- Email: [email protected]
An analysis of direct and indirect measurements of left atrial filling pressure
Volume 71, Number 5
AlGV
1 9 / 6
The Journal of T H O R A C I C A N D
CARDIOVASCULAR
SURGERY
An analysis of direct and indirect measurements of left atrial filling pressure Left ventricular function may be assessed by direct catheter measurements of left atrial pressure or by indirect measurements of pulmonary artery wedge pressure or pulmonary artery end-diastolic pressure. Controversy exists as to how closely the indirect measurements correlate with true left atrial pressure and to which is the most accurate. To clarify this problem, we studied 43 patients undergoing cardiac surgical procedures with cardiopulmonary bypass. Both left atrial catheters for direct measurement and Swan-Ganz catheters were placed at the time of surgery. All patients were monitored continuously for 48 hours and hourly measurements were recorded. The resultant 1,620 left atrial pressure and pulmonary artery wedge pressure figures and 1,860 left atrial pressure and pulmonary artery end-diastolic wedge pressure measurements were subjected to computer analysis. The following conclusions have been found: (1) Pulmonary artery wedge pressure is a better indirect measure of left atrial pressure than is pulmonary artery end-diastolic wedge pressure (pooled correlation coefficient 0.629); (2) direct left atrial pressure measurement is more reliable and has fewer complications than indirect measurements; (3) there is no consistent correlation between left atrial pressure and central venous pressure (pooled correlation coefficient 0.3). A discussion of our results and the problems associated with left atrial catheters and Swan-Ganz catheters is presented.
Chester B. Humphrey, Commander (MC) USNR, James H. Oury, Commander (MC) USN, Richard W. Virgilio, Commander (MC) USN, James A. Gibbons, Commander (MC) USN, Theodore L. Folkerth, Commander (MC) USN, Alan R. Shapiro, M.D., and Richard G. Fosburg, Captain (MC) USN, San Diego, Calif.
From the Naval Regional Medical Center (Department of Cardiothoracic Surgery; Trauma Research Unit, Department of General Surgery; and Clinical Investigation Center), San Diego, Calif., and from the Division of Thoracic Surgery, University of California at San Diego, Calif.
The opinions or assertions expressed herein are those of the authors and are not to be construed as official or as reflecting the views of the Navy Department or the naval service at large. R e a d a t t n e F i r s t A n n u a l Meeting of The Samson Thoracic Surgical Society, Santa Barbara, Calif., May 28-30, 1975.
Supported by Office of Naval Research Contract N-0014-690A0200-6046 and by U. S. Public Health Service Grants GM 17284 and HL 13172.
Address for reprints: Chester B. Humphrey, Commander (MC) USNR, Department of Cardiothoracic Surgery, Naval Regional Medical Center, San Diego, Calif. 92134.
643
644
The Journal of Thoracic and Cardiovascular Surgery
Humphrey et al.
Left ventricular function and subsequently cardiac output are determined by heart rate, myocardial contractility, and filling pressure.1 In the postoperative patient, heart rate can be monitored electrocardiographically and adjusted with chronotropic drugs or with a pacemaker, while myocardial contractility can be adjusted with inotropic drugs.2 The filling pressure, or left ventricular end-diastolic pressure, is a guide to end-diastolic volume and wall tension and thus determines left ventricular function.3 This pressure can be obtained by measuring left atrial pressure and, in the absence of significant mitral valve disease, can be used as an additional means of following cardiac performance. Thus the only nonmeasurable variable of left ventricular function is the state of myocardial contractility. During the 1960's, central venous pressure was introduced and accepted as a means of measuring fluid replacement and adequacy of blood volume in the cardiac patient.4 Central venous pressure measured at the bedside provided a faster and easier measurement than calculation of total blood volume. The value of left atrial pressure was recognized but required either thoracotomy or transeptal puncture for direct measurements.5 With the introduction of the pulmonary artery balloon catheter, left atrial pressure measurements could be obtained indirectly from the pulmonary artery capillary level as a pulmonary artery wedge pressure.6 Results of comparison studies between these various pressures were initially conflicting. Bouchard and co-authors7 showed that no estimate of left ventricular end-diastolic pressure could be made from the pulmonary artery end-diastolic pressure and that pulmonary artery end-diastolic pressure often failed to reflect acute alterations in left ventricular end-diastolic pressure. Falicov and Resnekov,8 performing similar studies, demonstrated that left ventricular end-diastolic pressure did not correlate well with pulmonary artery enddiastolic pressure in patients with left ventricular dysfunction, although there was a correlation between pulmonary artery end-diastolic pressure and pulmonary artery wedge pressure. A study conducted on patients who were in shock revealed that the pulmonary artery end-diastolic pressure did reflect the left ventricular end-diastolic pressure and was a useful therpeutic tool.9 Herbert10 demonstrated that the limitation of pulmonary artery end-diastolic pressure as a reflection of left ventricular end-diastolic pressure was related to pulmonary vascular disease, right bundle branch block, aortic or mitral valvular disease, and a left atrial pressure below the critical closing pressure of the pulmonary capillary
Table I. Patients studied No. of patients Valve replacement Mitral Aortic Double (aortic and mitral) Coronary artery bypass Resection of left ventricular aneurysm Congenital
18 13 4 1 22 2
Total
43
1
bed. Retrospective studies at Duke University showed a significant correlation between pulmonary artery wedge and left atrial pressure when the wedge pressure was in the lower ranges of normal. There was considerable error when the wedge pressure was higher. n In order to assess further the correlation between direct and indirect measurements of left atrial filling pressures, we compared left atrial pressure, pulmonary capillary wedge pressure, and pulmonary artery enddiastolic pressure in 43 patients subjected to a variety of cardiac procedures with cardiopulmonary bypass. Methods and clinical material The clinical material (Table I) comprised 18 patients undergoing valve replacements: 13 mitral, 4 aortic, and one mitral plus aortic; 22 patients with aorta-coronary bypasses to single or multiple vessels; 2 patients undergoing left ventricular aneurysmectomy and concomitant aorta-coronary bypass; and one patient who had closure of an atrial septal defect. There were 30 male and 13 female patients. Selection for study was on a nearly random basis determined by the availability of a bed in the Trauma Research Unit on a day when a patient was scheduled for cardiac surgery. An element of selection did exist when, on a given day, two cardiac procedures were scheduled and there was room to study only one patient. In most instances, the patient selected was the more critically ill of the two. All patients were operated upon by the same surgical team in a similar fashion. Before the patient was weaned from cardiopulmonary bypass, a 17 gauge polyethylene long-line catheter was introduced percutaneously through the right anterior chest wall into the mediastinum and then placed into the left atrium via the right superior pulmonary vein, to which it was secured with a purse-string suture. In a few patients the catheter was brought directly out through the inferior
Volume 71 Number 5 May, 1976
Left atrial filling pressure
RLL PRTIENTS
RLL PATIENTS
Y = 0=S25*X - 0=027 (POOLED EQN„] POOLED CORRo C0EFF=: 0=623
K5 1
' l£o
PULMONARY
'
r
i3To
CAPILLARY
t
3>
,
IO.O
WEDGE PRESSURE
645
,
CHANGE
Y = 0=372«X - 0=008 (POOLED EQN=] POOLED C0RR= C0EFF=: 0=551
,—|
».o
(mm HG)
Fig. 1. Pooled left atrial pressure changes plotted against the pooled pulmonary capillary wedge pressure changes with the line of regression plotted. aspect of the midline incision. Direct left atrial pressures were thereby obtained. In all patients, a Swan-Ganz catheter was floated into the pulmonary artery for measurement of the pulmonary artery and pulmonary artery wedge pressures. During the early part of our study these catheters were introduced approximately 2 to 4 hours after arrival in the Trauma Research Unit, when the patient had achieved a relatively stable state. Subsequently, in the majority of patients, the Swan-Ganz catheter was inserted in the operating room via the right internal jugular vein after the induction of general anesthesia. In addition, in all patients, a 14 gauge long-line polyethylene catheter was inserted percutaneously via the left brachial vein for intra- and postoperative measurements of central venous pressure. An 18 gauge polyethylene catheter was placed percutaneously in the radial artery for monitoring arterial blood pressure and determining blood gas values. All patients were left undisturbed for 5 minutes prior to each hourly interval, at which time pressure readings were obtained. All measurements were recorded on a Hewlett-Packard Model 7700 monitor and recorder. All patients had mechanical volume ventilation for at
-26.0
*
PULMONARY
-id.o
'
DIASTOLIC
3
'
PRESSURE
ioTo CHANGE
'
iSo
(mm HG)
Fig. 2. Pooled left atrial pressure changes plotted against the pooled pulmonary diastolic pressure changes with the line of regression plotted. least 12 hours. The majority were extubated on the first postoperative day, although in a few patients respiratory insufficiency necessitated prolonged ventilator support. No patients were subjected to positive end-expiratory pressure. At the completion of the study the available data were analyzed by the Computer Center of the University of California at San Diego. Results There were 1,620 data points, comparing left atrial pressure with pulmonary artery wedge pressure, and 1,860 data points for left atrial pressure and pulmonary artery end-diastolic pressure. Ninety-five per cent of all changes in left atrial and pulmonary artery wedge pressure were between - 5 and +5 mm. Hg, whereas 95 per cent of the pulmonary artery end-diastolic changes were between —6 and +6 mm. Hg. Central venous pressure. There was no significant correlation between central venous pressure and left atrial pressure, pulmonary capillary wedge pressure, or pulmonary artery end-diastolic pressure.
646
The Journal of Thoracic and Cardiovascular Surgery
Humphrey et at.
rzzi PAW
S*55? PAEDP
100
89.3
81,7
80
63.1 un
60
—
40
24 20
—
±1 LAP
MINUS
PAW
OR
PAEDP
(mm
HG)
Fig. 3. Correlation of direct and indirect measurements of left atrial pressure (LAP). This graph reflects the percentage of time that the LAP was identical to the pulmonary artery wedge (PA W) and pulmonary artery end-diastolic pressure (PAEDP) or within ± 1 to ±4 mm. Hg. Table II. Complications in 43 patients Patients No. Swan-Ganz catheter Correctable "Failures"
Per cent
4 8
9 19
12
28
Left atrial catheter Broken on removal "Failures"
2 1
5 2
Total
3
7
Total
Left atrial pressure. The pooled correlation coefficient for the 1,620 observations of left atrial and pulmonary artery wedge pressures (Fig. 1) was 0.629, with the pooled regression line being Y = 0.525 x -0.027. The pooled correlation coefficient for the 1,860 left atrial pressure and pulmonary artery
end-diastolic pressure data (Fig. 2) was only 0.551, and the pooled regression line was Y = 0.327 x —0.008. Reliability of a change in an indirect measurement. Left atrial pressure was identical to pulmonary capillary wedge pressure and pulmonary artery enddiastolic pressure in 24 per cent and 18.5 per cent of the measurements, respectively. It was within ±4 mm. Hg of the pulmonary capillary wedge pressure and pulmonary artery end-diastolic pressure in 94.6 per cent and 89 per cent of the measurements, respectively (Fig. 3). Problems occurred with the Swan-Ganz catheter in over one fourth of the 43 patients, and in 19 per cent (8 patients) it was not possible to obtain complete measurements for the 48 hour study period. Most of the catheter failures were due to intraluminal clotting or broken balloons. Additionally, one patient accidentally pulled out the Swan-Ganz catheter, and another patient selected for study could not be included after several unsuccessful attempts to insert the Swan-Ganz catheter.
Volume 71 Number 5 May, 1976
One left atrial catheter failed to function properly for the 48 hour study period. This catheter had been brought out through the inferior aspect of the midline incision and was occluded by a fascial suture. Complications associated with removal of the left atrial catheter occurred twice and were limited to breakage of the catheter. Only one broken catheter required surgical removal under local anesthesia (Table II). A comparison of those patients undergoing valve replacement with those undergoing aorta-coronary bypass produced no statistically significant differences. In addition, the degree of left ventricular dysfunction preoperatively had no significant bearing on whether there was a good or poor correlation between left atrial pressure and pulmonary artery wedge or pulmonary artery end-diastolic pressures. Discussion This study supports the report of Lappas and associates, 12 whose 161 measurements in 18 patients showed that the pulmonary artery wedge pressure more closely reflected left atrial pressure than did the pulmonary artery end-diastolic pressure. In the absence of direct measurement of left ventricular end-diastolic pressure, the left atrial pressure measurement remains the next best alternative. Since it is rare clinically to be able to obtain direct left atrial pressure, our results (Fig. 3) demonstrate that the indirect left atrial pressure measurements are a useful and reliable reflection of left atrial pressure. The most common malfunction of the Swan-Ganz catheter was the inability to obtain a pulmonary wedge pressure. However, since the pulmonary artery end-diastolic pressure still has a relatively good correlation, the Swan-Ganz catheter should not be abandoned as a useful tool in the management of patients. The following conclusions can be made: 1. The Swan-Ganz catheter is a reliable clinical tool for estimation of left atrial pressure in patients with cardiac disease. 2. Determination of the pulmonary artery wedge pressure is a more reliable indirect measurement of left atrial pressure than is the pulmonary artery enddiastolic pressure measurement. 3. There is no consistent correlation between the left atrial pressure and the central venous pressure. REFERENCES 1 Oldham, H. N., Jr., Wechsler, A. S., Wolfe, W. G., Anderson, R. W., and Sabiston, D. C , Jr.: Left
Left atrial filling pressure
647
Ventricular Filling Pressure After Aorto-coronary Grafting, J. THORAC. CARDIOVASC. SURG. 65: 343,
1973.
2 Sarin, C. L., Yalav, E., Clement, A. J., and Braimbridge, M. V.: The Necessity for Measurement of Left Atrial Pressure After Cardiac Valve Surgery, Thorax 25: 185, 1970. 3 Braunwald, E., and Ross, J., Jr.: The Ventricular End-Diastolic Pressure: Appraisal of Its Valve in the Recognition of Ventricular Failure in Man, Am. J. Med. 34: 147, 1963. 4 Gazzaniga, A. B., Byrd, C. L., Stewart, D. R., and O'Connor, N. E.: Evaluation of Central Venous Pressure as a Guide to Volume Replacement in Children Following Cardiopulmonary Bypass, Ann. Thorac. Surg. 13: 148, 1972. 5 Braunwald, E., Brockenbrough, E. C , Frahm, C. J., and Ross, J., Jr.: Left Atrial and Left Ventricular Pressures in Subjects Without Cardiovascular Disease: Observations on Eighteen Patients Studied by Transeptal Left Heart Catheterization, Circulation 24: 267, 1961. 6 Swan, H. J. C , Ganz, W., Forrester, J., Marcus, H., Diamond, G., and Chonette, D.: Catheterization of the Heart in Man With Use of a Flow-Directed BalloonTipped Catheter, N. Engl. J. Med. 283: 447, 1970. 7 Bouchard, R. J., Gault, J. H., and Ross, J., Jr.: Evaluation of Pulmonary Arterial End-Diatolic Pressure as an Estimate of Left Ventricular End-Diastolic Pressure in Patients With Normal and Abnormal Left Ventricular Performance, Circulation 44: 1072, 1971. 8 Falicov, R. E., and Resnekov, L.: Relationship of the Pulmonary Artery End-Diastolic Pressure to the Left Ventricular End-Diastolic and Mean Filling Pressures in Patients With and Without Left Ventricular Dysfunction, Circulation 42: 65, 1970. 9 Scheinman, M. Evans, G. T., Weiss, A., and Rapaport, E.: Relationship Between Pulmonary Artery EndDiastolic Pressure and Left Ventricular Filling Pressure in Patients in Shock, Circulation 47: 317, 1973. 10 Herbert, W. H.: Limitations of Pulmonary Artery End-Diastolic Pressure as the Reflection of Left Ventricular End-Diastolic Pressure, N.Y. State J. Med. 72: 229, 1972. 11 Walston, A., II, and Kendall, M. E.: Comparison of Pulmonary Wedge and Left Atrial Pressure in Man, Am. Heart J. 86: 159, 1973. 12 Lappas, D., Lell, W. A., Gabel, J. E., Civetta, J. M., and Lowenstein, E.: Indirect Measurement of Left-Atrial Pressure in Surgical Patients—Pulmonary-Capillary Wedge and Pulmonary-Artery Diastolic Pressures Compared With Left-Atrial Pressure, Anesthesiology 38: 394, 1973.
AlGV
1 9 / 6
The Journal of T H O R A C I C A N D
CARDIOVASCULAR
SURGERY
An analysis of direct and indirect measurements of left atrial filling pressure Left ventricular function may be assessed by direct catheter measurements of left atrial pressure or by indirect measurements of pulmonary artery wedge pressure or pulmonary artery end-diastolic pressure. Controversy exists as to how closely the indirect measurements correlate with true left atrial pressure and to which is the most accurate. To clarify this problem, we studied 43 patients undergoing cardiac surgical procedures with cardiopulmonary bypass. Both left atrial catheters for direct measurement and Swan-Ganz catheters were placed at the time of surgery. All patients were monitored continuously for 48 hours and hourly measurements were recorded. The resultant 1,620 left atrial pressure and pulmonary artery wedge pressure figures and 1,860 left atrial pressure and pulmonary artery end-diastolic wedge pressure measurements were subjected to computer analysis. The following conclusions have been found: (1) Pulmonary artery wedge pressure is a better indirect measure of left atrial pressure than is pulmonary artery end-diastolic wedge pressure (pooled correlation coefficient 0.629); (2) direct left atrial pressure measurement is more reliable and has fewer complications than indirect measurements; (3) there is no consistent correlation between left atrial pressure and central venous pressure (pooled correlation coefficient 0.3). A discussion of our results and the problems associated with left atrial catheters and Swan-Ganz catheters is presented.
Chester B. Humphrey, Commander (MC) USNR, James H. Oury, Commander (MC) USN, Richard W. Virgilio, Commander (MC) USN, James A. Gibbons, Commander (MC) USN, Theodore L. Folkerth, Commander (MC) USN, Alan R. Shapiro, M.D., and Richard G. Fosburg, Captain (MC) USN, San Diego, Calif.
From the Naval Regional Medical Center (Department of Cardiothoracic Surgery; Trauma Research Unit, Department of General Surgery; and Clinical Investigation Center), San Diego, Calif., and from the Division of Thoracic Surgery, University of California at San Diego, Calif.
The opinions or assertions expressed herein are those of the authors and are not to be construed as official or as reflecting the views of the Navy Department or the naval service at large. R e a d a t t n e F i r s t A n n u a l Meeting of The Samson Thoracic Surgical Society, Santa Barbara, Calif., May 28-30, 1975.
Supported by Office of Naval Research Contract N-0014-690A0200-6046 and by U. S. Public Health Service Grants GM 17284 and HL 13172.
Address for reprints: Chester B. Humphrey, Commander (MC) USNR, Department of Cardiothoracic Surgery, Naval Regional Medical Center, San Diego, Calif. 92134.
643
644
The Journal of Thoracic and Cardiovascular Surgery
Humphrey et al.
Left ventricular function and subsequently cardiac output are determined by heart rate, myocardial contractility, and filling pressure.1 In the postoperative patient, heart rate can be monitored electrocardiographically and adjusted with chronotropic drugs or with a pacemaker, while myocardial contractility can be adjusted with inotropic drugs.2 The filling pressure, or left ventricular end-diastolic pressure, is a guide to end-diastolic volume and wall tension and thus determines left ventricular function.3 This pressure can be obtained by measuring left atrial pressure and, in the absence of significant mitral valve disease, can be used as an additional means of following cardiac performance. Thus the only nonmeasurable variable of left ventricular function is the state of myocardial contractility. During the 1960's, central venous pressure was introduced and accepted as a means of measuring fluid replacement and adequacy of blood volume in the cardiac patient.4 Central venous pressure measured at the bedside provided a faster and easier measurement than calculation of total blood volume. The value of left atrial pressure was recognized but required either thoracotomy or transeptal puncture for direct measurements.5 With the introduction of the pulmonary artery balloon catheter, left atrial pressure measurements could be obtained indirectly from the pulmonary artery capillary level as a pulmonary artery wedge pressure.6 Results of comparison studies between these various pressures were initially conflicting. Bouchard and co-authors7 showed that no estimate of left ventricular end-diastolic pressure could be made from the pulmonary artery end-diastolic pressure and that pulmonary artery end-diastolic pressure often failed to reflect acute alterations in left ventricular end-diastolic pressure. Falicov and Resnekov,8 performing similar studies, demonstrated that left ventricular end-diastolic pressure did not correlate well with pulmonary artery enddiastolic pressure in patients with left ventricular dysfunction, although there was a correlation between pulmonary artery end-diastolic pressure and pulmonary artery wedge pressure. A study conducted on patients who were in shock revealed that the pulmonary artery end-diastolic pressure did reflect the left ventricular end-diastolic pressure and was a useful therpeutic tool.9 Herbert10 demonstrated that the limitation of pulmonary artery end-diastolic pressure as a reflection of left ventricular end-diastolic pressure was related to pulmonary vascular disease, right bundle branch block, aortic or mitral valvular disease, and a left atrial pressure below the critical closing pressure of the pulmonary capillary
Table I. Patients studied No. of patients Valve replacement Mitral Aortic Double (aortic and mitral) Coronary artery bypass Resection of left ventricular aneurysm Congenital
18 13 4 1 22 2
Total
43
1
bed. Retrospective studies at Duke University showed a significant correlation between pulmonary artery wedge and left atrial pressure when the wedge pressure was in the lower ranges of normal. There was considerable error when the wedge pressure was higher. n In order to assess further the correlation between direct and indirect measurements of left atrial filling pressures, we compared left atrial pressure, pulmonary capillary wedge pressure, and pulmonary artery enddiastolic pressure in 43 patients subjected to a variety of cardiac procedures with cardiopulmonary bypass. Methods and clinical material The clinical material (Table I) comprised 18 patients undergoing valve replacements: 13 mitral, 4 aortic, and one mitral plus aortic; 22 patients with aorta-coronary bypasses to single or multiple vessels; 2 patients undergoing left ventricular aneurysmectomy and concomitant aorta-coronary bypass; and one patient who had closure of an atrial septal defect. There were 30 male and 13 female patients. Selection for study was on a nearly random basis determined by the availability of a bed in the Trauma Research Unit on a day when a patient was scheduled for cardiac surgery. An element of selection did exist when, on a given day, two cardiac procedures were scheduled and there was room to study only one patient. In most instances, the patient selected was the more critically ill of the two. All patients were operated upon by the same surgical team in a similar fashion. Before the patient was weaned from cardiopulmonary bypass, a 17 gauge polyethylene long-line catheter was introduced percutaneously through the right anterior chest wall into the mediastinum and then placed into the left atrium via the right superior pulmonary vein, to which it was secured with a purse-string suture. In a few patients the catheter was brought directly out through the inferior
Volume 71 Number 5 May, 1976
Left atrial filling pressure
RLL PRTIENTS
RLL PATIENTS
Y = 0=S25*X - 0=027 (POOLED EQN„] POOLED CORRo C0EFF=: 0=623
K5 1
' l£o
PULMONARY
'
r
i3To
CAPILLARY
t
3>
,
IO.O
WEDGE PRESSURE
645
,
CHANGE
Y = 0=372«X - 0=008 (POOLED EQN=] POOLED C0RR= C0EFF=: 0=551
,—|
».o
(mm HG)
Fig. 1. Pooled left atrial pressure changes plotted against the pooled pulmonary capillary wedge pressure changes with the line of regression plotted. aspect of the midline incision. Direct left atrial pressures were thereby obtained. In all patients, a Swan-Ganz catheter was floated into the pulmonary artery for measurement of the pulmonary artery and pulmonary artery wedge pressures. During the early part of our study these catheters were introduced approximately 2 to 4 hours after arrival in the Trauma Research Unit, when the patient had achieved a relatively stable state. Subsequently, in the majority of patients, the Swan-Ganz catheter was inserted in the operating room via the right internal jugular vein after the induction of general anesthesia. In addition, in all patients, a 14 gauge long-line polyethylene catheter was inserted percutaneously via the left brachial vein for intra- and postoperative measurements of central venous pressure. An 18 gauge polyethylene catheter was placed percutaneously in the radial artery for monitoring arterial blood pressure and determining blood gas values. All patients were left undisturbed for 5 minutes prior to each hourly interval, at which time pressure readings were obtained. All measurements were recorded on a Hewlett-Packard Model 7700 monitor and recorder. All patients had mechanical volume ventilation for at
-26.0
*
PULMONARY
-id.o
'
DIASTOLIC
3
'
PRESSURE
ioTo CHANGE
'
iSo
(mm HG)
Fig. 2. Pooled left atrial pressure changes plotted against the pooled pulmonary diastolic pressure changes with the line of regression plotted. least 12 hours. The majority were extubated on the first postoperative day, although in a few patients respiratory insufficiency necessitated prolonged ventilator support. No patients were subjected to positive end-expiratory pressure. At the completion of the study the available data were analyzed by the Computer Center of the University of California at San Diego. Results There were 1,620 data points, comparing left atrial pressure with pulmonary artery wedge pressure, and 1,860 data points for left atrial pressure and pulmonary artery end-diastolic pressure. Ninety-five per cent of all changes in left atrial and pulmonary artery wedge pressure were between - 5 and +5 mm. Hg, whereas 95 per cent of the pulmonary artery end-diastolic changes were between —6 and +6 mm. Hg. Central venous pressure. There was no significant correlation between central venous pressure and left atrial pressure, pulmonary capillary wedge pressure, or pulmonary artery end-diastolic pressure.
646
The Journal of Thoracic and Cardiovascular Surgery
Humphrey et at.
rzzi PAW
S*55? PAEDP
100
89.3
81,7
80
63.1 un
60
—
40
24 20
—
±1 LAP
MINUS
PAW
OR
PAEDP
(mm
HG)
Fig. 3. Correlation of direct and indirect measurements of left atrial pressure (LAP). This graph reflects the percentage of time that the LAP was identical to the pulmonary artery wedge (PA W) and pulmonary artery end-diastolic pressure (PAEDP) or within ± 1 to ±4 mm. Hg. Table II. Complications in 43 patients Patients No. Swan-Ganz catheter Correctable "Failures"
Per cent
4 8
9 19
12
28
Left atrial catheter Broken on removal "Failures"
2 1
5 2
Total
3
7
Total
Left atrial pressure. The pooled correlation coefficient for the 1,620 observations of left atrial and pulmonary artery wedge pressures (Fig. 1) was 0.629, with the pooled regression line being Y = 0.525 x -0.027. The pooled correlation coefficient for the 1,860 left atrial pressure and pulmonary artery
end-diastolic pressure data (Fig. 2) was only 0.551, and the pooled regression line was Y = 0.327 x —0.008. Reliability of a change in an indirect measurement. Left atrial pressure was identical to pulmonary capillary wedge pressure and pulmonary artery enddiastolic pressure in 24 per cent and 18.5 per cent of the measurements, respectively. It was within ±4 mm. Hg of the pulmonary capillary wedge pressure and pulmonary artery end-diastolic pressure in 94.6 per cent and 89 per cent of the measurements, respectively (Fig. 3). Problems occurred with the Swan-Ganz catheter in over one fourth of the 43 patients, and in 19 per cent (8 patients) it was not possible to obtain complete measurements for the 48 hour study period. Most of the catheter failures were due to intraluminal clotting or broken balloons. Additionally, one patient accidentally pulled out the Swan-Ganz catheter, and another patient selected for study could not be included after several unsuccessful attempts to insert the Swan-Ganz catheter.
Volume 71 Number 5 May, 1976
One left atrial catheter failed to function properly for the 48 hour study period. This catheter had been brought out through the inferior aspect of the midline incision and was occluded by a fascial suture. Complications associated with removal of the left atrial catheter occurred twice and were limited to breakage of the catheter. Only one broken catheter required surgical removal under local anesthesia (Table II). A comparison of those patients undergoing valve replacement with those undergoing aorta-coronary bypass produced no statistically significant differences. In addition, the degree of left ventricular dysfunction preoperatively had no significant bearing on whether there was a good or poor correlation between left atrial pressure and pulmonary artery wedge or pulmonary artery end-diastolic pressures. Discussion This study supports the report of Lappas and associates, 12 whose 161 measurements in 18 patients showed that the pulmonary artery wedge pressure more closely reflected left atrial pressure than did the pulmonary artery end-diastolic pressure. In the absence of direct measurement of left ventricular end-diastolic pressure, the left atrial pressure measurement remains the next best alternative. Since it is rare clinically to be able to obtain direct left atrial pressure, our results (Fig. 3) demonstrate that the indirect left atrial pressure measurements are a useful and reliable reflection of left atrial pressure. The most common malfunction of the Swan-Ganz catheter was the inability to obtain a pulmonary wedge pressure. However, since the pulmonary artery end-diastolic pressure still has a relatively good correlation, the Swan-Ganz catheter should not be abandoned as a useful tool in the management of patients. The following conclusions can be made: 1. The Swan-Ganz catheter is a reliable clinical tool for estimation of left atrial pressure in patients with cardiac disease. 2. Determination of the pulmonary artery wedge pressure is a more reliable indirect measurement of left atrial pressure than is the pulmonary artery enddiastolic pressure measurement. 3. There is no consistent correlation between the left atrial pressure and the central venous pressure. REFERENCES 1 Oldham, H. N., Jr., Wechsler, A. S., Wolfe, W. G., Anderson, R. W., and Sabiston, D. C , Jr.: Left
Left atrial filling pressure
647
Ventricular Filling Pressure After Aorto-coronary Grafting, J. THORAC. CARDIOVASC. SURG. 65: 343,
1973.
2 Sarin, C. L., Yalav, E., Clement, A. J., and Braimbridge, M. V.: The Necessity for Measurement of Left Atrial Pressure After Cardiac Valve Surgery, Thorax 25: 185, 1970. 3 Braunwald, E., and Ross, J., Jr.: The Ventricular End-Diastolic Pressure: Appraisal of Its Valve in the Recognition of Ventricular Failure in Man, Am. J. Med. 34: 147, 1963. 4 Gazzaniga, A. B., Byrd, C. L., Stewart, D. R., and O'Connor, N. E.: Evaluation of Central Venous Pressure as a Guide to Volume Replacement in Children Following Cardiopulmonary Bypass, Ann. Thorac. Surg. 13: 148, 1972. 5 Braunwald, E., Brockenbrough, E. C , Frahm, C. J., and Ross, J., Jr.: Left Atrial and Left Ventricular Pressures in Subjects Without Cardiovascular Disease: Observations on Eighteen Patients Studied by Transeptal Left Heart Catheterization, Circulation 24: 267, 1961. 6 Swan, H. J. C , Ganz, W., Forrester, J., Marcus, H., Diamond, G., and Chonette, D.: Catheterization of the Heart in Man With Use of a Flow-Directed BalloonTipped Catheter, N. Engl. J. Med. 283: 447, 1970. 7 Bouchard, R. J., Gault, J. H., and Ross, J., Jr.: Evaluation of Pulmonary Arterial End-Diatolic Pressure as an Estimate of Left Ventricular End-Diastolic Pressure in Patients With Normal and Abnormal Left Ventricular Performance, Circulation 44: 1072, 1971. 8 Falicov, R. E., and Resnekov, L.: Relationship of the Pulmonary Artery End-Diastolic Pressure to the Left Ventricular End-Diastolic and Mean Filling Pressures in Patients With and Without Left Ventricular Dysfunction, Circulation 42: 65, 1970. 9 Scheinman, M. Evans, G. T., Weiss, A., and Rapaport, E.: Relationship Between Pulmonary Artery EndDiastolic Pressure and Left Ventricular Filling Pressure in Patients in Shock, Circulation 47: 317, 1973. 10 Herbert, W. H.: Limitations of Pulmonary Artery End-Diastolic Pressure as the Reflection of Left Ventricular End-Diastolic Pressure, N.Y. State J. Med. 72: 229, 1972. 11 Walston, A., II, and Kendall, M. E.: Comparison of Pulmonary Wedge and Left Atrial Pressure in Man, Am. Heart J. 86: 159, 1973. 12 Lappas, D., Lell, W. A., Gabel, J. E., Civetta, J. M., and Lowenstein, E.: Indirect Measurement of Left-Atrial Pressure in Surgical Patients—Pulmonary-Capillary Wedge and Pulmonary-Artery Diastolic Pressures Compared With Left-Atrial Pressure, Anesthesiology 38: 394, 1973.