- Email: [email protected]
Changes in superior mesenteric artery Doppler waveform during reduction of cardiac stroke volume and hypotension
Copyright
8
Ultrasound I996 World
in Med. L Bml.. Vol. 72. No. I, pp. I I-IX, lYY6 Federation for Ultrasound in Medicine & Biology Prmted tn the USA. All rights reserved KiOl-562Y/Y6 % I S.(N) + .oO
0301~5629( 95 ) 02037-3
ELSEVIER
l Original Contribution CHANGES IN SUPERIOR MESENTERIC ARTERY WAVEFORM DURING REDUCTION OF CARDIAC VOLUME AND HYPOTENSION
DOPPLER STROKE
M. J. PERKO, G. PERKO, S. JUST, N. H. SECHER and T. V. SCHROEDER Departments of Vascular Surgery, Radiology, and Anesthesia, The Copenhagen Muscle Research Center. Rigshospitalet, The National University of Copenhagen, Denmark (Received
IO April
1995; in jinal form
24
July
1995)
Abstract-Influence of stroke volume reduction and bypotension on the superior mesenteric artery (SMA) Doppler waveform was evaluated during bead-up tilt-induced central bypovolemia in 11 healthy volunteers. During normotensive reduction in stroke volume, peak systolic velocity ( pV) , mean velocity, pulsatility and resistivity indices decreased, while diastolic velocities increased. During hypotension, a further decrease in pV was accompanied by maintained elevation of diastolic velocities and reduction in pulsatility and resistivity indices. Power of backscattered Doppler wave was elevated throughout the hypovolemia. Alterations in pV and pulsatility indices were closely related to changes in stroke volume, and a negative correlation was found between diastolic velocities and stroke volume. Regression analysis showed no signigcant relation between variations in velocity parameters and blood pressure. Results of the study indicate that alterations in stroke volume induce consequential changes in the SMA Doppler waveform. These changes originate from both direct influence of stroke volume and/or pressure on blood flow velocity, and alterations in SMA peripheral resistance that follow variations in stroke volume. Presented interdependencies should be taken into consideration while studying mesenteric physiology with the use of Doppler technique and while interpreting the duplex results in patients suffering from diseases that may influence flow velocity and mimic or obscure Doppler effects of the SMA stenosis. Key Words: Ultrasonics,
Pulsatility
index, Resistivity
index, Blood flow, Hypovolemia.
reduction of the resistivity indices (Bawersox et al. 1991; Moneta et al. 1991; Nicholls et al. 1986). Alterations in the SMA blood flow velocity (and Doppler waveform) occur, however, not only due to local arterial disease. In clinical practice, we meet patients suspected for mesenteric ischemia who suffer dehydration, hypotension, aortic valve stenosis, and congestive heart disease with low cardiac output that can influence mesenteric artery blood velocity and flow. This study was undertaken to detect whether hypovolemic reductions in stroke volume and arterial blood pressure influence SMA Doppler waveforms.
INTRODUCTION Duplex ultrasound of the superior mesenteric artery (SMA) allows a noninvasive measurement of blood flow velocity and volume through the artery in humans (Jager et al. 1986). This simple-to-apply and relatively inexpensive technique gains acceptance as a method for studying mesenteric physiology and recognition of SMA obstructive disease (Lilly et al. 1989; Moneta et al. 1991). Stenosis of the artery is characterized by elevation of the blood flow velocity over the stenotic area, reduction of the velocity caudally to the stenosis and reduction of the peripheral mesenteric resistivity. In ultrasound Doppler examination, these changes are best reflected by increase in peak systolic ( pV), diastolic and mean velocity (MV) over the stenotic area; reduction in velocities caudally to the stenosis; and Address correspondence to: Dr. Mario J. Perko, University Copenhagen, KAS Gentofte, Department R, Niels Andersens 65, DK-2900 Hellerup, Denmark.
MATERIALS
AND METHODS
Eleven healthy, fasting ( ~-8 h), nonsmoking and nonmedicated volunteers with a mean age of 26 ( 1932) y, weight 65 (48-85) kg, height 175 (164-187) cm and a male-to-female ratio of 8:3, underwent duplex ultrasonographic examination of the SMA during passive head-up tilt. The study was approved by the
of Vej
11
I7
Ultrasound
in Medicine
and Biology
Volume
22. Number
I. 1996
Ethical Committee of Medical Research in Copenhagen, and all subjects gave informed consent in accordance with the Helsinki II Declaration.
with no footboard. Leg movements were not allowed to reduce venous return (Matzen et al. 199 1). Throughout the study, all variables were recorded every minute.
Study design The investigation was comprised of four stages: 1) 3Omin supine rest; 2) passive head-up tilt to 30” for 2 min; 3) increasing the head-up tilt to 50” and sustained tilt until presyncopal symptoms appeared (dizziness, nausea, pallor and sweating); and 4) supine recovery until stabilization of cardiovascular variables. During the head-up tilt, the subjects were supported by a bicycle saddle but
Equipment and variables Impedance cardiography is a noninvasive, reliable method for estimation of alterations in fluid volume, heart rate, stroke volume, cardiac output and cardiac index, which correlates closely with thermodilution technique (Bernstein 1986; Clancy et al. 1991 1. The method is based on a mathematical analysis of both thoracic volume and the rate of change of thoracic
A
8
”
4
8
4
I
Fig. 1. Doppler waveform of the superior mesenteric artery: (a) at rest, and (b) during stroke volume reduction and hypotension (presyncope) . pV, peak systolic velocity; pdV, peak diastolic velocity; minV, minimal velocity; MV, mean height over the cardiac cycle; S, maximum height of the waveform; T, length of the cardiac cycle; ts, duration of the systolic peak (measured between half-amplitude points).
Superior mesenteric artery changes 0 M. J. PERKO et al.
electrical impedance during the cardiac cycle. We used a Simonsen & Wee1 (Copenhagen, Denmark) impedance cardiomonitor which records Zo at 100 kHz and calculates stroke volume from the beat-to-beat pulsatile changes in thoracic impedance (Kubicek et al. 1966; Matzen et al. 1990). Changes in Zo reflect alterations in intrathoracic fluid volume (Perk0 et al. 1991) _ Systolic and diastolic arterial blood pressures were measured by sphygmomanometry on the left arm and mean arterial pressure (pulse pressure 3 -’ + diastolic pressure ) was calculated. Ultrasound duplex evaluation was conducted with an Ultramark 9 (Advanced Technology Laboratories, Bothell, WA, USA) using a 2.5-MHz phased-array probe. In this system, the elements of the array are used to generate a short-gate pulsed-wave Doppler beam that can be steered to the desired location in the sector scanned with constant control of the angle of insonation. Estimation of blood flow by means of the duplex scan carries some potential errors (Evans 1989). To reduce random errors and minimize variability and relative inaccuracies during assessment of the Doppler parameters, all measurements were standardized as follows: the SMA was examined 2-3 cm distally from its origin, at a 60” angle of insonation (the angle between ultrasound beam and the long axis of the vessel), and at a depth of 4-8 cm; samples of ultrasound beam were placed entirely within the artery; high-pass filters were used to remove low-frequency (O-100 Hz) signals arising from the vessel wall; repetitive measurements were collected from four to six cardiac cycles (Perk0 and Just 1993 ) . The Doppler parameters recorded included blood flow velocities: peak systolic (pV>, minimal (minV ) , peak diastolic ( pdV) , end-diastolic (edV) and mean (MV) ; and their derivatives: pulsatility index [PI = (pV - minV) MV -’ 1, resistivity index [RI = ( pV - minV) pV -’ 1, constant flow ratio [PI (RI - PI))‘] and height width index [PI (Tt,;‘)], where T is the time of a single wave and t, is the duration of the systolic peak (Evans et al. 1989) (Fig. 1). The power (decibels) of backscattered Doppler wave was measured off-line from digitized Doppler spectral data. Diameter of the artery was assessed at long arterial axis display and, for standardization of the procedure, during systole. MV was determined by off-line planimen-y. The SMA blood flow volume was determined as MV &r2, and systemic vascular resistance was the ratio between mean arterial pressure and cardiac output. Duplex scan examinations were videotape recorded for computerized graphical off-line processing in a software program written in C. Calculations
Comparative statistics within the individual parameters were made for time-specific points (from rest to 8
13
min after appearance of presyncopal symptoms) of all examinations as shown in Fig. 2. Data were first analyzed using one-way analysis of variance for repeated measurements, and if the variance of a parameter was significant, differences between time-specific points were evaluated by a post hoc multiple comparison test ( Newman-Keuls test). Uni- and multiple regression analyses were used to detect interdependencies between the parameters. The p of <0.05 was considered significant. Results are given as meant standard error or with range. RESULTS
The course of selected variables and Doppler indices during head-up tilt is presented in Fig. 2. Alterations in cardiac output and cardiac index corresponded closely to changes in stroke volume. The onset of presyncopal symptoms occurred within 16 (5-47) min of head-up tilt. Return to the supine position initiated immediate recovery and normalization of circulatory and Doppler variables within a few minutes. Systemic hemodynamics
Continuous significant increase in thoracic impedance throughout the head-up tilt from 35.0 -+ 0.7 s1 at rest to 38.0 t: 0.7 R at presyncope (p < 0.01) indicated progression of central hypovolemia. Until 1 min before occurrence of presyncopal symptoms, in spite of tachycardia, a successive decrease in stroke volume produced reduction in cardiac output from 5.7 2 0.4 to 3.9 + 0.3 L min-’ (p < 0.01). A simultaneous increase in systemic vascular resistance from 16 t 2 to 22 + 4 mmHg min L-’ (p < 0.01) maintained arterial pressure. Occurrence of presyncopal symptoms was associated with a decrease in arterial pressure and further reduction in stroke volume. Systemic vascular resistance and heart rate decreased, re-establishing pretilt level (relative bradycardia) . SMA ultrasound parameters
Peak systolic velocity decreased throughout the head-up tilt from 211 + 14 cm s ’ at rest to the lowest value of 87 + 5 cm s-’ at presyncope (p < 0.01). The diastolic velocities increased gradually from rest to presyncope as follows: minV from -9 + 10 to 28 + 2cms-‘; pdV from 19 + 5 to 34 -+ 2 cm ss’; and edV from 19.2 to 32.6 2 2.3 cm s-’ (p < 0.01). Mean velocity decreased from 54 ? 6 cm s-’ at rest to 33 2 1 cm s-’ at presyncope (p < 0.01) . This reduction in MV was entirely due to decrease in mean velocity of the systolic phase of cardiac cycle (Fig. 2). Opposite changes in systolic (decrease) and diastolic (increase) velocities resulted in conversion of the triphasic Doppler waveform to bi- or monophasic during presyncope
Ultrasound
in Medicine
and Biology
Volume
22. Number
1. 1996
0.8 a
0.6
0I -1.2 &
-,::
““yz,
U
-1.6-
a
-1% -2-
OJ Dia.&c
Fig. 2. Cardiovascular and Doppler variables during head-up tilt at rest, tilting to 30” and 50”, maintained 50” head-up tilt for 8 min before presyncopal symptoms, during the onset of presyncopal symptoms, and in the recovery period. (0) Significant difference with rest; (0) significant difference from preceding value. MAP, mean arterial pressure; pV, peak velocity; minV, minimal velocity; MV, mean velocity; PI, pulsatility index; RI, resistivity index; HWI, height width index; CPR, constant flow ratio.
(Fig. 1) . Return to the supine position induced a brief (several seconds), rapid increase in pV to nearly double of the resting value, followed by fast stabilization at the pretilt level. PI and RI decreased gradually throughout head-up tilt, reaching the lowest values at presyncope. The diameter of the SMA increased shortly after initiation of 50” head-up tilt from 5.8 ? 0.2 to 6.9 + 0.4 cm (p < 0.05)) and returned to resting
value within 2 min. Blood flow decreased temporarily during presyncope from 0.88 t 0.13 L min-’ at rest to 0.41 + 0.03 L mitt-’ (p < 0.05). Power of backscattered Doppler signals increased throughout the headup tilt and was maintained elevated during the first 4 (2- 10) min of recovery. Mean relative changes (A between the rest and sympathoexcitatory phase of hypovolemia) in power of specific frequencies are given
Superior mesenteric artery changes 0 M. J. PERKO
in Table 1. An example of a Doppler spectrogram obtained from a single subject at rest and during hypovolemic reduction in stroke volume is presented in Figs. 3 and 4. Relations between variables
Mean arterial pressure, pulse pressure and stroke volume were selected as independent variables for regression analyses. Univariate analysis indicated no significant relation between alterations in velocity parameters and arterial pressure, and no relation between changes in blood velocity and pulse pressure (Table 2 ) . On the other hand, a strong positive correlation was found between systolic velocity and stroke volume, and a negative correlation between diastolic velocities and stroke volume. In the forward stepwise multiple regression, arterial pressure was rejected from most of the models, while stroke volume had a positive association with all Doppler indices (Table 3 ) . DISCUSSION The results of the present study indicate that: 1) SMA systolic blood flow velocity can be reduced by about 50% during reduction in inflow volume (stroke volume) despite maintained blood pressure; 2) SMA systolic velocity can be reduced by about 60% during severe reduction in stroke volume inducing hypotension; 3) SMA diastolic velocity increases during inflow volume reduction-most probably due to reduction in peripheral mesenteric resistance; 4) SMA backscattered power of the Doppler signal increases during reduction in stroke volume and hypotension; 5) SMA Doppler-derived frequency indices can be influenced by changes in stroke volume. Systemic hemodynamics
During the first stage of head-up tilt, the so-called sympathoexcitatory phase of central hypovolemia, the decrease in stroke volume opposed the increase in systemic vascular resistance, which maintained arterial pressure. During the second sympathoinhibitory phase of central hypovolemia, the decrease in stroke volume was accompanied by a reduction in systemic vascular resistance
Table 1. Relative change (A from rest to hypovolernia) in the power of the highest frequencies present. Phase
A Frequency power (dB)
Peak systolic Minimal diastolic Peak diastolic End diastolic
1.8 -+ 0.7 1.4 2 0.2 4.6 t 0.6
3.1 2 0.8
P <0.005 <0.0005
er al.
15
resulting in hypotension. These observations correspond well with earlier findings (Pawelczyk et al. 1994 ). Doppler waveform parameters
Decreases in pV and MV during the normotensive phase of hypovolemia indicated an influence of the inflow (to the SMA) volume on these two parameters. This observation is confirmed by regression analysis, which revealed dependency of pV and MV on changes in stroke volume. Simultaneously, all of the diastolic velocities (minV, pdV, edV) increased, indicating reduction of the peripheral mesenteric resistance. This reduction in wave amplitude resulted in decreases in PI and RI, while blood flow did not change. During the hypotensive phase of hypovolemia, decreases in pV and MV were accompanied by maintained elevation in diastolic velocities resulting in continuous depletion of wave amplitude-and a further reduction in PI and RI. Thus, the present study indicates that pulsatility indices are influenced by both inflow volume (stroke volume) and regional peripheral resistance. This deduction is valid on the condition that increase in diastolic velocity truly reflects decreasing peripheral resistance. An alternative explanation for increase in diastolic velocity could have been local vasoconstriction, but this is unlikely as vasoconstriction would increase pV and MV, and reduce diameter of the vessel, neither of which took place. Power of the ultrasound signal backscattered by blood depends on transmitted frequency power, presence of red blood cell aggregation, compressibility and density of red cells and plasma, hematocrit, packing factor and volume of the scatters (Shung et al. 1993). The present study design allows the assumption that all but last variable were constant during examinations, thus the power of the reflected Doppler wave within a particular frequency depended predominantly on the number of blood cells moving with velocity producing this frequency. Thus, accentuation of the power of the diastolic frequencies, as seen in Figs. 3 and 4, reflected increased volumetric blood flow during the diastolic phase of the cardiac cycle. One can argue that increased power, and therefore increased flow volume, should be accompanied by local increase in diameter of the artery, which was not observed in the present study. The possible explanation may be that volumetric blood flow is proportional to the squared diameter, thus minor changes in diameter of an artery may produce remarkable changes in blood flow volume. Head-up tilt-induced changes in height width index resembled that of PI, however, the variance did not reach statistical significance. This “smoothing” of height width index originated from a shortening of the systolic phase and a relative elongation of the diastolic phase of the
Ultrasound
in Medicine
and Biology
Volume
22, Number
I. 1996
Frequency(Hz) Fig. 3. An example of spectral content and power of frequencies over a single cardiac cycle: (A) at rest and during (B) normotensive reduction in stroke volume.
Doppler wave. Thus, height width index takes into account prolonged diastolic flow and in that way is more “sensitive” to increased diastolic velocity. Inclusion of the systolic-to-diastolic velocity ratio in an index equation makes the index independent of the angle of insonation. On the other hand, a disadvantage is that the index becomes dependent on both inflow volume and peripheral resistance. In situations in which both of these factors may alter or deviate from
“normal” due to diseases (e.g., heart insufficiency + peripheral arterial stenosis), evaluation of peripheral resistance by means of Doppler ultrasound may be erroneous. In clinical practice, erroneous interpretation of mesenteric Doppler examination may be verified by X-ray arteriography. In physiological studies, however, verification of Doppler ultrasound observations is, for ethical reasons, very difficult. Lilly et al. (1989) re-
Superior 8 6 4
A
mesenteric
artery
I
I
I
6A
early diastolic (239 ms)
4 2
I
I
I
I
I
8
6A
peak diastolic
4
:& R
6A
end diastolic (864 ms)
4
6 4 2 0 8 6 4 2
B
MAP
,
8
peak systolic
17
er al.
Table 2. Univariate regression analysis.
peak systolic (67 ms)
2 o-
o- !
changes 0 M. J. PRRKO
Pulse pressure
Parameter
r
P
Peak systolic V Minimal V Peak diastolic V End diastolic V Mean V-systolic Mean V-diastolic Mean V Pulsatility index Resistivity index HWI
0.08 -0.01 -0.04 -0.04 -0.02
NS NS NS NS NS
CFR Diameter Blood flow
-0.21 < 0.05 0.06 NS 0.10 0.14 0.08
NS NS NS
0.33 0.33 0.17
e < NS
r
P
0.23 < 0.03 -0.25 < 0.01 -0.14 NS -0.14 NS 0.11
NS
-0.17 NS 0.04 NS 0.21 < 0.05 0.27 e 0.20 NS 0.33 NS 0.31 e 0.24 < 0.02
Stroke volume r
P
0.68 -0.48 -0.41 -0.41
-3 4 e e
0.44
e
-0.33 6 0.27 < 0.51 < 0.63 < 0.34 Q 0.46 -S -0.15 NS 0.11 NS
MAP, mean arterial pressure; V, velocity; HWI, height width index; CFR. constant flow ratio; G. p < 0.001; NS, not significant; r, correlation coefficent.
(83 m.9 I
I
I
1
B
O_rflI
I
I
I
early diastolic (232 ms) I
I
I
I
,
/
peak diastolic
0
516 1033 1550 2067 2583 3100 Frequency(Hz)
Fig. 4. Relative change in the power of frequencies from Fig. 3: (A) at rest and during (B) normotensive reduction in stroke volume.
ported about 40% increase in pV after glucagon infusion and suggested that pharmacological stimulation by glucagon may substitute for the physiologic stimulation of a meal. However, the diastolic velocity did not increase significantly after glucagon infusion in contrast to the meal response (Lilly et al. 1989). In the light of the present study, pV is predominantly related to the inflow alterations (stroke volume), while diastolic velocities are rather inflow-independent. Thus, the response pattern to glucagon infusion may represent changes in cardiac output rather than peripheral mesenteric vasodilation. PI is the ratio between velocity amplitude and MV. In the present study, MV decreased during hypo-
volemia and hypotension, thus the decrease in PI was entirely due to amplitude reduction. Qamar et al. ( 1986) reported a 46% decrease in PI following ingestion of a meal and related this observation to vasodilation in the splanchnic bed and decreased downstream resistance. The authors did not indicate whether the fall in PI was a result of depletion of wave amplitude or an increase in MV, but from another investigation (Jlger et al. 1986) it is known that a postprandial moderate increase in amplitude (from 120.0 to 125.6 cm s-’ ) is accompanied by a more than twofold increase in MV (from 22.2 to 57.0 cm s-l). Thus, in this study, the postprandial decrease in PI was an effect of increased MV. The increase in MV could be due to reduction in peripheral resistance, but it could also be due to postprandially increased cardiac output, and most probably due to both. Another possibility is that reduction in peripheral resistance may be due to arteriolar dilation, but it could also be due to decrease in blood pressure. CONCLUSIONS Significant changes in the SMA Doppler waveform may occur during alterations in stroke volume and blood pressure. These Doppler waveform changes may originate from direct influence of inflow (to the SMA) volume and/or pressure on blood flow velocity, but may also occur due to alterations in peripheral resistance that follow variations in cardiac output. These interdependencies should be taken into consideration while studying mesenteric physiology with the use of the Doppler technique and while interpreting the SMA duplex results in patients suffering from hy-
Ultrasound in Medicine and Biology Table 3. Multiple
Volume 22, Number I. 1996
regression analysis (forward
MAP Dependent parameter
P
Peak systolic V Minimal V Peak diastolic V End diastolic V Mean V-systolic Mean V-diastolic Mean V Pulsatility index Resistivity index HWI CFR Diameter Blood flow
rejected 0.17 (0.11) rejected rejected rejected -0.16 (0.10) rejected rejected rejected rejected 0.27 (0.09) 0.22 (0.11) rejected
stepwise method).
Pulse pressure P
NS
NS
< : 0.01 NS
B
rejected -0.20 (0.11) rejected rejected rejected rejected rejected rejected rejected rejected rejected 0.28 (0.12) 0.24 (0.10)
Stroke
I’
P
NS
-0.45 (0. IO) -0.41 (0.10) -0.41 (0.10) 0.44 (0.09) -0.30 (0. IO) 0.27 (0.10) 0.51 (0.09) 0.63 (0.08) 0.34 (0. IO) 0.42 (0.09) -0.27 (0. IO) rejected
volume
I’
0.68 (0.08)
< 0.03 < 0.03
MAP, mean arterial pressure; V, velocity; HWI, height width index; CFR, constant flow ratio; <. p < 0.001; NS, not significant; p, standardized regression coefficient.
po- and hypertension, hypo- and hyperthyroidism, cardiac insufficiency and other diseases that may influence cardiac output and/or pressure. work was supported by Gerda and Aage Haensch’s Fond; Fonden til Sygdmomsbekrempelse uden Dyreforseg; and Ferd. & Ellen Hindsgauls Fond.
Acknowledgements-This
REFERENCES Bernstein DP. Continuous noninvasive real-time monitoring of stroke volume and cardiac output by thoracic electrical bioimpedance. Crit Care Med 1986; 14:898-901. Bowersox JC, Zwolak RM, Walsh DB, Schneider JR, Musson A. Duplex ultrasonography in the diagnosis of celiac and mesenteric artery occlusive disease. J Vast Surg 1991; 14:780-788. Clancy TV, Norman K, Reynolds R, Covington D, Maxwell JG. Cardiac output measurements in critical care patients: thoracic electrical bioimpedance versus thermodilution. J Trauma 1991;31:1116-1119. Evans DH, McDicken WN, Skidmore R, Wodcock JP. Doppler ultrasound. Physics, instrumentation, and clinical applications. New York: Wiley, 1989. Jager K, Bollinger A, Valli C, Amman n R. Measurement of mesenteric blood flow by duplex scanning. J Vast Surg 1986;3:462469. Kubicek WG. Karnegis JN, Patterson RP, Witsoe DA, Matterson
RH. Development and evaluation of an impedance cardiac output system. Aerospace Med 1966;37:1208- 1212. Lilly MP, Harvard TRS, Flinn WR, Blackburn DR, Astleford PM. Duplex ultrasound measurement of changes in mesenteric flow velocity with pharmacologic and physiologic alteration of intestinal blood flow in man. J Vast Surg 1989;9:18-25. Matzen S, Perk0 G, Groth S, Friedman DB, Secher NH. Blood volume distribution during head-up tilt induced central hypovolaemia in men. Clin Physiol 1991; 11:41 I-422. Moneta GL, Yeager RA, Dalman R, Antonovic R, Hall LD. Duplex ultrasound criteria for diagnosis of splanchnic artery stenosis or occlusion. J Vast Surg 1991; 14:511-520. Nicholls SC, Kohler TR, Martin RL, Strandness DEJ. Use of hemodynamic parameters in the diagnosis of mesenteric insufficiency. J Vast Surg 1986;3:507-510. Pawelczyk JA, Matzen S, Friedman DB, Secher NH. Cardiovascular and hormonal responses to central hypovolemia in humans. In: Secher NH, Pawelczyk JA, Lundbrook J, eds. Blood loss and shock. London: Edward Arnold, 1994:25-37. Perko G, Perk0 MJ, Jansen E, Secher NH. Thoracic impedance as an index of body fluid balance during cardiac surgery. Acta Anaesthesiol Stand l991;35:568-571. Perk0 MJ, Just S. Duplex ultrasound of superior mesenteric artery: interobserver variability. J Ultrasound Med 1993;5:259-263. Qamar MI, Read AE, Skidmore R, Evans JM, Wells PNT. Pulsatility index of superior mesenteric artery blood velocity waveforms. Ultrasound Med Biol 1986; 12:773-776. Shung KK, Kuo IY, Cloutier G. Ultrasonic scattering properties of blood. In: Roelandt J, Gussenhoven EJ, Born N, eds. Intravascular ultrasound. Dordrecht: Kluwer 1993:119- 139.
8
Ultrasound I996 World
in Med. L Bml.. Vol. 72. No. I, pp. I I-IX, lYY6 Federation for Ultrasound in Medicine & Biology Prmted tn the USA. All rights reserved KiOl-562Y/Y6 % I S.(N) + .oO
0301~5629( 95 ) 02037-3
ELSEVIER
l Original Contribution CHANGES IN SUPERIOR MESENTERIC ARTERY WAVEFORM DURING REDUCTION OF CARDIAC VOLUME AND HYPOTENSION
DOPPLER STROKE
M. J. PERKO, G. PERKO, S. JUST, N. H. SECHER and T. V. SCHROEDER Departments of Vascular Surgery, Radiology, and Anesthesia, The Copenhagen Muscle Research Center. Rigshospitalet, The National University of Copenhagen, Denmark (Received
IO April
1995; in jinal form
24
July
1995)
Abstract-Influence of stroke volume reduction and bypotension on the superior mesenteric artery (SMA) Doppler waveform was evaluated during bead-up tilt-induced central bypovolemia in 11 healthy volunteers. During normotensive reduction in stroke volume, peak systolic velocity ( pV) , mean velocity, pulsatility and resistivity indices decreased, while diastolic velocities increased. During hypotension, a further decrease in pV was accompanied by maintained elevation of diastolic velocities and reduction in pulsatility and resistivity indices. Power of backscattered Doppler wave was elevated throughout the hypovolemia. Alterations in pV and pulsatility indices were closely related to changes in stroke volume, and a negative correlation was found between diastolic velocities and stroke volume. Regression analysis showed no signigcant relation between variations in velocity parameters and blood pressure. Results of the study indicate that alterations in stroke volume induce consequential changes in the SMA Doppler waveform. These changes originate from both direct influence of stroke volume and/or pressure on blood flow velocity, and alterations in SMA peripheral resistance that follow variations in stroke volume. Presented interdependencies should be taken into consideration while studying mesenteric physiology with the use of Doppler technique and while interpreting the duplex results in patients suffering from diseases that may influence flow velocity and mimic or obscure Doppler effects of the SMA stenosis. Key Words: Ultrasonics,
Pulsatility
index, Resistivity
index, Blood flow, Hypovolemia.
reduction of the resistivity indices (Bawersox et al. 1991; Moneta et al. 1991; Nicholls et al. 1986). Alterations in the SMA blood flow velocity (and Doppler waveform) occur, however, not only due to local arterial disease. In clinical practice, we meet patients suspected for mesenteric ischemia who suffer dehydration, hypotension, aortic valve stenosis, and congestive heart disease with low cardiac output that can influence mesenteric artery blood velocity and flow. This study was undertaken to detect whether hypovolemic reductions in stroke volume and arterial blood pressure influence SMA Doppler waveforms.
INTRODUCTION Duplex ultrasound of the superior mesenteric artery (SMA) allows a noninvasive measurement of blood flow velocity and volume through the artery in humans (Jager et al. 1986). This simple-to-apply and relatively inexpensive technique gains acceptance as a method for studying mesenteric physiology and recognition of SMA obstructive disease (Lilly et al. 1989; Moneta et al. 1991). Stenosis of the artery is characterized by elevation of the blood flow velocity over the stenotic area, reduction of the velocity caudally to the stenosis and reduction of the peripheral mesenteric resistivity. In ultrasound Doppler examination, these changes are best reflected by increase in peak systolic ( pV), diastolic and mean velocity (MV) over the stenotic area; reduction in velocities caudally to the stenosis; and Address correspondence to: Dr. Mario J. Perko, University Copenhagen, KAS Gentofte, Department R, Niels Andersens 65, DK-2900 Hellerup, Denmark.
MATERIALS
AND METHODS
Eleven healthy, fasting ( ~-8 h), nonsmoking and nonmedicated volunteers with a mean age of 26 ( 1932) y, weight 65 (48-85) kg, height 175 (164-187) cm and a male-to-female ratio of 8:3, underwent duplex ultrasonographic examination of the SMA during passive head-up tilt. The study was approved by the
of Vej
11
I7
Ultrasound
in Medicine
and Biology
Volume
22. Number
I. 1996
Ethical Committee of Medical Research in Copenhagen, and all subjects gave informed consent in accordance with the Helsinki II Declaration.
with no footboard. Leg movements were not allowed to reduce venous return (Matzen et al. 199 1). Throughout the study, all variables were recorded every minute.
Study design The investigation was comprised of four stages: 1) 3Omin supine rest; 2) passive head-up tilt to 30” for 2 min; 3) increasing the head-up tilt to 50” and sustained tilt until presyncopal symptoms appeared (dizziness, nausea, pallor and sweating); and 4) supine recovery until stabilization of cardiovascular variables. During the head-up tilt, the subjects were supported by a bicycle saddle but
Equipment and variables Impedance cardiography is a noninvasive, reliable method for estimation of alterations in fluid volume, heart rate, stroke volume, cardiac output and cardiac index, which correlates closely with thermodilution technique (Bernstein 1986; Clancy et al. 1991 1. The method is based on a mathematical analysis of both thoracic volume and the rate of change of thoracic
A
8
”
4
8
4
I
Fig. 1. Doppler waveform of the superior mesenteric artery: (a) at rest, and (b) during stroke volume reduction and hypotension (presyncope) . pV, peak systolic velocity; pdV, peak diastolic velocity; minV, minimal velocity; MV, mean height over the cardiac cycle; S, maximum height of the waveform; T, length of the cardiac cycle; ts, duration of the systolic peak (measured between half-amplitude points).
Superior mesenteric artery changes 0 M. J. PERKO et al.
electrical impedance during the cardiac cycle. We used a Simonsen & Wee1 (Copenhagen, Denmark) impedance cardiomonitor which records Zo at 100 kHz and calculates stroke volume from the beat-to-beat pulsatile changes in thoracic impedance (Kubicek et al. 1966; Matzen et al. 1990). Changes in Zo reflect alterations in intrathoracic fluid volume (Perk0 et al. 1991) _ Systolic and diastolic arterial blood pressures were measured by sphygmomanometry on the left arm and mean arterial pressure (pulse pressure 3 -’ + diastolic pressure ) was calculated. Ultrasound duplex evaluation was conducted with an Ultramark 9 (Advanced Technology Laboratories, Bothell, WA, USA) using a 2.5-MHz phased-array probe. In this system, the elements of the array are used to generate a short-gate pulsed-wave Doppler beam that can be steered to the desired location in the sector scanned with constant control of the angle of insonation. Estimation of blood flow by means of the duplex scan carries some potential errors (Evans 1989). To reduce random errors and minimize variability and relative inaccuracies during assessment of the Doppler parameters, all measurements were standardized as follows: the SMA was examined 2-3 cm distally from its origin, at a 60” angle of insonation (the angle between ultrasound beam and the long axis of the vessel), and at a depth of 4-8 cm; samples of ultrasound beam were placed entirely within the artery; high-pass filters were used to remove low-frequency (O-100 Hz) signals arising from the vessel wall; repetitive measurements were collected from four to six cardiac cycles (Perk0 and Just 1993 ) . The Doppler parameters recorded included blood flow velocities: peak systolic (pV>, minimal (minV ) , peak diastolic ( pdV) , end-diastolic (edV) and mean (MV) ; and their derivatives: pulsatility index [PI = (pV - minV) MV -’ 1, resistivity index [RI = ( pV - minV) pV -’ 1, constant flow ratio [PI (RI - PI))‘] and height width index [PI (Tt,;‘)], where T is the time of a single wave and t, is the duration of the systolic peak (Evans et al. 1989) (Fig. 1). The power (decibels) of backscattered Doppler wave was measured off-line from digitized Doppler spectral data. Diameter of the artery was assessed at long arterial axis display and, for standardization of the procedure, during systole. MV was determined by off-line planimen-y. The SMA blood flow volume was determined as MV &r2, and systemic vascular resistance was the ratio between mean arterial pressure and cardiac output. Duplex scan examinations were videotape recorded for computerized graphical off-line processing in a software program written in C. Calculations
Comparative statistics within the individual parameters were made for time-specific points (from rest to 8
13
min after appearance of presyncopal symptoms) of all examinations as shown in Fig. 2. Data were first analyzed using one-way analysis of variance for repeated measurements, and if the variance of a parameter was significant, differences between time-specific points were evaluated by a post hoc multiple comparison test ( Newman-Keuls test). Uni- and multiple regression analyses were used to detect interdependencies between the parameters. The p of <0.05 was considered significant. Results are given as meant standard error or with range. RESULTS
The course of selected variables and Doppler indices during head-up tilt is presented in Fig. 2. Alterations in cardiac output and cardiac index corresponded closely to changes in stroke volume. The onset of presyncopal symptoms occurred within 16 (5-47) min of head-up tilt. Return to the supine position initiated immediate recovery and normalization of circulatory and Doppler variables within a few minutes. Systemic hemodynamics
Continuous significant increase in thoracic impedance throughout the head-up tilt from 35.0 -+ 0.7 s1 at rest to 38.0 t: 0.7 R at presyncope (p < 0.01) indicated progression of central hypovolemia. Until 1 min before occurrence of presyncopal symptoms, in spite of tachycardia, a successive decrease in stroke volume produced reduction in cardiac output from 5.7 2 0.4 to 3.9 + 0.3 L min-’ (p < 0.01). A simultaneous increase in systemic vascular resistance from 16 t 2 to 22 + 4 mmHg min L-’ (p < 0.01) maintained arterial pressure. Occurrence of presyncopal symptoms was associated with a decrease in arterial pressure and further reduction in stroke volume. Systemic vascular resistance and heart rate decreased, re-establishing pretilt level (relative bradycardia) . SMA ultrasound parameters
Peak systolic velocity decreased throughout the head-up tilt from 211 + 14 cm s ’ at rest to the lowest value of 87 + 5 cm s-’ at presyncope (p < 0.01). The diastolic velocities increased gradually from rest to presyncope as follows: minV from -9 + 10 to 28 + 2cms-‘; pdV from 19 + 5 to 34 -+ 2 cm ss’; and edV from 19.2 to 32.6 2 2.3 cm s-’ (p < 0.01). Mean velocity decreased from 54 ? 6 cm s-’ at rest to 33 2 1 cm s-’ at presyncope (p < 0.01) . This reduction in MV was entirely due to decrease in mean velocity of the systolic phase of cardiac cycle (Fig. 2). Opposite changes in systolic (decrease) and diastolic (increase) velocities resulted in conversion of the triphasic Doppler waveform to bi- or monophasic during presyncope
Ultrasound
in Medicine
and Biology
Volume
22. Number
1. 1996
0.8 a
0.6
0I -1.2 &
-,::
““yz,
U
-1.6-
a
-1% -2-
OJ Dia.&c
Fig. 2. Cardiovascular and Doppler variables during head-up tilt at rest, tilting to 30” and 50”, maintained 50” head-up tilt for 8 min before presyncopal symptoms, during the onset of presyncopal symptoms, and in the recovery period. (0) Significant difference with rest; (0) significant difference from preceding value. MAP, mean arterial pressure; pV, peak velocity; minV, minimal velocity; MV, mean velocity; PI, pulsatility index; RI, resistivity index; HWI, height width index; CPR, constant flow ratio.
(Fig. 1) . Return to the supine position induced a brief (several seconds), rapid increase in pV to nearly double of the resting value, followed by fast stabilization at the pretilt level. PI and RI decreased gradually throughout head-up tilt, reaching the lowest values at presyncope. The diameter of the SMA increased shortly after initiation of 50” head-up tilt from 5.8 ? 0.2 to 6.9 + 0.4 cm (p < 0.05)) and returned to resting
value within 2 min. Blood flow decreased temporarily during presyncope from 0.88 t 0.13 L min-’ at rest to 0.41 + 0.03 L mitt-’ (p < 0.05). Power of backscattered Doppler signals increased throughout the headup tilt and was maintained elevated during the first 4 (2- 10) min of recovery. Mean relative changes (A between the rest and sympathoexcitatory phase of hypovolemia) in power of specific frequencies are given
Superior mesenteric artery changes 0 M. J. PERKO
in Table 1. An example of a Doppler spectrogram obtained from a single subject at rest and during hypovolemic reduction in stroke volume is presented in Figs. 3 and 4. Relations between variables
Mean arterial pressure, pulse pressure and stroke volume were selected as independent variables for regression analyses. Univariate analysis indicated no significant relation between alterations in velocity parameters and arterial pressure, and no relation between changes in blood velocity and pulse pressure (Table 2 ) . On the other hand, a strong positive correlation was found between systolic velocity and stroke volume, and a negative correlation between diastolic velocities and stroke volume. In the forward stepwise multiple regression, arterial pressure was rejected from most of the models, while stroke volume had a positive association with all Doppler indices (Table 3 ) . DISCUSSION The results of the present study indicate that: 1) SMA systolic blood flow velocity can be reduced by about 50% during reduction in inflow volume (stroke volume) despite maintained blood pressure; 2) SMA systolic velocity can be reduced by about 60% during severe reduction in stroke volume inducing hypotension; 3) SMA diastolic velocity increases during inflow volume reduction-most probably due to reduction in peripheral mesenteric resistance; 4) SMA backscattered power of the Doppler signal increases during reduction in stroke volume and hypotension; 5) SMA Doppler-derived frequency indices can be influenced by changes in stroke volume. Systemic hemodynamics
During the first stage of head-up tilt, the so-called sympathoexcitatory phase of central hypovolemia, the decrease in stroke volume opposed the increase in systemic vascular resistance, which maintained arterial pressure. During the second sympathoinhibitory phase of central hypovolemia, the decrease in stroke volume was accompanied by a reduction in systemic vascular resistance
Table 1. Relative change (A from rest to hypovolernia) in the power of the highest frequencies present. Phase
A Frequency power (dB)
Peak systolic Minimal diastolic Peak diastolic End diastolic
1.8 -+ 0.7 1.4 2 0.2 4.6 t 0.6
3.1 2 0.8
P <0.005 <0.0005
er al.
15
resulting in hypotension. These observations correspond well with earlier findings (Pawelczyk et al. 1994 ). Doppler waveform parameters
Decreases in pV and MV during the normotensive phase of hypovolemia indicated an influence of the inflow (to the SMA) volume on these two parameters. This observation is confirmed by regression analysis, which revealed dependency of pV and MV on changes in stroke volume. Simultaneously, all of the diastolic velocities (minV, pdV, edV) increased, indicating reduction of the peripheral mesenteric resistance. This reduction in wave amplitude resulted in decreases in PI and RI, while blood flow did not change. During the hypotensive phase of hypovolemia, decreases in pV and MV were accompanied by maintained elevation in diastolic velocities resulting in continuous depletion of wave amplitude-and a further reduction in PI and RI. Thus, the present study indicates that pulsatility indices are influenced by both inflow volume (stroke volume) and regional peripheral resistance. This deduction is valid on the condition that increase in diastolic velocity truly reflects decreasing peripheral resistance. An alternative explanation for increase in diastolic velocity could have been local vasoconstriction, but this is unlikely as vasoconstriction would increase pV and MV, and reduce diameter of the vessel, neither of which took place. Power of the ultrasound signal backscattered by blood depends on transmitted frequency power, presence of red blood cell aggregation, compressibility and density of red cells and plasma, hematocrit, packing factor and volume of the scatters (Shung et al. 1993). The present study design allows the assumption that all but last variable were constant during examinations, thus the power of the reflected Doppler wave within a particular frequency depended predominantly on the number of blood cells moving with velocity producing this frequency. Thus, accentuation of the power of the diastolic frequencies, as seen in Figs. 3 and 4, reflected increased volumetric blood flow during the diastolic phase of the cardiac cycle. One can argue that increased power, and therefore increased flow volume, should be accompanied by local increase in diameter of the artery, which was not observed in the present study. The possible explanation may be that volumetric blood flow is proportional to the squared diameter, thus minor changes in diameter of an artery may produce remarkable changes in blood flow volume. Head-up tilt-induced changes in height width index resembled that of PI, however, the variance did not reach statistical significance. This “smoothing” of height width index originated from a shortening of the systolic phase and a relative elongation of the diastolic phase of the
Ultrasound
in Medicine
and Biology
Volume
22, Number
I. 1996
Frequency(Hz) Fig. 3. An example of spectral content and power of frequencies over a single cardiac cycle: (A) at rest and during (B) normotensive reduction in stroke volume.
Doppler wave. Thus, height width index takes into account prolonged diastolic flow and in that way is more “sensitive” to increased diastolic velocity. Inclusion of the systolic-to-diastolic velocity ratio in an index equation makes the index independent of the angle of insonation. On the other hand, a disadvantage is that the index becomes dependent on both inflow volume and peripheral resistance. In situations in which both of these factors may alter or deviate from
“normal” due to diseases (e.g., heart insufficiency + peripheral arterial stenosis), evaluation of peripheral resistance by means of Doppler ultrasound may be erroneous. In clinical practice, erroneous interpretation of mesenteric Doppler examination may be verified by X-ray arteriography. In physiological studies, however, verification of Doppler ultrasound observations is, for ethical reasons, very difficult. Lilly et al. (1989) re-
Superior 8 6 4
A
mesenteric
artery
I
I
I
6A
early diastolic (239 ms)
4 2
I
I
I
I
I
8
6A
peak diastolic
4
:& R
6A
end diastolic (864 ms)
4
6 4 2 0 8 6 4 2
B
MAP
,
8
peak systolic
17
er al.
Table 2. Univariate regression analysis.
peak systolic (67 ms)
2 o-
o- !
changes 0 M. J. PRRKO
Pulse pressure
Parameter
r
P
Peak systolic V Minimal V Peak diastolic V End diastolic V Mean V-systolic Mean V-diastolic Mean V Pulsatility index Resistivity index HWI
0.08 -0.01 -0.04 -0.04 -0.02
NS NS NS NS NS
CFR Diameter Blood flow
-0.21 < 0.05 0.06 NS 0.10 0.14 0.08
NS NS NS
0.33 0.33 0.17
e < NS
r
P
0.23 < 0.03 -0.25 < 0.01 -0.14 NS -0.14 NS 0.11
NS
-0.17 NS 0.04 NS 0.21 < 0.05 0.27 e 0.20 NS 0.33 NS 0.31 e 0.24 < 0.02
Stroke volume r
P
0.68 -0.48 -0.41 -0.41
-3 4 e e
0.44
e
-0.33 6 0.27 < 0.51 < 0.63 < 0.34 Q 0.46 -S -0.15 NS 0.11 NS
MAP, mean arterial pressure; V, velocity; HWI, height width index; CFR. constant flow ratio; G. p < 0.001; NS, not significant; r, correlation coefficent.
(83 m.9 I
I
I
1
B
O_rflI
I
I
I
early diastolic (232 ms) I
I
I
I
,
/
peak diastolic
0
516 1033 1550 2067 2583 3100 Frequency(Hz)
Fig. 4. Relative change in the power of frequencies from Fig. 3: (A) at rest and during (B) normotensive reduction in stroke volume.
ported about 40% increase in pV after glucagon infusion and suggested that pharmacological stimulation by glucagon may substitute for the physiologic stimulation of a meal. However, the diastolic velocity did not increase significantly after glucagon infusion in contrast to the meal response (Lilly et al. 1989). In the light of the present study, pV is predominantly related to the inflow alterations (stroke volume), while diastolic velocities are rather inflow-independent. Thus, the response pattern to glucagon infusion may represent changes in cardiac output rather than peripheral mesenteric vasodilation. PI is the ratio between velocity amplitude and MV. In the present study, MV decreased during hypo-
volemia and hypotension, thus the decrease in PI was entirely due to amplitude reduction. Qamar et al. ( 1986) reported a 46% decrease in PI following ingestion of a meal and related this observation to vasodilation in the splanchnic bed and decreased downstream resistance. The authors did not indicate whether the fall in PI was a result of depletion of wave amplitude or an increase in MV, but from another investigation (Jlger et al. 1986) it is known that a postprandial moderate increase in amplitude (from 120.0 to 125.6 cm s-’ ) is accompanied by a more than twofold increase in MV (from 22.2 to 57.0 cm s-l). Thus, in this study, the postprandial decrease in PI was an effect of increased MV. The increase in MV could be due to reduction in peripheral resistance, but it could also be due to postprandially increased cardiac output, and most probably due to both. Another possibility is that reduction in peripheral resistance may be due to arteriolar dilation, but it could also be due to decrease in blood pressure. CONCLUSIONS Significant changes in the SMA Doppler waveform may occur during alterations in stroke volume and blood pressure. These Doppler waveform changes may originate from direct influence of inflow (to the SMA) volume and/or pressure on blood flow velocity, but may also occur due to alterations in peripheral resistance that follow variations in cardiac output. These interdependencies should be taken into consideration while studying mesenteric physiology with the use of the Doppler technique and while interpreting the SMA duplex results in patients suffering from hy-
Ultrasound in Medicine and Biology Table 3. Multiple
Volume 22, Number I. 1996
regression analysis (forward
MAP Dependent parameter
P
Peak systolic V Minimal V Peak diastolic V End diastolic V Mean V-systolic Mean V-diastolic Mean V Pulsatility index Resistivity index HWI CFR Diameter Blood flow
rejected 0.17 (0.11) rejected rejected rejected -0.16 (0.10) rejected rejected rejected rejected 0.27 (0.09) 0.22 (0.11) rejected
stepwise method).
Pulse pressure P
NS
NS
< : 0.01 NS
B
rejected -0.20 (0.11) rejected rejected rejected rejected rejected rejected rejected rejected rejected 0.28 (0.12) 0.24 (0.10)
Stroke
I’
P
NS
-0.45 (0. IO) -0.41 (0.10) -0.41 (0.10) 0.44 (0.09) -0.30 (0. IO) 0.27 (0.10) 0.51 (0.09) 0.63 (0.08) 0.34 (0. IO) 0.42 (0.09) -0.27 (0. IO) rejected
volume
I’
0.68 (0.08)
< 0.03 < 0.03
MAP, mean arterial pressure; V, velocity; HWI, height width index; CFR, constant flow ratio; <. p < 0.001; NS, not significant; p, standardized regression coefficient.
po- and hypertension, hypo- and hyperthyroidism, cardiac insufficiency and other diseases that may influence cardiac output and/or pressure. work was supported by Gerda and Aage Haensch’s Fond; Fonden til Sygdmomsbekrempelse uden Dyreforseg; and Ferd. & Ellen Hindsgauls Fond.
Acknowledgements-This
REFERENCES Bernstein DP. Continuous noninvasive real-time monitoring of stroke volume and cardiac output by thoracic electrical bioimpedance. Crit Care Med 1986; 14:898-901. Bowersox JC, Zwolak RM, Walsh DB, Schneider JR, Musson A. Duplex ultrasonography in the diagnosis of celiac and mesenteric artery occlusive disease. J Vast Surg 1991; 14:780-788. Clancy TV, Norman K, Reynolds R, Covington D, Maxwell JG. Cardiac output measurements in critical care patients: thoracic electrical bioimpedance versus thermodilution. J Trauma 1991;31:1116-1119. Evans DH, McDicken WN, Skidmore R, Wodcock JP. Doppler ultrasound. Physics, instrumentation, and clinical applications. New York: Wiley, 1989. Jager K, Bollinger A, Valli C, Amman n R. Measurement of mesenteric blood flow by duplex scanning. J Vast Surg 1986;3:462469. Kubicek WG. Karnegis JN, Patterson RP, Witsoe DA, Matterson
RH. Development and evaluation of an impedance cardiac output system. Aerospace Med 1966;37:1208- 1212. Lilly MP, Harvard TRS, Flinn WR, Blackburn DR, Astleford PM. Duplex ultrasound measurement of changes in mesenteric flow velocity with pharmacologic and physiologic alteration of intestinal blood flow in man. J Vast Surg 1989;9:18-25. Matzen S, Perk0 G, Groth S, Friedman DB, Secher NH. Blood volume distribution during head-up tilt induced central hypovolaemia in men. Clin Physiol 1991; 11:41 I-422. Moneta GL, Yeager RA, Dalman R, Antonovic R, Hall LD. Duplex ultrasound criteria for diagnosis of splanchnic artery stenosis or occlusion. J Vast Surg 1991; 14:511-520. Nicholls SC, Kohler TR, Martin RL, Strandness DEJ. Use of hemodynamic parameters in the diagnosis of mesenteric insufficiency. J Vast Surg 1986;3:507-510. Pawelczyk JA, Matzen S, Friedman DB, Secher NH. Cardiovascular and hormonal responses to central hypovolemia in humans. In: Secher NH, Pawelczyk JA, Lundbrook J, eds. Blood loss and shock. London: Edward Arnold, 1994:25-37. Perko G, Perk0 MJ, Jansen E, Secher NH. Thoracic impedance as an index of body fluid balance during cardiac surgery. Acta Anaesthesiol Stand l991;35:568-571. Perk0 MJ, Just S. Duplex ultrasound of superior mesenteric artery: interobserver variability. J Ultrasound Med 1993;5:259-263. Qamar MI, Read AE, Skidmore R, Evans JM, Wells PNT. Pulsatility index of superior mesenteric artery blood velocity waveforms. Ultrasound Med Biol 1986; 12:773-776. Shung KK, Kuo IY, Cloutier G. Ultrasonic scattering properties of blood. In: Roelandt J, Gussenhoven EJ, Born N, eds. Intravascular ultrasound. Dordrecht: Kluwer 1993:119- 139.