(Received March 7; revision accepted June 24, 1968.)
Supported in part by a U.S. Public Health Service Grant (HD-00050) from the National Institute of Child
Health and Human Development and by the Association for the Aid of Crippled Children.
N.M.N. is a recipient of a Career Development Award (1-KO3-HD-38867) of the National Institute
of Child Health and Human Development.
ADDRESS: 221 Longwood Avenue, Boston, Massachusetts 02115.
PEDIATRICS, Vol. 42, No. 6, December 1968 934
ON
THE
INDIRECT
DETERMINATION
OF
SYSTOLIC
AND
DIASTOLIC
BLOOD
PRESSURE
IN
THE
NEWBORN
INFANT
A Simple
Bedside
Electronic
Oscillometer
Nicholas M. Nelson, M.D.
The Department of Pediatrics, Harvard Medical School, and the Boston Hospital for Women, Lying-in Division, Boston
ABSTRACT. Classical osciilometr was used to de-termine indirect systolic/diastolic blood pressure in nine newborn infants. The measurement was facili-tated by a relatively simple and inexpensive elec-tronic osciilometer, the construction of which is de-scribed. Comparisons between indirect leg and
di-rect intra-aortic blood pressure revealed a systolic difference of 0 to 5 mm Hg and a diastolic differ-ence of 0 to 10 mm Hg, depending on respiratory variations and accuracy of calibration. Pediatrics,
42:934, 1968, NEWBORN, BLOOD PRESSURE, OSCIL-LOMETBY.
T
HE SIMPLICITY of the auscultatorymethod for the indirect estimation of
arterial blood pressure and its widespread
acceptance on that account have obscured
the facts that it is not very precisely
cone-lated with intra-arterial pressures,1’2 that
the criteria for diastolic pressure are still
disputed, and that it can be totally
unreli-able in low flow states.3 Previous to
Korotkoff’s description in 1910 of the
arte-rial sounds during decompression,
oscillom-etry was the standard method for clinical
determinations of blood pressure; the two
designs of mechanical oscillometers
cur-rently available (Pachon and von
Recklin-hausen) stem from that era.4 The method
has been shown to yield accurate estimates
of intra-arterial pressure.4’5 As suprasystolic
tissue pressure around an artery is
de-creased during decompression of the
con-stricting cuff, the artery will expand slightly
as a systolic peak of the pressure pulse
ar-rives at the constriction.4 This is shown in
the second line of Figure 1. (For
illustra-tive purposes in this and the subsequent
two figures, the quiescent middle finger of
the author was substituted for a restless
in-fant extremity in the apparatus to be
de-scribed.
)
As tissue pressure is furtherde-creased, a proportionately greater fraction
of the pressure pulse appears until the
di-erotic notch becomes evident. As tissue
pressure reaches diastolic, the oscillations
of the arterial wall become maximal, since
the artery is fully open at systole but fully
closed at diastole. At subdiastolic tissue
pressure (last line, Fig. 1), the artery is
never fully closed so that the oscillations
wane in amplitude (Marey’s principle).
This waxing and waning of arterial wall
os-cillations is what the observer sees
im-pressed on the mercury manometer column
during standard sphygmomanometry and is
illustrated in the upper portion of Figure 2.
Unfortunately, even at suprasystolic cuff
pressures, the centrifugal pressure pulse
impinges on the proximal edge of the
compression cuff and is transmitted to the
system, thereby giving rise to a spurious
in-dication of oscillation of the arterial wall
beneath the cuff. This problem was long
ago solved by placing a second compression
cuff proximal to the main compression and
pressure at all times but not in pneumatic
continuity, the central pressure pulse is
in-tercepted before reaching the main cuff.
The effect of such a proximal cuff in
lower-ing the apparent indicated systolic pressure
is shown in the lower portion of Figure 2.
In the plethysmographic or palpatory
method some sort of pulse indicator
(fingers, strain gage, plethysmograph at
constant pressure) is placed distal to a
compression cuff. As the proximal artery is
decompressed below systolic, the first crest
of the pulse wave passes distally and is
sensed by the pulse indicator. The distal
pulse subsequently builds up to maximal
(and thereafter constant) amplitude as
compression cuff pressure decreases below
diastolic. This is shown in Figure 3.
All of these methods have beell
at-tempted in the Ile\vborn and young infant
with varying degrees of success; very few
have been validated by comparisons with
intra-arterial measurements. In view of the
prominence of arterial hypotension in the
respiratory distress syndrome7 and the
re-current clinical problems where accurate
knowledge of pulse pressure would be
helpful, it seemed important to attempt
de-velopment of relatively simple and
econom-ical apparatus for the indirect l)edside de-termination of systolic and diastolic
pres-sure and to validate its performance. The
oscillometric method was selected, and an
electronically amplified oscillometer, whose
accuracy has been validated as satisfactory,
was developed.
MATERIAL AND METHODS
The present apparatus comprises two
cuffs applied to an extremity and most
con-veniently separated by the knee or elbow.
The pneumatic system allows use of the
cuffs separately or together as an effectively
continuous single cuff. Tile two cuffs may
be decompressed together (oscillometry) or
the distal sensing cuff may be held at
con-stant subdiastolic pressure while the
proxi-mal occluding cuff is decompressed
(palpa-tory method). This latter technique is a
convenient and rapidly repeatable
proce-FIG. 1. Oscillometric tracings from an adult finger at occlusion pressures
de-creasing from 130 mm Jig (to1)) to 70 mm Hg in 10 mm steps. Pressure
936 BLOOD PRESSURES OF NEWBORN
TABLE I
COMPARATIVE BLOOD PRESSURE MEASUHEMENTS
IN NEwB01IN INFANTS
. Subject Number Blood Pressure Aorlzc Blood Difference-indirect rressvre ilznus Aortie
(,,tiit hg) (iiun JIg)
Graphic Recording* 5-6 4.3 35-42 0.3 46-55 5.5 1.8 46-54 6.1 3 2.’2 4-52 7.4 4 10.4 60-70 0 5 -40-45 1 Visual lionitoringt 60 6
-40 47 7 -27 46 8 -‘24 60 9 -38 0 -0 0 -3 2 0 -2 -2
* These were non-distressed infants of diabetic
mothers delivered by elective cesarean section at 37
weeks’ gestation and studied at less than H hours of age.
t These were premature infatits with respiratory
(us-tress syndrome in the first 3 days of life. All survived.
dure for following systolic pressure after
initial determination of systolic/diastolic
pressure by classical oscillometry.
The pulse indicator is a piezoelectric
crystal whose output is amplified by an
in-expensive operational amplifier so that the
arterial pulsations may be transmitted to a
meter needle whose oscillations may be
in-creased to 3 to 4 cm and are hence easily
visible at onset and maximum. Details of
construction and use are included in the
Appendix. Total cost for pneumatic and
electronic parts is less than $100. The
pneu-matic apparatus can be stored inside a
stan-dard sphygmomanometer case. The
elec-tronic box measures 2 X 3 X 6 inches.
Comparisons of indirect estimations of
blood pressure in the extremities were
made with intra-aortic blood pressure in
nine babies whose aortae were cannulated
for diverse therapeutic purposes. The
can-nulae were side-hole polyvinyl feeding
tubes* passed through an umbilical artery to
the approximate level of the diaphragm
and attached to a Statham P23Db strain
gage. After appropriate calibration the
out-put of the strain gage was recorded visuallyt
in Subjects 6 to 9 or on a direct-writing
Sanborn oscil1ograph in Subjects 1 to 6.
The natural frequency of this latter system
was 15 to 30 Hz with a damping ratio of
0.3-0.4 as determined by explosive
de-compression
(
“pop” technique)
. Usablefre-quency response thus extended to
approxi-mately 10 Hz. Arterial pulsations were
monitored visually with the electronic
oscil-lometer. In order to obviate human reading
errors in some babies, the piezoelectric
pulse indicator was substituted by a
differ-ential manometer whose oscillations were
recorded along with cuff pressure on a
di-rect writing oscillograph.
RESU LTS
The results are shown in Table I. The
graphic recordings compared arterial
oscil-lations of the right leg (impressed by the
pneumatic system described below on a
dif-ferential manometer acting as the pulse
transducer) with an intra-aortic pressure
tracing. Cuff pressure and true aortic blood
pressure were recorded from separate strain
gages. Since reflection of standing pressure
waves normally produces some
amplifica-o No. 5 Fr, 15 to :30 cm, Pharmaseal, Glendale, California, and Argyle, St. Louis, Missouri.
f
A-V Monitor, Med-Science, St. Louis, Missouri. Hewlett-Packard, Waltham, Massachusetts.Sanborn Differential Gas Tranducer,
I)iastolie
l,1uni1e? Clinical Disadvanhiges Method
Os(illornetric
Ausetiltatory
Palpatory
Pletliysmograpliic
Flush
Yes
Arguable
No
Yes
N
Amplification often necessary.
Very difficult in newborn.
Satisfactory for systolic only. Encessive instrumentation;
serious systolic overestimate.
Crude correlation, rapid
repeatability awkward. FIG. 2. Oscillometric tracings from the adult finger. Top, “single cuff”
sys-tefll showing spuriously early systolic indication (see text). Bottom, isolated
double cuff system-proximal pulse is obliterated when proximal cuff is iso-lated from distal cuff (between large spikes). Note respiratory variations in pulse amplitude. In both tracings the diastolic shift of tile dicrotic notch2’
and decrease in pulse amplitude
(
Marev’s principle) are evident.tion of pulse pressure in the peripheral
arterial tree, the systolic discrepancy
be-tween central and femoral arterial pressure
was anticipated. Ideal comparisons between
direct femoral pressure and oscillometric
determinations were not attempted in these
infants.
The visual monitoring system employed
the oscillometer precisely as described in
the Appendix and these indirect readings
(also from the right leg) were compared to
direct aortic pressure as monitored by a
very stable strain-gage electronic readout
device. II The apparent improvement in
correlation of central direct and peripheral
indirect arterial blood pressure may
repre-sent more stable calibration or perhaps is
due to visual averaging by the observer
during respiratory variations of the
periph-eral pulsations (only one reading was
re-ported per infant, since repeat
determina-tions varied less than 2 mm Hg).
Alterna-H
Statham P23Db Strain Gage. I lato Rev, PuertoRico, and A-V Monitor, Med-Science, St. Louis,
Missouri.
tively, since the cardiovascular status of
in-I ants with respiratory distress often
in-cludes hypotension and poor peripheral
flow, these factors could have served to
abolish the normal peripheral amplification of pulse pressure in these infants who were mildly ill.
DISCUSSION
The available methods for indirect blood
pressure estimation are presented in Table
II. At one time or another all have l)een at-tempted in infants. The flush method is
per-TABLE II
938 BLOOD PRESSURES OF NEWBORN
FIG. 3. “Palpatory” tracing. The distal cuff when held at constant subdiastolic pressure allows registration of the standard plilebogram seen in lower line trace at a higher sweep speed. When proximal occlusion cuff is bled down from supra-systolic levels, the onset of pulsations marks systolic pressure.
Again, note respiratory variations in pulse amplitude.
haps most widely used but it is difficult to
repeat rapidly, gives no indication of pulse
pressure, and is only fitfully correlated with
mean blood pressure.
The plethysmographic method has
en-joyed recent pediatric resurgence in both
standard8 and strain gage forms, despite its
known overestimation of systolic pressure.’#{176}
This is shown in Figure 4 where the
pleth-ysmographic and oscillometric methods are
compared. The increase in limb volume
oc-curring at a suprasystolic pressure and
be-fore the onset of oscillations is probably the
result of arterial inflow to the limbs through
arteries buried within or near bones and
hence incompressible by the occluding cuff.
Since vascular resistance is calculated as
the ratio of blood pressure/blood flow, any
determinations of resistance based on
pleth-ysmographic blood pressure
determina-tions’1’12 must necessarily be overestimates.
The palpatory method for systolic
pres-sure has been notoriously unreliable where
human fingers have been the sensors. But
when the fingers have been supplanted by
a sensitive pulse indicator,13 accuracy,
re-peatability, and simplicity have followed.
Indeed, the only defect of the
Ashworth-Neligan apparatus has been its inability to
provide a reliable estimate of diastolic
pres-sure. The apparatus described in the
pres-ent report can be used, if desired, as just
such a palpatory sensor with the
advan-tages of rapid repeatability and
considera-bly more visibility than the standard xylol
bead pulse indicator.
Experience with auscultatory
sphygmo-manometry has been uniformly
disappoint-ing in infants. Although extreme observer
dedication and subject cooperation have
oc-casionally combined to produce a reading,
the method has never proven reliable.14
Most probably the critical variables of
arte-rial wall tension, pressure drop across the
occluding cuff, and flow rate through the
compressed artery3 are insufficiently high in
the newborn infant to cause the regular
culminat-939
ing in Korotkoff’s sounds. Indeed, the poor
correlation of Korotkoff’s sounds with
in-tra-arterial measurements in adults’ ‘‘
bears testimony to their general
unreliabil-ity (especially at diastolic pressure).
De-spite these problems, electronic
amplifica-tion and recording of auscultatory blood
pressure has been used in babies without
the necessary intra-arterial validation.’
The oscillometric method has long been
used in newborns,1619 but persistent defects
have been instrumental insensitivity and
lack of intra-arterial validation. The present
results would seem to establish the
oscillo-metric method’s reliability. In animal and
adult human material,’ intra-arterial
com-parisons have long since demonstrated the
clear superiority of the oscillometric over
other indirect methods. Its total
replace-ment in clinical medicine by the
ausculta-tory method is, of course, the result of the
latter’s extreme simplicity.
This advantage does not apply to the
mm Pig
newborn so that the instrumentation
re-quired in the present method can perhaps
be excused. Indeed, the insufficient
sensitiv-ity (in the author’s hands) of commercially
available mechanical oscillometersl9 has
prompted the present work. Once diastolic
pressure has been determined, the
de-scribed apparatus may be used as a
palpa-tory sensor, for which purpose it offers no
advantage over the Ashworth-Neligan
ap-paratus except readability.
In all methods, proper cuff size is
per-haps the most important factor in achieving
a true approximation of intra-arterial
pres-sure. The standard recommendation for cuff
width in adults is about one-third the limb
circumference, or approximately 2 to 3 cm
for a newborn infant. Many studies have
substantiated this flgure,13 so 2.5 or 3.0 cm
cuffs were used throughout the present
study.
The two main sources of error in the use
of the apparatus are observer error and true
FIG. 4. Comparison of plethysmographic and oscillometric methods of indi-rect blood pressure estimation in the right leg of an infant with a direct aortic blood pressure of 60/35. Top: cuff pressure. Center: mercury-in-rubber plethysmographic tracing. Bottom: oscillonetric tracing. Plethysmo-graphic blood pressure is 81/45 (onset/offset of limb volume increase). Oscillometric blood pressure is 69/31 (appearance/diminishment of limb
A
CD
FIG. 5. Pneunlatic circuit (see Appendix for explanation).
10 K
940 BLOOD PRESSURES OF NEWTBORN
Fic. 6. Electronic circuit (see Appendix for explanation).
respiratory variations in peripheral blood
pressure. The observer errors can be held to
less than 2 to 5 mm Hg by using a slow
leak and frequent interruptions of mercury
descent with a check valve (Fig. 5, tap B).
Respiratory variations can amount to about
6 mm Hg. Consistent recording of that
oc-cluding pressure at which any (rather than
all) pulsations show the proper criterion
(appearance of pulse or its waning) will at
least permit consistent determinations of
systolic/diastolic pressure.
It is of perhaps more than simply
nostal-gic interest that in 1913 Balard made
ac-Ima
APPENDIX
Construction Details
sold for transistor radios. Its output is shunted ELECTRONIC (schematic in Fig. 6): The pulse across an RC low pass filter in order to
attenu-transducer is an inexpensive crystal earphone ate the unwanted higher frequencies which
curate measurements of systolic/ diastolic
blood pressure in normal and low birth
weight infants, including a description of
the slow postnatal rise in blood pressure
only recently accepted as standard.’#{176} Balard
used the oscillometric method, which
unac-countably later fell into clinical disuse. The
present validation and instrumentation will
perhaps encourage and permit the
renais-sance which the oscillometric technique so
richly deserves.
SUMMARY
Classical oscillometry has been employed
to determine indirect systolic/diastolic
blood pressure in newborn infants. A
rela-tively simple and inexpensive electronic
Os-cillometer, which facilitates the
determina-tion, is described. Comparisons between
in-tra-aortic and (indirect-direct) leg blood
pressure in nine babies revealed a systolic
difference of 0 to 5 mm Hg and a diastolic
difference of 0 to 10 mm Hg, depending on
the accuracy of calibration.
REFERENCES
1. Roberts, L. N., Smiley, B. A., and Manning, C. W.: A comparison of direct and indirect blood pressure determinations. Circulation,
8:232, 1953.
2. Simpson, J. A., Jamieson, G., Dickhaus, D. W., and Grover, R. F.: Effect of cuff bladder on
accuracy of measurement of indirect blood
pressure. Amer. Heart J., 70:208, 1965.
3. Cohn, J. N.: Blood pressure measurement in shock. J. A. M. A., 199:972, 1967.
4. Burch, G. E, and DePasquale, N. P.: Primer
of Clinical Measurement of Blood Pressure. St. Louis, Missouri: C. V. Mosby Co., 1962. 5. Van Bergen, F. H., Watherhead, D. S.,
Tre-bar, A. E., Dobkin, A. B., and Buckles’, J. J.: Comparison of indirect and direct meth-ods of measuring arterial blood pressure. Circulation, 10:481, 1954.
6. Christensen, B. C.: Oscillometric studies. Acta Med. Scand., 120:474, 1945.
7. Rudolph, A. M., Drorbaugh, J. E., Auld, P. A. M., Rudolph, A. J., Nadas, A. S., Smith, C.
A., and Hubbell, J. P.: Studies on the circu-bation in the neonatal period. The circulation in the respiratory distress syndrome. PirnT-IIICS, 27:551, 1961.
8. Celander, 0., and Thunell, G: A plethysmo-graphic method for measuring the systolic and diastolic blood pressure in newborn in-fants. Acta Paediat. Scand., 49:497, 1960. 9. Levison, H., Kidd, B. S. L., Gemmell, P. A.,
and Swyer, P. R. : Blood pressure in normal
full-term and premature infants. Amer. J.
Dis. Child., 111:374, 1966.
10. Landowne, M., and Katz, L. N. : A critique of the plethysmographic method of measuring blood flow in the extremities of man. Amer. Heart J., 23:644, 1942.
11. Kidd, L., Levison, H., Gemmell, P., Aharon, A., and Swyer, P. R. : Limb blood flow in the normal and sick newborn. Amer. J. Dis.
Child., 112:402, 1966.
12. Celander, 0. : Blood flow in the foot and calf of the newborn. Acta. Pediat. Scand., 49:488, 1960.
13. Ashworth, A. M., Neligan, C. A., and Rogers, J. E.: Sphygmomanometer for the newborn.
Lancet, 1:801, 1959.
14. Moss, A. J., and Adams, F. H.: Index of indi-rect estinlation of diastolic blood pressure.
Amer. J. Dis. Child., 106:364, 1963.
15. Morse, R. 0., Brownell, C. L., and Currens,
J.
H.: The blood pressure of newborn infants.Indirect determination by an automatic
re-corder. PEDIATRICS, 25:50, 1960.
16. Balard, P.: Le pouls et Ia tension arterielle
de l’enfant et du nouveau-ne. Caz. Hop.,
86:837, 1913.
17. Rucker, M. P., and Connell, J. W.: Blood
pres-sure in the newborn. Amer. J. Dis. Child., 27:6, 1924.
18. Bevis, D. C. A., and Schofield, S. F.: The renal
function of infants delivered of toxaemic
mothers. Arch. Dis. Child., 26:109, 1951. 19. Kafka, H.: A simple method for blood pressure
measurements in the premature and
new-born infant. PEDIATRICS, 40:106, 1967. 20. Contis, C., and Lind, J.: Study of systolic
blood pressure, heart rate, body temperature of normal newborn infants through the first week of life. Acta Paediat. Scand. (Suppl.),
146:41, 1963.
942 BLOOD PRESSURES OF NEWBORN
abound in most any clinical acoustic-electronic
environment. Generator impedence is kept
mod-erately low in order to minimize error voltages.
The conditioned signal is then amplified by a
battery-driven operational amplifier0 connected
as a voltage follower with a closed-up gain
capability of 1,000. This gain has proven to be
more than ample to produce full-scale
deflec-tion with the smallest pulses. The output signal is rectified in order to simplify interpretation of
the otherwise complicated wave form as seen
on the meter needle. Because the dicrotic notch
so prominent in the record from an adult finger
(Fig. 1 and 3) is not seen in the infant leg (Fig.
4), rectification of the pulse wave-form was felt to result in no serious loss of information. Input
impedance of the amplifier in this configuration
is approximately 5 megohms and expected
bat-tery life is 1,000 hours.
PNEUMATIC: The design is depicted in Fig.
5
and is an adaptation of that given byChris-tensen.6 The polyvinyl cuffs are commercially
available.
f
Rubber cuffs may well be toocorn-pliant for accuracy in oscillometry. Tubing is of
plastic and latex leftovers from intravenous ap-paratus. The manometer is the standard clinical portable variety. T-joints are small-bore plastic.t
The three-way tap is of disposable plastic, and
the inexpensive “needle valve” is reclaimed from an exchange transfusion set. The crystal
microphone (earphone) is sealed with epoxy
resin on all seams because as purchased it is
insufficiently robust to withstand the required
pressures. The four-way stopcock is machined
from a standard metal three-way stopcock and
a stop is attached so that only a 45#{176}rotation is
possible (i.e., all four ports are in mutual
com-munication or are totally isolated). However, a
less accurate version which is simpler to use substitutes two equal lengths No. 21 gage
cap-illary needles for the four-way tap B. The cuffs
are inflated by the bulb inserted via a T-tube
between the capillaries. The capillaries thus
act as a low pass filter between the cuffs, which
allows a continuous bleed through C (which in
this version must be moved to a position
be-tween the capillaries and the inflating bulb).
#{176}No. 2LV-1, Nexus, Dedham, Massachusetts.
Sage, White Plains, New York. Bolab, Reading, Massachusetts.
Use
PRELIMINARY ADJUSTMENTS: Cuffs are care-fully applied above and below the knee or el-bow and are connected as indicated in Figure
5. With both cuffs in communication (through
taps A and B) tap C is closed, the cuffs are in-flated, tap C is then opened, and the minor
leak at C is adjusted carefully for a bleed of
about 2 to 3 mm Hg per second. This should
rarely need readjustment. The transducer is
plugged into the amplifier, which is then
switched on. With the amplifier at approxi-mately half-gain, the meter needle is brought to slightly above zero with the zero control. The gain control is then turned down to zero. The cuffs are then inflated to approximate mean pressure between expected systolic and diastolic pressures, and the gain is advanced until the needle pulsations are brought to nearly full
scale. This should need only minor
readjust-ment.
OSCILLOMETRIC BLOOD PRESSURE (systolic/ diastolic) : The cuffs are in communication through taps A and B. Tap D is closed and the cuffs are inflated above systolic. By alternately
opening and closing tap B while observing the
meter needle, the cuffs are deflated in 2 to 5 mm steps, and systolic pressure is taken as that at which first pulsations are seen with B closed. Respiratory variations will often be encoun-tered.
After noting the systolic pressure, B is opened,
and the cuffs are bled continuously until pulse
amplitude first begins to wane, at which point
tap B is again closed (if desired) while the dia-stolic pressure is noted. Tap B is again opened and the cuffs are rapidly deflated tllrough tap D.
PALPATORY BLOOD PRESSURE (systolic) :‘
The cuffs are in communication through taps A and B. The cuffs are inflated to a pressure
be-tween systolic and diastolic (or even slightly
below diastolic) and tap A is turned to isolate the distal sensing cuff. The proximal occluding
cuff is deflated through D and B, which latter
is then left in open position. Further rapid sys-tolic readings are made by inflating the
proxi-mal cuff and deflating through D again, noting
onset of pulsations as the systolic pressure