
60
40
20
0
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43
NIBP MAP
IBP MAP
1
measurement
NIBP MAP
IBP MAP
5
7
9 11 13 15 17 19 21 23
NIBP MAP
IBP MAP
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
measurement
MAP Trend (patient 6)
80
70
60
50
40
30
20
10
0
mmHg
mmHg
MAP Trend (patient 5)
NIBP MAP
IBP MAP
3
5
7
9
11 13 15
9 11 13 15 17 19 21 23
160
140
120
100
80
60
40
20
0
measurement
1
7
MAP Trend (patient 4)
mmHg
mmHg
160
140
120
100
80
60
40
20
0
3
5
Measurement
MAP Trend (patient 3)
1
3
160
140
120
100
80
60
40
20
0
NIBP MAP
IBP MAP
1 3 5 7 9 11 13 15 17 19 21 23 25
17 19
measurement
measurement
Figure 4 a-c: Randomly selected MAP trends for multiple patients.
14
Discussion
When the O-NIBP device is used in a clinical setting it will surely provide the basis for
decisions about fluid therapy and/or vasopressive therapy and inotropic drug use. Hence it is
important that the O-NIBP (in this case the Datex Ohmeda GE S/5 O-NIBP module) provide
consistent and reliable results. The American Association of Medical Instrumentation (AAMI) and
the British Hypertension Society have determined standards for diverse medical equipment
including NIBP-measuring devices. According to these standards any NIBP device should predict
the values within a range of 5+/- 8 mmHg of error compared to the golden standard (IBP). Only a
few veterinary NIBP monitors have been able to comply to these standards [Deflandre, C.J.A., 2008].
It should be noted that (in contrast to other authors, like Deflandre et al.) we looked at the
percentage of bias and 95%-limits of agreements and not the absolute numbers. To put it simply, a
bias of 5 mmHg is relatively limited on a blood pressure of 200 (2,5%). However it is quite large on
50 mmHg (10%). As such we prevent the misinterpretation on the extent of the bias especially in
the lower blood pressures.
The variability for SAP, DAP and MAP differs substantially. MAP is relatively unbiased
especially on the tail base, while SAP and DAP are more variable. This can be attributed to the fact
that when the MAP and cuff-pressure are (nearly) equal, the oscillations in the cuff are maximized
and thus the signal is most distinct. Furthermore, the MAP is least distorted in the IABP because
the dynamic response characteristics and distortions in the recording system are of little
importance [Gardner, R.M., 1981]. The problems with the accuracy of the O-NIBP in light of the SAP can
in part be attributed to the pulse wave. The peak pressure is relatively short lived and can easily be
missed in an inflation-deflation cycle of the O-NIBP [Mandigers, P., 2005] [Binns S.H., 1995][Grosenbaugh D.A.,
1998] [Stephien, R.L.,1999]. Furthermore the formation of oscillations takes more energy than pressure
waves in the IABP doing there work on the transducer, as such explaining further signal loss in the
O-NIBP [Binns S.H., 1995]. Lastly, there is still the aspect of vasoconstriction, which is clearly
demonstrated by the results from the different anaesthetic protocols.
It should be noted that the measurements on the hind limb are based upon five
consecutive measurements and the tail base and front limb measurements on three. This
difference means that the amount tail base and front limb data is considerably less then hind limb
data.
In this study several types of anesthetic protocols where used, one was a protocol of
methadone and isoflurane (twelve patients). The others were dexmedetomidine in conjunction with
propofol (nine patients) or alfaxalone (nine patients), respectively. It was found that in the
dexmedetomidine groups the blood pressure levels where higher then in the methadone-isoflurane
group (Table 2).
Since dexmedetomidine is a potent vasoconstrictive agent, like all other α2-agonists, it
could inevitably influence the results due to vasoconstriction in the periphery where the O-NIBP is
to get a signal [Uilenreef, J.J., 2008][Lin, G.Y., 2008]. As mentioned in the results, the first time failure rate
in the protocols in which dexmedetomidine was used is higher than in the methadone-isoflurane
protocol (Table 3), which could mostly likely be attributed to the vasoconstriction caused by
dexmedetomidine.
Alfaxalone, which was used for induction and maintenance of anaesthesia, has
cardiovascular effects as well. Alfaxalone produces a significant drop in blood pressure, especially
in higher dosages (e.g 20 mg kg-1) [Muir, W., 2008]. In the dosages of alfaxalone used in this study
(0.9 – 2.0 mg kg-1 and 2.6 – 7.6 mg kg-1 hour-1 for induction and maintenance, respectively) the
effects are minimal and α2-agonistic effects (the standard premedication dexmedetomidine dose
was 5 μg kg-1 in the alfaxalone study, followed by a continuous rate infusion (CRI) of 1 μg kg -1
hour-1) on the cardiovascular system are dominant.
Methadone has minimal effects on the cardiovascular system when given slowly
(intravenous route) over the course of a minute [Hall, L.W., 2001].
Propofol, isoflurane and sevoflurane all give a decrease in blood pressure due to actions on
the heart itself and vasodilatation which can is not be prevented by premedication compensated
when these are combined with methadone. However, the combination of propofol and
dexmedetomidine shows higher blood pressures, which can be attributed to the α2-agonistic effects
on the vascular system (peripheral vasoconstriction) [Hall, L.W., 2001].
A strange phenomenon remains being that the data collected during surgery are more
variable then the data collected with out surgical stimulation. Theoretically the patient should
remain as cardiovascular stable during surgery as before surgery. The exact reason is obscured
since the exact amount and timing of surgical stimulation was not scored and recorded as goes for
the amount and type of fluids administered to the patient.
15
Location of the cuff has effects on the accuracy of the O-NIBP. The use of the tail base
produced the best results followed by the hind limb and then the front limb. The accuracy of the
tail base has been demonstrated by other authors in anesthetized [Bodey,A.R., 1994][Haberman, C.E., 2006]
and conscious dogs [Bodey,A.R., 1996]. It has been demonstrated that (at least in conscious dogs) that
cuff placement on front- or hind limbs leads to a preoccupation with removal of the cuff in dogs.
Since the dogs where anesthetized the only relevant movement would be shivering. Shivering was
never seen however during measurements, as such one would expect similar results on hind limb
and tail base. Perhaps the pressure-waves in the medial caudal artery are relatively strong when
compared to the dorsal pedal artery. Thus creating a better signal and more reliable results at the
tail base cuff site.
16
Conclusions
Concluding from all above, the Datex Ohemda Ge S/5 O-NIBP blood pressure measurement
device tends to under- or overestimate the IABP given the data in table 4. There is also a tendency
to be less accurate during times of surgical stimulation. Accuracy, in contrast to the Goal one
hypothesis (“The oscillometric Arterial blood Pressure device is only reliable at values equal or
lower then normotension in comparison to the IABP”), tends to be lower in the low-normotensive
and hypotensive range and the accuracy increases with increasing arterial blood pressure.
It has also been determined that the tail base is the most accurate position to place the
cuff and that it is wise to take the recommended cuff size for a given circumference into account.
The anesthetic protocol used also influences the O-NIBP’s ability to get first time readings
(not the accuracy of the readings), which can be attributed to the vasoconstrictive action of α2agonists on the vascular system. First time failure rate in the dexmedetomidine group (32%) is
more then three times higher then in the methadone groups (9,5%).
17
Clinical relevance
The use of the Datex Ohmeda Ge S/5 O-NIBP is only advisable under certain
circumstances. The IBP remains to be the standard blood pressure measuring method for patients
with ASA-III or higher.
The Datex O-NIBP can provide trend information about the patient’s blood pressure (with
some considerations to its accuracy) and can be used for patients that are undergoing anesthesia
for minor surgery or non surgical procedures for example MRI, ultrasonography or X-ray’s.
If placement of the arterial catheter fails, the Datex O-NIBP can be used as an alternative
device to the IABP if properly used (e.g. cuff placed on the tail base and the correct cuff size is
used). Single measurements yield no reliable data, it is best to consider trends for decision making.
Sudden peaks or dips don’t always indicate that a patient is in trouble. When dips or peaks are
maintained over longer periods or when other vital signs are changing in conjunction to a single
measurement it could spell trouble and action can be taken. The MAP needs to be the main focus
when monitoring the blood pressure with the Datex since this parameter is the least biased and
least variable of the three arterial blood pressure parameters.
Anesthetic protocols have some influence on the system’s functioning, and although they
do not change the amount of variation, they can change the Datex ability to quickly provide
results. Especially protocols with dexmedetomidine influence the O-NIBP system so this should be
considered when selecting a time interval for measurements. The best interval time for the Datex
is 2,5 to 5 minute at maximum otherwise the intervals become to great to provide up-to-date
information. Since the average time for an inflation-deflation cycle is approximately one minute
shorter than 2,5 minute intervals are not advisable.
18
Acknowledgements
I would like to thank the following persons for supporting me during my research.
Rob Sap for supporting me on a day to day basis with my problems and questions as well
for is support during the writing of research-proposal, this research paper and the article that
(hopefully) is going to follow from it. Furthermore, my thanks for the placement of many arterial
catheters on the patients, otherwise measuring would not have been possible.
On that same matter (placement of arterial catheters) I would also like to thank Ies
Akkerdaas and Joost Uilenreef. And of course Ies also for suggesting the NIBP-subject in the first
place. And the both of them for suggestions and tips given during my time on the OR.
Of course the rest of the staff working on the OR during my “onderzoeksstage” should be
mentioned as well for making it so enjoyable. Noting better than having a good time when you are
doing your work.
And last but not least I would like to thank Erik Teske for his help and support with the
statistical analysis of the collected data. He helped me a great deal with his insights and his time
input for analysing the data. Knowing myself it would have turned out a complete disaster when I
was left by my self to do the statistics on my own. Thanks again.
19
Appendix
Difference (gemSAP-NIBP AP - gemSAP-IBP AP) / Mean of All
Difference Plot
50%
Identity
40%
Bias (-1.0%)
30%
95% CI
20%
95% Limits of agreement
(-34.7% to 32.7%)
10%
95% CI
0%
-10%
-20%
-30%
-40%
-50%
60
80
100
120
140
160
180
200
220
Mean of All
Appendix Figure 1a: SAP difference plot for the O-NIBP (during surgical stimulation) at the hind limb
Difference (gemDAP-NIBP AP - gemDAP-IBP AP) / Mean of All
Difference Plot
80%
Identity
60%
Bias (-15.3%)
40%
95% CI
20%
95% Limits of agreement
(-71.0% to 40.5%)
0%
95% CI
-20%
-40%
-60%
-80%
-100%
20
40
60
80
100
120
Mean of All
Appendix figure 1b: DAP difference plot for the O-NIBP (during surgical stimulation) at the hind limb
20
Difference Plot
Difference (gemMAP-NIBP AP - gemMAP-IBP AP) / Mean of
All
50%
Identity
40%
Bias (2.9%)
30%
95% CI
20%
95% Limits of agreement
(-28.1% to 34.0%)
10%
95% CI
0%
-10%
-20%
-30%
-40%
-50%
50
70
90
110
130
150
Mean of All
Appendix figure 1c: MAP difference plot for the O-NIBP (during surgical stimulation) at the hind limb
Difference Plot
Difference (gemSAP-NIBP_VP - gemSAP-IBP_VP) / Mean of
All
40%
Identity
30%
Bias (-7.8%)
20%
95% CI
10%
95% Limits of agreement
(-41.7% to 26.1%)
0%
95% CI
-10%
-20%
-30%
-40%
-50%
-60%
80
100
120
140
160
180
200
Mean of All
Appendix figure 2a: SAP difference plot for the O-NIBP (post surgical stimulation) at the front limb
21
Difference (gemDAP-NIBP_VP - gemDAP-IBP_VP) / Mean of
All
Difference Plot
40%
Identity
20%
Bias (-27.4%)
95% CI
0%
95% Limits of agreement
(-75.0% to 20.1%)
-20%
95% CI
-40%
-60%
-80%
-100%
20
40
60
80
100
Mean of All
Appendix figure 2b: DAP difference plot for the O-NIBP (post surgical stimulation) at the front limb
Difference (gemMAP-NIBP_VP - gemMAP-IBP_VP) / Mean of
All
Difference Plot
40%
Identity
30%
Bias (-7.3%)
20%
95% CI
10%
95% Limits of agreement
(-35.7% to 21.0%)
0%
95% CI
-10%
-20%
-30%
-40%
-50%
60
70
80
90
100
110
120
130
Mean of All
Appendix figure 2c: MAP difference plot for the O-NIBP (post surgical stimulation) at the front limb
22
Difference (gemSAP-NIBP_SB - gemSAP-IBP_SB) / Mean of
All
Difference Plot
30%
Identity
20%
Bias (-11.2%)
10%
95% CI
0%
95% Limits of agreement
(-36.0% to 13.6%)
95% CI
-10%
-20%
-30%
-40%
-50%
60
80
100
120
140
160
180
200
Mean of All
Appendix figure 3a: SAP difference plot for the O-NIBP (post surgical stimulation) at the tail base
Difference (gemDAP-NIBP_SB - gemDAP-IBP_SB) / Mean of
All
Difference Plot
30%
Identity
20%
Bias (-22.2%)
10%
95% CI
0%
-10%
95% Limits of agreement
(-56.3% to 12.0%)
-20%
95% CI
-30%
-40%
-50%
-60%
-70%
20
40
60
80
100
Mean of All
Appendix figure 3b: DAP difference plot for the O-NIBP (post surgical stimulation) at the tail base
23
Difference (gemMAP-NIBP_SB - gemMAP-IBP_SB) / Mean of
All
Difference Plot
30%
Identity
20%
Bias (-4.6%)
95% CI
10%
95% Limits of agreement
(-23.6% to 14.3%)
0%
95% CI
-10%
-20%
-30%
-40%
40
60
80
100
120
140
Mean of All
Appendix figure 3c: MAP difference plot for the O-NIBP (post surgical stimulation) at the tail base
Appendix table 1: Values for scatter plots for each location. Values from tail base (xxx-xxxx-TB) based upon
three measurements instead of five for the hind limb (xxx-xxxx-HL), hind limb presurgical (xxx-xxxx-preHL)
and front limb (xxx-xxxx-FL).
Parameter
Combination
R
Adjusted-R2
Significance
Equation
SAP-IABP-preHL  SAP-NIBP-preHL
0,841
0,697
<0,001
Y=15.884 + 0.851X
DAP-IABP-preHL  DAP-NIBP-preHL
0,801
0,628
<0,001
Y=-8.118 + 0.904X
MAP-IABP-preHL  MAP-NIBP-preHL
0,852
0,715
<0,001
Y=2.875 + 0,981X
SAP-IABP-HL  SAP-NIBP-HL
0,747
0,536
<0,001
Y=36.262 +0.707X
DAP-IABP-HL  DAP-NIBP-HL
0,730
0,510
<0,001
Y=2,394 +0.853X
MAP-IABP-HL  MAP-NIBP-HL
0,850
0,709
<0,001
Y=12.751 + 0.888X
SAP-IABP-FL  SAP-NIBP-FL
0,654
0,289
0,003
Y=55.04 + 0.949X
DAP-IABP-FL  DAP-NIBP-FL
0,596
0,327
0,002
Y=16.206 + 0.533X
MAP-IABP-FL  MAP-NIBP-FL
0,659
0,409
<0,001
Y=37.551 + 0.490X
SAP-IABP-TB  SAP-NIBP- TB
0,873
0,752
<0,001
Y=-19.895 + 1.06X
DAP-IABP-TB  DAP-NIBP- TB
0,919
0,839
<0,001
Y=-12.488 + X
MAP-IBP-TB  MAP-NIBP- TB
0,933
0,870
<0,001
Y=-7.899 + 1.056X
24
References
[Anonymous]. (2005)
Standards for basic anesthetic monitoring.
http://www.asahq.org/publicationsAndServices/standards/02.pdf. Accessed 12/8, 2008
BERNE R. M., LEVY M. N., KOEPPEN B. M. & STANTON B. A. (2004) Physiology. In Arterial System.
5th edn. Anonymous St. Louis, Mosby-Elsevier. pp 355-367
BILLIET E. & COLARDYN F. (1998) Pressure measurement evaluation and accuracy validation: The
gabarith test. Intensive Care Medicine 24, 1323-1326
BINNS S. H., SISSON D. D., BUOSCIO D. A. & SCHAEFFER D. J. (1995) Doppler ultrasonographic,
oscillometric sphygmomanometric, and photoplethysmographic techniques for noninvasive blood
pressure measurement in anesthetized cats. Journal of Veterinary Internal Medicine / American
College of Veterinary Internal Medicine 9, 405-414
BIRKS R. J. S., GEMMELL L. W., O’SULLIVAN E. P., ROWBOTHAM D. J. & SNEYD J. R. (2007)
http://www.aagbi.org/publications/guidelines/docs/standardsofmonitoring07.pdf. Accessed 12/8,
2008
BLAND J. M. & ALTMAN D. G. (1986) Statistical methods for assessing agreement between two
methods of clinical measurement. Lancet 1, 307-310
BODEY A. R., YOUNG L. E., BARTRAM D. H., DIAMOND M. J. & MICHELL A. R. (1994) A comparison
of direct and indirect (oscillometric) measurements of arterial blood pressure in anaesthetised
dogs, using tail and limb cuffs. Research in Veterinary Science 57, 265-269
CHALIFOUX A., DALLAIRE A., BLAIS D., LARIVIERE N. & PELLETIER N. (1985) Evaluation of the
arterial blood pressure of dogs by two noninvasive methods. Canadian Journal of Comparative
Medicine.Revue Canadienne De Medecine Comparee 49, 419-423
DEFLANDRE C. J. & HELLEBREKERS L. J. (2008) Clinical evaluation of the surgivet V60046, a non
invasive blood pressure monitor in anaesthetized dogs. Veterinary Anaesthesia and Analgesia 35,
13-21
GAINS M. J., GRODECKI K. M., JACOBS R. M., DYSON D. & FOSTER R. A. (1995) Comparison of
direct and indirect blood pressure measurements in anesthetized dogs. Canadian Journal of
Veterinary Research = Revue Canadienne De Recherche Veterinaire 59, 238-240
GARDNER R. M. (1981) Direct blood pressure measurement--dynamic response requirements.
Anesthesiology 54, 227-236
GEDDES L. A., COMBS W., DENTON W., WHISTLER S. J. & BOURLAND J. D. (1980) Indirect mean
arterial pressure in the anesthetized dog. The American Journal of Physiology 238, H664-6
GROSENBAUGH D. A. & MUIR W. W.,3RD. (1998) Accuracy of noninvasive oxyhemoglobin
saturation, end-tidal carbon dioxide concentration, and blood pressure monitoring during
experimentally induced hypoxemia, hypotension, or hypertension in anesthetized dogs. American
Journal of Veterinary Research 59, 205-212
HABERMAN C. E., KANG C. W., MORGAN J. D. & BROWN S. A. (2006) Evaluation of oscillometric
and doppler ultrasonic methods of indirect blood pressure estimation in conscious dogs. Canadian
Journal of Veterinary Research = Revue Canadienne De Recherche Veterinaire 70, 211-217
HALL L. W., CLARKE K. W. & TRIM C. M. (2001) Veterinary anaesthesia. 10th edn. London,
Saunders
25
KUUSELA E., VAINIO O., SHORT C. E., LEPPALUOTO J., HUTTUNEN P., STROM S., HUJU V.,
VALTONEN A. & RAEKALLIO M. (2003) A comparison of propofol infusion and propofol/isoflurane
anaesthesia in dexmedetomidine premedicated dogs. Journal of Veterinary Pharmacology and
Therapeutics 26, 199-204
LIN G. Y., ROBBEN J. H., MURRELL J. C., ASPEGREN J., MCKUSICK B. C. & HELLEBREKERS L. J.
(2008) Dexmedetomidine constant rate infusion for 24 hours during and after propofol or isoflurane
anaesthesia in dogs. Veterinary Anaesthesia and Analgesia 35, 141-153
MANDIGERS P. (2005) Methoden van niet-invasieve bloeddrukmeting bij de hond en kat. 130, 198203
MEURS K. M., MILLER M. W. & SLATER M. R. (1996) Comparison of the indirect oscillometric and
direct arterial methods for blood pressure measurements in anesthetized dogs. Journal of the
American Animal Hospital Association 32, 471-475
MUIR W., LERCHE P., WIESE A., NELSON L., PASLOSKE K. & WHITTEM T. (2008) Cardiorespiratory
and anesthetic effects of clinical and supraclinical doses of alfaxalone in dogs. Veterinary
Anaesthesia and Analgesia
NELSON R. W. & COUTO C. G. (2003) Small animal internal medicine. In Systemic Arterial
Hypertension. 3rd edn. Anonymous St. Louis, Mosby. pp 199-205
PETRIE A. & WATSON P. (1999) Statistics for veterinary and animal science. In Additional Topics.
1st edn. Anonymous Malden (USA), Blackwell Science Ltd. pp 168-181
SAWYER D. C., GUIKEMA A. H. & SIEGEL E. M. (2004) Evaluation of a new oscillometric blood
pressure monitor in isoflurane-anesthetized dogs. Veterinary Anaesthesia and Analgesia 31, 27-39
SHARN VETERINARY INC. (2008) http://www.sharnvet.com/CustomerService/faqs.aspx. Accessed
11/24, 2008
STEPIEN R. L. & RAPOPORT G. S. (1999) Clinical comparison of three methods to measure blood
pressure in nonsedated dogs. Journal of the American Veterinary Medical Association 215, 16231628
UILENREEF J. J., MURRELL J. C., MCKUSICK B. C. & HELLEBREKERS L. J. (2008) Dexmedetomidine
continuous rate infusion during isoflurane anaesthesia in canine surgical patients. Veterinary
Anaesthesia and Analgesia 35, 1-12
VALERIO F., MARISCOLI M. & PETRIZZI L. (2006) Comparative evaluation of the accuracy of
oscillometric and direct methods for arterial blood pressure monitoring during anaesthesia in dogs.
30, 321-323
WAGNER A. E. & BRODBELT D. C. (1997) Arterial blood pressure monitoring in anesthetized
animals. Journal of the American Veterinary Medical Association 210, 1279-1285
26