- Email: [email protected]
Inaccuracies and variability of indirect pressure measurements during cardioplegia administration
Inaccuracies and Variability of Indirect Pressure Measurements During Cardioplegia Administration Norman S. Kato, MD, Gerald D. Buckberg, MD, Christine K. Cushen, CCP, and Colin R. Whitwam, CCP Division of Cardiothoracic Surgery, Department of Surgery, University of California School of Medicine, Los Angeles, California
This study shows cardioplegic delivery requires direct measurement of intravascular pressure, rather than its estimation by calibration of individual delivery systems or by aortic palpation. The effects of temperature, hematocrit, flow rate, and cannula type were tested in vitro after recording intravascular pressure during routine cardiac operations. Inaccuracies were introduced by estimating intravascular pressure, as changes in blood viscosity are affected by hematocrit and temperature, and delivery system pressure varied in relation with the type of cannula, direction of perfusion, and flow rate. Additionally, clinical delivery introduces the variable of intravascular resistance. The variability of direct intravascular
pressure versus predicted pressure increased as flow rate was raised. These inaccuracies were overcome completely by directly monitoring intravascular pressure from the side ports of antegrade and retrograde cannulas. We conclude that (1) monitoring cardioplegic delivery device pressure is useful primarily to detect potential obstruction in the delivery system that must be corrected intraoperatively and (2) predicted pressure (by either palpation or in vitro calibration) is an unreliable method of determining intravascular pressure during cardioplegic delivery.
C
drop across a cardioplegic system or by palpation of the aorta and the actual intraoperative intravascular pressure. Data will be presented emphasizing the importance of direct intravascular pressure measurement by confirming the confounding effects of varying temperature, flow rate, hematocrit, and direction of perfusion on pressure measured in the cardioplegic system, thereby documenting the inaccuracy of estimating actual intravascular pressure by indirect methods.
urrently, more than 95% of cardiac surgeons employ cardioplegia techniques, and the use of warm and cold antegrade and retrograde cardioplegic delivery has enhanced the safety of operation in the increasing number of high-risk patients requiring surgical intervention [1]. The effectiveness of cardioplegic strategies depends on adequate cardioplegic solution distribution to all areas of the myocardium. This objective requires sufficient flow rates and perfusion pressure for delivery into arterial and venous beds with different vascular resistances. Presently, perfusionists at most cardiac centers estimate intravascular pressure by measuring pressure drops across the delivery system. This pressure drop calibrated preoperatively is subtracted from the observed delivery system pressure while infusing cardioplegic solution to the patient. The imposed changes in delivery temperature, vascular bed resistance, or direction of perfusion via different types and sizes of antegrade and retrograde cannulas make indirect methods of determining intravascular pressure by preoperative or postoperative calibration techniques both difficult and cumbersome. Alternatively, the surgeon estimates antegrade pressure by aortic palpation, but this assessment cannot be done during retrograde cardioplegic delivery, which is used currently in more than 60% of operations in the United States as described in a recent survey of more than 1,400 cardiac surgeons [2]. This study tests the hypothesis that there is no correlation between the estimated intravascular pressure determined either by subtracting a predetermined pressure Accepted for publication July 8, 1994. Address reprint requests to Dr Buckberg, Department of Surgery, UCLA Medical Center, Room B2-375 CHS, Los Angeles, CA 90024-1741.
© 1994 by The Society of Thoracic Surgeons
(Ann Thorae Surg 1994;58:1188-91)
Material and Methods Blood cardioplegia was administered to 10 consecutive patients undergoing elective coronary bypass grafting using a 4:1 blood:cardioplegia ratio, with hematocrit ranging between 20% and 22%. A standard cardioplegic delivery system (Shiley BCD Plus; Sorin Biomedical Inc, Irvine, CA) and a standard roller pump (Cobe-Stockert; Cobe Cardiovascular Inc, Harvada, CO) were employed to deliver the solution into the aortic root through an antegrade cardioplegia cannula with a pressure monitoring side port (ATCOllMV; Research Medical Inc, Salt Lake City, UT) or into the coronary sinus through a retrograde cardioplegia cannula with a pressure measuring side port (RC014T; Research Medical Inc). Intraoperative pressure measurements were recorded simultaneously from the pressure ports on the cannulas (electronic transducer; ViggoSpectromed, Oxnard, CA) and pressure ports in the cardioplegic system itself using a DLP 6000 monitor (DLP, Grand Rapids, MI) positioned at the heart-lung machine. Antegrade cardioplegic pressure was estimated by digital palpation of the ascending aorta by three different surgeons (each with experience in more than 1,000 cardiac 0003-4975/94/$7.00
HOW TO DO IT KATO ET AL INACCURATE INDIRECT PRESSURE MEASUREMENTS
Ann Thorae Surg 1994;58:1188-91
operations) while the perfusionist recorded the aortic pressure measured from the pressure monitoring port of the antegrade cardioplegic cannula. This information was not transmitted to the surgeon intraoperatively, to avoid having this knowledge influence the predicted pressure during the next antegrade infusion. At the conclusion of the surgical procedure, an in vitro calibration procedure was carried out by measuring the pressure within the delivery system while cardioplegia was infused through the antegrade and retrograde cannulas into a beaker on the operating table positioned at the heart level. The system was zeroed by taking into account the difference in height between the delivery system pressure port and the heart level. The predicted intravascular pressure was determined as the difference between the cardioplegic delivery system pressure measured during the surgical procedure and the pressure drop measured in the delivery system during the calibration procedure. These predictions were compared with the actual intravascular pressures recorded during intraoperative cardioplegic administration. After cardiopulmonary bypass, pressure measurements in the cardioplegic delivery system were made at a hematocrit of approximately 20%, the average intraoperative hematocrit of the cardioplegic solution in a patient with 25% hematocrit blood mixed 4:1 with cardioplegic solution. Asanguineous cardioplegia with zero hematocrit was simulated by administering normal saline solution to correspond to delivery of a crystalloid cardioplegic solution. Measurements were obtained at flow rates varying from 50 to 350 mL/min and 50 to 200 mL/min for antegrade and retrograde flow, respectively, and at temperatures 8°C and 37°C to correspond to conditions existing during operation. Data were analyzed with StatView V2.0 on an Apple Macintosh lIci. Analysis of variance was employed for comparisons between groups. The relationship between observed and predicted pressure was tested by linear regression analysis. Differences were considered significant at a probability level less than 0.05. Group data are expressed as mean ± standard error of the mean.
Results In Vitro Measurements Cardioplegic delivery system pressure never exceeded 120 mm Hg during in vitro calibration procedure at flow rates up to 350 mL/min. However, under all conditions, the cardioplegic delivery system pressure was higher with retrograde than antegrade delivery, as we used a retrograde cannula that was designed to impose a distal resistance to accomplish the self-inflating capacity of the balloon. At normothermia (37°C) varying hematocrit between 0% (crystalloid cardioplegia) and 20% (blood cardioplegia) produced negligible difference in cardioplegic delivery system pressure (Fig 1). Conversely, cardioplegic system pressure was significantly higher at all flow rates with cold blood cardioplegia (8°C) than with cold crystalloid cardioplegia, with the difference in pressure averaging 40 mm Hg at the maximum antegrade and retrograde flow rates (350
1189
140 120
Ci
100
:I:
E
..
.§.
80
Hct=20%
Gl
::l III III
e c.
60
~
~Hct=O%
40
20 0 0
50
100
150
200
250
300
350
400
Flow (ml/minute)
3rc
Fig 1. Delivery system pressure during in vitro infusion of blood and crystalloid cardioplegia at flow rates ranging from 50 to 350 ml.lmin. Note comparable delivery system pressure with normothermic blood and crystalloid perfusion. (n = 5 for each obseroaiion.)
mL/min and 250 mL/min, respectively), as shown in Figure 2. Pressure varied negligibly during the calibration procedure, attesting to the reproducibility of in vitro delivery system pressure measurement.
In Vivo Measurements Figure 3 shows the marked difference between predicted intravascular pressure and using the in vitro calibration subtracted from the cardioplegic delivery system pressure recorded during infusions and actual intravascular pressure during antegrade and retrograde delivery of blood cardioplegia at 8°C intraoperatively. Direct intravascular pressure differed from predicted pressure during antegrade delivery as much as -20 mm Hg and +70 mm Hg with antegrade delivery at flows ranging from 50 to 350 mL/min. These differences ranged from -20 mm Hg to + 15 mm Hg during retrograde delivery at flows ranging from 50 to 250 mL/min. Consequently, there was marked variability and limited correlation (r 2 = 0.56 and 0.76, respectively) between predicted and measured pressure with both routes of delivery. Figure 4 shows the difference between predicted aortic pressure, estimated by digital palpation of the aorta, and direct intraaortic pressure measurement. The correlation between predicted and observed pressure was weak and even more variable than when intravascular pressure was estimated by subtracting the calculated pressure drop in vitro from the observed delivery system pressure during intraoperative antegrade cardioplegic delivery.
Comment These data show that direct intravascular pressure measurement is the only reliable method for determining either aortic or coronary sinus pressure during cardiople-
1190
HOW TO DO IT KATO ET AL INACCURATE INDIRECT PRESSURE MEASUREMENTS
gic delivery, and document the inaccuracy of indirect estimation of p.ressure by either calibration of the delivery system or d~!Sttf!:f palpation of the aorta. The pressHn~-f1ow relationships for any cardioplegic delivery system are unique to that system and vary minimally during in vitro calibration. Delivery system pressure, however, is dependent on flow rate, temperature, viscosity, and compliance of the tubing system, which changes with its thickness, length, and temperature. Consequently, delivery system pressure with cold blood cardioplegia is higher because hypothermia increases blood viscosity and reduces tubing compliance and elastic recoil. The influence of viscosity lessens with normothermia, and the outflow resistance of the employed cannula becomes important (ie, antegrade versus retrograde cannulas). We
Ann Thorac Surg 1994;58:1188-91
250
_
200
Cl :I:
E
.§. l!!
150
Observed
:::l
VI VI
e :e0 a. o
«
100
50
0 0 140
50
100
150
250
200
Predicted Aortic Pressure (mmHg)
A 120 150
Hct=20%
Ci
100
Ci
80
.§.
J:
J:
E
.§.
e VI VI e a. :::l
I Predicted
E 125
e VI VI e a.
:::l 100
60 40
VI
75
:::l
C
C;;
20
~
I
50
~
0
50
100
150
A
200
250
300
350
400
0
0
Flow (ml/minute)
Ci E
§. l!! :::I
Hct = 20%
100
VI VI
/
80 60
l!!
D..
40
..
20 0 0
B
50
./
I
/ 100
~
/
150
/
;N
/
200
Hct = 0%
/
250
r2 = 0.76
1\
25
0
:I:
Observed
0
140 120
I
I
I
I
I /-
CO C
0
I
I
I
I
I
300
350
Flow (ml/minute)
Fig 2. Delivery system pressure during in vitro infusion of 8°C blood and crystalloid cardioplegia through antegrade (A) and retrograde (B) cannulas. Note higher delivery system pressure with hypothermic blood cardioplegia (p < 0.05 at all flow rates >50 ml.lmin) and negligible variability at each rate of perfusion. (n = 5 for each observation.)
B
25
50
75
100
125
150
Predicted Coronary Sinus Pressure (mmHg)
Fig 3. Comparison between measured intravascular aortic and coronary sinus pressure and predicted pressures during clinical antegrade (A) and retrograde (B) warm and cold blood cardioplegia perfusions. Hatched line of identity would occur if predicted and direct pressure were identical. These data for antegrade delivery are expressed by the equation y = 28.19 + 0.37x with an r value of 0.56, and for retrograde delivery by the equation y = 11.93 + 0.27x with an r value of 0.76. Both slopes and the intercepts are significantly (p < 0.05) different from the slope of identity.
tested only one type of antegrade and retrograde cannula in our system and would suspect similar directional changes with other cannulas and systems available commercially, especially because cannula resistance is low in manually inflating retrograde cannulas. This difference, however, would not be sufficient reason to avoid direct intravascular pressure measurements as will be discussed subsequently. In contrast to in vitro findings, significant variability of pressure-flow relationships occurred when cardioplegia
HOW TO DO IT KATO ET AL INACCURATE INDIRECT PRESSURE MEASUREMENTS
Ann Thorae Surg 1994;58:1188-91
100
Ci
:I:
• • • •
80
E
.s l!!~
UI UI
e a. :e0 (,)
«
.
•
90
• • • • • ,(Predicted • • .: Observed •• • .~ •r =0.32 2
.. ..
70 60
/
'/ • • •
./ •
50 40
/
/
30
• • •
/'
•
• •
20 +----,----,------,--,---,---,---,--_ 20
30
40
50
60
70
80
90
100
Palpated Aortic Pressure (mmHg)
Fig 4. Comparison between pressure estimated by aortic palpation and direct intraaortic pressure measured during antegrade infusions. Hatched line of identity would occur if predicted (or estimated) pressure and direct pressure were identical. These data are expressed by the equation y = 16.66 + 0.77x with an r 2 value of 0.32. Both the slope and intercept are significantly different (p < 0.05) from the slope of identity.
was delivered into the aorta or coronary sinus intraoperatively, because intravascular resistance changes independently from the pressure drop imparted by the delivery system containing the antegrade and retrograde cannulas. This is especially true with antegrade blood cardioplegia, because patients with coronary disease have variable degrees of fixed vascular obstruction and aortic compliance. Furthermore, arteriolar vascular resistance can change from dose to dose in the same heart, which also may generate different vascular pressures at similar flow rates, especially when high initial flow rates are used to induce antegrade cardioplegic arrest. Conversely, the relative lower variance of predicted versus measured pressure for retrograde cardioplegia is related probably to the use of lower maximum flow rates (200 to 250 mL/min) into a venous bed that is devoid of obstruction, unless the heart is retracted and there is physical distortion of the coronary sinus outflow. Variability of intravascular pressure increased as flow rate was raised with both antegrade and retrograde blood cardioplegic delivery, but the lack of correlation between predicted and measured intravascular pressure existed at all delivery rates (see Fig 3). Furthermore, there was no correlation between palpated aortic pressure and measured aortic pressure (see Fig 4). These observations reinforce the need to measure pressure directly from the aorta during antegrade delivery. A low aortic pressure (ie, <30 mm Hg) may signify unsuspected aortic insufficiency or aortic valve incompetence produced
1191
by a malpositioned aortic clamp, distortion of the noncoronary cusp by the venous cannula, or inadvertent failure to close the vent clamp or to switch from the retrograde to the antegrade route of delivery. It would seem impossible to estimate pressure by palpation of the coronary venous system, so that direct pressure measurement is essential. Direct coronary sinus pressure measurement is important so that the perfusionist can tell the surgeon if the pressure is too high (ie, >50 mm Hg) or too low «15 mm Hg). If the coronary sinus pressure exceeds 50 mm Hg, the perfusionist can reduce the flow to prevent disruption of the coronary sinus or production of myocardial edema while simultaneously alerting the surgeon that the catheter may be wedged and require either retraction or repositioning. Communication that the coronary sinus pressure is less than 15 mm Hg will allow the surgeon to determine if the cannula has retracted into the right atrium, the balloon is not inflated, or there is a persistent left superior vena cava. Our findings should not be interpreted to discourage measuring pressure in the cardioplegic delivery system. This measurement should be monitored routinely, because obstruction in the delivery system (ie, kinking or failure to open a flow clamp) will raise delivery system pressure and produce subsequent disruption of the system if not remedied immediately by stopping the flow and determining the source of obstruction. We recommend, therefore, that both cardioplegic delivery system pressure be monitored to detect such potential disruption and intravascular pressure be measured to optimize delivery of either antegrade or retrograde cardioplegia. Additionally, we continue to palpate the aorta during antegrade delivery and observe the distention of the coronary veins during retrograde delivery, as these clinical tools remain useful deterrents to total reliance upon strain gauges that occasionally can be calibrated incorrectly. The availability of pressure measuring capability in current antegrade and retrograde cardioplegic cannulas allows for direct measurements intraoperatively, and this study shows that neither in vitro calibration nor direct palpation will provide this information. These observations have led to our measurement of aortic and coronary sinus pressure whenever cardioplegia is delivered either into the aorta or into the coronary sinus. Direct intravascular pressure measurements circumvent the errors imparted by indirect assessments and provide for reliable pressure measurement, and hopefully will improve the effectiveness and safety of antegrade and retrograde cardioplegic delivery.
References 1. Buckberg GO. Myocardial protection: Thorac Cardiovasc Surg 1993;5:98-106. 2. Robinson LA. Cardioplegia solutions in spectives and national trends. Presented vation: Current Technology and Future Oct 16, 1992.
an overview. Semin the 90's: current perat Myocardial PreserTrends, Atlanta, GA,
This study shows cardioplegic delivery requires direct measurement of intravascular pressure, rather than its estimation by calibration of individual delivery systems or by aortic palpation. The effects of temperature, hematocrit, flow rate, and cannula type were tested in vitro after recording intravascular pressure during routine cardiac operations. Inaccuracies were introduced by estimating intravascular pressure, as changes in blood viscosity are affected by hematocrit and temperature, and delivery system pressure varied in relation with the type of cannula, direction of perfusion, and flow rate. Additionally, clinical delivery introduces the variable of intravascular resistance. The variability of direct intravascular
pressure versus predicted pressure increased as flow rate was raised. These inaccuracies were overcome completely by directly monitoring intravascular pressure from the side ports of antegrade and retrograde cannulas. We conclude that (1) monitoring cardioplegic delivery device pressure is useful primarily to detect potential obstruction in the delivery system that must be corrected intraoperatively and (2) predicted pressure (by either palpation or in vitro calibration) is an unreliable method of determining intravascular pressure during cardioplegic delivery.
C
drop across a cardioplegic system or by palpation of the aorta and the actual intraoperative intravascular pressure. Data will be presented emphasizing the importance of direct intravascular pressure measurement by confirming the confounding effects of varying temperature, flow rate, hematocrit, and direction of perfusion on pressure measured in the cardioplegic system, thereby documenting the inaccuracy of estimating actual intravascular pressure by indirect methods.
urrently, more than 95% of cardiac surgeons employ cardioplegia techniques, and the use of warm and cold antegrade and retrograde cardioplegic delivery has enhanced the safety of operation in the increasing number of high-risk patients requiring surgical intervention [1]. The effectiveness of cardioplegic strategies depends on adequate cardioplegic solution distribution to all areas of the myocardium. This objective requires sufficient flow rates and perfusion pressure for delivery into arterial and venous beds with different vascular resistances. Presently, perfusionists at most cardiac centers estimate intravascular pressure by measuring pressure drops across the delivery system. This pressure drop calibrated preoperatively is subtracted from the observed delivery system pressure while infusing cardioplegic solution to the patient. The imposed changes in delivery temperature, vascular bed resistance, or direction of perfusion via different types and sizes of antegrade and retrograde cannulas make indirect methods of determining intravascular pressure by preoperative or postoperative calibration techniques both difficult and cumbersome. Alternatively, the surgeon estimates antegrade pressure by aortic palpation, but this assessment cannot be done during retrograde cardioplegic delivery, which is used currently in more than 60% of operations in the United States as described in a recent survey of more than 1,400 cardiac surgeons [2]. This study tests the hypothesis that there is no correlation between the estimated intravascular pressure determined either by subtracting a predetermined pressure Accepted for publication July 8, 1994. Address reprint requests to Dr Buckberg, Department of Surgery, UCLA Medical Center, Room B2-375 CHS, Los Angeles, CA 90024-1741.
© 1994 by The Society of Thoracic Surgeons
(Ann Thorae Surg 1994;58:1188-91)
Material and Methods Blood cardioplegia was administered to 10 consecutive patients undergoing elective coronary bypass grafting using a 4:1 blood:cardioplegia ratio, with hematocrit ranging between 20% and 22%. A standard cardioplegic delivery system (Shiley BCD Plus; Sorin Biomedical Inc, Irvine, CA) and a standard roller pump (Cobe-Stockert; Cobe Cardiovascular Inc, Harvada, CO) were employed to deliver the solution into the aortic root through an antegrade cardioplegia cannula with a pressure monitoring side port (ATCOllMV; Research Medical Inc, Salt Lake City, UT) or into the coronary sinus through a retrograde cardioplegia cannula with a pressure measuring side port (RC014T; Research Medical Inc). Intraoperative pressure measurements were recorded simultaneously from the pressure ports on the cannulas (electronic transducer; ViggoSpectromed, Oxnard, CA) and pressure ports in the cardioplegic system itself using a DLP 6000 monitor (DLP, Grand Rapids, MI) positioned at the heart-lung machine. Antegrade cardioplegic pressure was estimated by digital palpation of the ascending aorta by three different surgeons (each with experience in more than 1,000 cardiac 0003-4975/94/$7.00
HOW TO DO IT KATO ET AL INACCURATE INDIRECT PRESSURE MEASUREMENTS
Ann Thorae Surg 1994;58:1188-91
operations) while the perfusionist recorded the aortic pressure measured from the pressure monitoring port of the antegrade cardioplegic cannula. This information was not transmitted to the surgeon intraoperatively, to avoid having this knowledge influence the predicted pressure during the next antegrade infusion. At the conclusion of the surgical procedure, an in vitro calibration procedure was carried out by measuring the pressure within the delivery system while cardioplegia was infused through the antegrade and retrograde cannulas into a beaker on the operating table positioned at the heart level. The system was zeroed by taking into account the difference in height between the delivery system pressure port and the heart level. The predicted intravascular pressure was determined as the difference between the cardioplegic delivery system pressure measured during the surgical procedure and the pressure drop measured in the delivery system during the calibration procedure. These predictions were compared with the actual intravascular pressures recorded during intraoperative cardioplegic administration. After cardiopulmonary bypass, pressure measurements in the cardioplegic delivery system were made at a hematocrit of approximately 20%, the average intraoperative hematocrit of the cardioplegic solution in a patient with 25% hematocrit blood mixed 4:1 with cardioplegic solution. Asanguineous cardioplegia with zero hematocrit was simulated by administering normal saline solution to correspond to delivery of a crystalloid cardioplegic solution. Measurements were obtained at flow rates varying from 50 to 350 mL/min and 50 to 200 mL/min for antegrade and retrograde flow, respectively, and at temperatures 8°C and 37°C to correspond to conditions existing during operation. Data were analyzed with StatView V2.0 on an Apple Macintosh lIci. Analysis of variance was employed for comparisons between groups. The relationship between observed and predicted pressure was tested by linear regression analysis. Differences were considered significant at a probability level less than 0.05. Group data are expressed as mean ± standard error of the mean.
Results In Vitro Measurements Cardioplegic delivery system pressure never exceeded 120 mm Hg during in vitro calibration procedure at flow rates up to 350 mL/min. However, under all conditions, the cardioplegic delivery system pressure was higher with retrograde than antegrade delivery, as we used a retrograde cannula that was designed to impose a distal resistance to accomplish the self-inflating capacity of the balloon. At normothermia (37°C) varying hematocrit between 0% (crystalloid cardioplegia) and 20% (blood cardioplegia) produced negligible difference in cardioplegic delivery system pressure (Fig 1). Conversely, cardioplegic system pressure was significantly higher at all flow rates with cold blood cardioplegia (8°C) than with cold crystalloid cardioplegia, with the difference in pressure averaging 40 mm Hg at the maximum antegrade and retrograde flow rates (350
1189
140 120
Ci
100
:I:
E
..
.§.
80
Hct=20%
Gl
::l III III
e c.
60
~
~Hct=O%
40
20 0 0
50
100
150
200
250
300
350
400
Flow (ml/minute)
3rc
Fig 1. Delivery system pressure during in vitro infusion of blood and crystalloid cardioplegia at flow rates ranging from 50 to 350 ml.lmin. Note comparable delivery system pressure with normothermic blood and crystalloid perfusion. (n = 5 for each obseroaiion.)
mL/min and 250 mL/min, respectively), as shown in Figure 2. Pressure varied negligibly during the calibration procedure, attesting to the reproducibility of in vitro delivery system pressure measurement.
In Vivo Measurements Figure 3 shows the marked difference between predicted intravascular pressure and using the in vitro calibration subtracted from the cardioplegic delivery system pressure recorded during infusions and actual intravascular pressure during antegrade and retrograde delivery of blood cardioplegia at 8°C intraoperatively. Direct intravascular pressure differed from predicted pressure during antegrade delivery as much as -20 mm Hg and +70 mm Hg with antegrade delivery at flows ranging from 50 to 350 mL/min. These differences ranged from -20 mm Hg to + 15 mm Hg during retrograde delivery at flows ranging from 50 to 250 mL/min. Consequently, there was marked variability and limited correlation (r 2 = 0.56 and 0.76, respectively) between predicted and measured pressure with both routes of delivery. Figure 4 shows the difference between predicted aortic pressure, estimated by digital palpation of the aorta, and direct intraaortic pressure measurement. The correlation between predicted and observed pressure was weak and even more variable than when intravascular pressure was estimated by subtracting the calculated pressure drop in vitro from the observed delivery system pressure during intraoperative antegrade cardioplegic delivery.
Comment These data show that direct intravascular pressure measurement is the only reliable method for determining either aortic or coronary sinus pressure during cardiople-
1190
HOW TO DO IT KATO ET AL INACCURATE INDIRECT PRESSURE MEASUREMENTS
gic delivery, and document the inaccuracy of indirect estimation of p.ressure by either calibration of the delivery system or d~!Sttf!:f palpation of the aorta. The pressHn~-f1ow relationships for any cardioplegic delivery system are unique to that system and vary minimally during in vitro calibration. Delivery system pressure, however, is dependent on flow rate, temperature, viscosity, and compliance of the tubing system, which changes with its thickness, length, and temperature. Consequently, delivery system pressure with cold blood cardioplegia is higher because hypothermia increases blood viscosity and reduces tubing compliance and elastic recoil. The influence of viscosity lessens with normothermia, and the outflow resistance of the employed cannula becomes important (ie, antegrade versus retrograde cannulas). We
Ann Thorac Surg 1994;58:1188-91
250
_
200
Cl :I:
E
.§. l!!
150
Observed
:::l
VI VI
e :e0 a. o
«
100
50
0 0 140
50
100
150
250
200
Predicted Aortic Pressure (mmHg)
A 120 150
Hct=20%
Ci
100
Ci
80
.§.
J:
J:
E
.§.
e VI VI e a. :::l
I Predicted
E 125
e VI VI e a.
:::l 100
60 40
VI
75
:::l
C
C;;
20
~
I
50
~
0
50
100
150
A
200
250
300
350
400
0
0
Flow (ml/minute)
Ci E
§. l!! :::I
Hct = 20%
100
VI VI
/
80 60
l!!
D..
40
..
20 0 0
B
50
./
I
/ 100
~
/
150
/
;N
/
200
Hct = 0%
/
250
r2 = 0.76
1\
25
0
:I:
Observed
0
140 120
I
I
I
I
I /-
CO C
0
I
I
I
I
I
300
350
Flow (ml/minute)
Fig 2. Delivery system pressure during in vitro infusion of 8°C blood and crystalloid cardioplegia through antegrade (A) and retrograde (B) cannulas. Note higher delivery system pressure with hypothermic blood cardioplegia (p < 0.05 at all flow rates >50 ml.lmin) and negligible variability at each rate of perfusion. (n = 5 for each observation.)
B
25
50
75
100
125
150
Predicted Coronary Sinus Pressure (mmHg)
Fig 3. Comparison between measured intravascular aortic and coronary sinus pressure and predicted pressures during clinical antegrade (A) and retrograde (B) warm and cold blood cardioplegia perfusions. Hatched line of identity would occur if predicted and direct pressure were identical. These data for antegrade delivery are expressed by the equation y = 28.19 + 0.37x with an r value of 0.56, and for retrograde delivery by the equation y = 11.93 + 0.27x with an r value of 0.76. Both slopes and the intercepts are significantly (p < 0.05) different from the slope of identity.
tested only one type of antegrade and retrograde cannula in our system and would suspect similar directional changes with other cannulas and systems available commercially, especially because cannula resistance is low in manually inflating retrograde cannulas. This difference, however, would not be sufficient reason to avoid direct intravascular pressure measurements as will be discussed subsequently. In contrast to in vitro findings, significant variability of pressure-flow relationships occurred when cardioplegia
HOW TO DO IT KATO ET AL INACCURATE INDIRECT PRESSURE MEASUREMENTS
Ann Thorae Surg 1994;58:1188-91
100
Ci
:I:
• • • •
80
E
.s l!!~
UI UI
e a. :e0 (,)
«
.
•
90
• • • • • ,(Predicted • • .: Observed •• • .~ •r =0.32 2
.. ..
70 60
/
'/ • • •
./ •
50 40
/
/
30
• • •
/'
•
• •
20 +----,----,------,--,---,---,---,--_ 20
30
40
50
60
70
80
90
100
Palpated Aortic Pressure (mmHg)
Fig 4. Comparison between pressure estimated by aortic palpation and direct intraaortic pressure measured during antegrade infusions. Hatched line of identity would occur if predicted (or estimated) pressure and direct pressure were identical. These data are expressed by the equation y = 16.66 + 0.77x with an r 2 value of 0.32. Both the slope and intercept are significantly different (p < 0.05) from the slope of identity.
was delivered into the aorta or coronary sinus intraoperatively, because intravascular resistance changes independently from the pressure drop imparted by the delivery system containing the antegrade and retrograde cannulas. This is especially true with antegrade blood cardioplegia, because patients with coronary disease have variable degrees of fixed vascular obstruction and aortic compliance. Furthermore, arteriolar vascular resistance can change from dose to dose in the same heart, which also may generate different vascular pressures at similar flow rates, especially when high initial flow rates are used to induce antegrade cardioplegic arrest. Conversely, the relative lower variance of predicted versus measured pressure for retrograde cardioplegia is related probably to the use of lower maximum flow rates (200 to 250 mL/min) into a venous bed that is devoid of obstruction, unless the heart is retracted and there is physical distortion of the coronary sinus outflow. Variability of intravascular pressure increased as flow rate was raised with both antegrade and retrograde blood cardioplegic delivery, but the lack of correlation between predicted and measured intravascular pressure existed at all delivery rates (see Fig 3). Furthermore, there was no correlation between palpated aortic pressure and measured aortic pressure (see Fig 4). These observations reinforce the need to measure pressure directly from the aorta during antegrade delivery. A low aortic pressure (ie, <30 mm Hg) may signify unsuspected aortic insufficiency or aortic valve incompetence produced
1191
by a malpositioned aortic clamp, distortion of the noncoronary cusp by the venous cannula, or inadvertent failure to close the vent clamp or to switch from the retrograde to the antegrade route of delivery. It would seem impossible to estimate pressure by palpation of the coronary venous system, so that direct pressure measurement is essential. Direct coronary sinus pressure measurement is important so that the perfusionist can tell the surgeon if the pressure is too high (ie, >50 mm Hg) or too low «15 mm Hg). If the coronary sinus pressure exceeds 50 mm Hg, the perfusionist can reduce the flow to prevent disruption of the coronary sinus or production of myocardial edema while simultaneously alerting the surgeon that the catheter may be wedged and require either retraction or repositioning. Communication that the coronary sinus pressure is less than 15 mm Hg will allow the surgeon to determine if the cannula has retracted into the right atrium, the balloon is not inflated, or there is a persistent left superior vena cava. Our findings should not be interpreted to discourage measuring pressure in the cardioplegic delivery system. This measurement should be monitored routinely, because obstruction in the delivery system (ie, kinking or failure to open a flow clamp) will raise delivery system pressure and produce subsequent disruption of the system if not remedied immediately by stopping the flow and determining the source of obstruction. We recommend, therefore, that both cardioplegic delivery system pressure be monitored to detect such potential disruption and intravascular pressure be measured to optimize delivery of either antegrade or retrograde cardioplegia. Additionally, we continue to palpate the aorta during antegrade delivery and observe the distention of the coronary veins during retrograde delivery, as these clinical tools remain useful deterrents to total reliance upon strain gauges that occasionally can be calibrated incorrectly. The availability of pressure measuring capability in current antegrade and retrograde cardioplegic cannulas allows for direct measurements intraoperatively, and this study shows that neither in vitro calibration nor direct palpation will provide this information. These observations have led to our measurement of aortic and coronary sinus pressure whenever cardioplegia is delivered either into the aorta or into the coronary sinus. Direct intravascular pressure measurements circumvent the errors imparted by indirect assessments and provide for reliable pressure measurement, and hopefully will improve the effectiveness and safety of antegrade and retrograde cardioplegic delivery.
References 1. Buckberg GO. Myocardial protection: Thorac Cardiovasc Surg 1993;5:98-106. 2. Robinson LA. Cardioplegia solutions in spectives and national trends. Presented vation: Current Technology and Future Oct 16, 1992.
an overview. Semin the 90's: current perat Myocardial PreserTrends, Atlanta, GA,