|
|
||||||||||
|
J Am Coll Cardiol, 1998; 32:840-864 © 1998 by the American College of Cardiology Foundation |
a Writing Group Member, USA
b Technology and Practice Executive Committee, USA
| Contents |
|---|
|
|
|---|
| Statement of technology practice Executive Committee |
|---|
|
|
|---|
| Present use of bedside right heart catheterization in patients with cardiac disease |
|---|
|
|
|---|
Evaluation of case-control study by Connors et al. (1)
The Expert Consensus Committee addresses the following issues:
1) RHC is a diagnostic and monitoring device and as such potentially affects outcome only to the extent that it triggers an intervention. Computed tomographic scanning in patients with closed head trauma may well select a population at greater risk but does not confer a higher mortality because of the diagnostic procedure itself. 2) Unexplained clinical factors that prompt clinicians to perform RHC may contribute to the risk of mortality and increased resource consumption. The decision to perform RHC is itself a marker that identifies severely ill patients. Although Connors et al. (1) attempted to eliminate what they term treatment selection bias, the Committee cannot conceive of any method, with the possible exception of randomization, by which such selection bias might be eliminated. Moreover, the computation of the propensity score did not include a measure of responses to therapy before the performance of RHC. 3) The effect of unmeasured confounders was miscalculated, thus overstating the statistical impact required for a variable to overcome the increased mortality. 4) Patients analyzed by Connors et al. (1) may not be characteristic of patients undergoing RHC. Eighty-four percent had either acute respiratory failure or multiorgan failure, neither of which is a specific diagnosis or entity that has a specific treatment and both of which are associated with high mortality.
None of these criticisms vitiates the possibility that there are adverse effects of RHC; however, detailed examination of the data does not provide justification for a moratorium on RHC. The Expert Consensus Committee is concerned that multicenter randomized studies are unlikely to resolve the issues raised by Connors et al. (1) because of the difficulty of standardizing protocols sufficiently to control effects of therapeutic interventions and to accommodate the very large number of patients that would have to be enrolled.
Indications and recommendations for use of bedside RHC
For specific recommendations, see Table 1.
|
RHC is helpful in the management of patients with cardiac disease and failure who are undergoing noncardiac surgery, particularly patients with decompensated heart failure undergoing intermediate or high risk operation. RHC aids in risk stratification of patients who are considered for heart transplantation. Patients with substantial fixed elevation of pulmonary vascular resistance have a poor prognosis; patients with reversible pulmonary vasoconstriction have a more favorable outcome. Preoperative RHC identifies patients with high pulmonary vascular resistance and determines its reversibility in response to vasodilator agents.
Pericardial tamponade constitutes a special case of the diagnosis and management of heart failure. Tamponade is suspected on clinical grounds and is best confirmed by echocardiography. RHC is warranted when clinical and echocardiographic findings are equivocal. In hemodynamically unstable patients with suspected pericardial tamponade, timely therapy should not be delayed by the performance of RHC.
Acute myocardial infarction
The use of RHC in acute myocardial infarction with hemodynamic instability depends on the underlying cause and the course of hemodynamic compromise. In patients with either transient hypotension or hypotension that responds promptly to intravascular volume expansion and/or low doses of inotropic drugs, RHC is not required. If patients do not improve rapidly, RHC is warranted to distinguish cardiogenic from other forms of shock and to guide the assessment of the short-term response to pharmacologic agents and mechanical support (intraaortic balloon pumping). Similarly, transient hypotension and clinical findings of low cardiac output in the setting or right ventricular infarction can often be managed without RHC, using cautious volume loading, small doses of inotropic drugs and, if required, temporary pacing. RHC is warranted when hemodynamic compromise persists despite initial therapy. Careful monitoring of cardiac output is important because ventricular interaction and pericardial constraint observed in right ventricular infarction alter diastolic properties of the left ventricle such that pulmonary artery occlusive pressure (PAOP) is not a reliable indicator of left ventricular preload.
In the setting of hemodynamic instability caused by acute mechanical complications after acute myocardial infarction, echocardiography with Doppler imaging is the primary procedure for the diagnosis. Acute mitral regurgitation with or without valve disruption, ventricular septal rupture and pericardial effusion or tamponade can be diagnosed rapidly by this modality. The main indications for RHC in these settings relate to short-term hemodynamic monitoring of preoperative and postoperative pharmacologic and mechanical support. RHC is helpful for monitoring the effect of therapeutic interventions on the magnitude of shunting through the ventricular septal defect.
The use of RHC for the guidance of pharmacologic therapy in patients with heart failure or pulmonary edema (including those requiring mechanical ventilation), and for prolonged monitoring of patients with persistent hemodynamic instability who are not considered for coronary revascularization or surgical repair of structural complications, represent intermediate indications for which considerable disagreement exists.
Perioperative use in cardiac surgery
Patients undergoing cardiac surgery may develop disturbances of perfusion that can be life-threatening and/or place the function of vital organs at risk. Identification of the underlying hemodynamic derangements determines therapeutic strategy. RHC facilitates both the diagnosis and management of low cardiac output states that may not always be adequately assessed by clinical markers of perfusion after cardiac surgery. Identification of abnormalities of cardiac output as well as right and left ventricular afterload or preload is particularly important to guide pharmacologic and other therapy aimed at optimizing perfusion. Finally, cardiac surgical techniques involving cardiopulmonary bypass may result in the accumulation of extravascular lung water, and RHC often serves as a useful guide for evaluating the effects of fluid management on global variables of perfusion after cardiac surgery.
Available data do not support the routine use of RHC in the perioperative period in the hemodynamically stable patient with good ventricular function. RHC is useful in patients undergoing cardiac surgery with previous evidence of ventricular dysfunction of any cause, complex coronary artery disease or valvular disease (especially when it is associated with significant coronary artery disease or pulmonary hypertension) and in patients requiring repeat cardiac surgery.
In the postoperative period, RHC is warranted for differentiation among causes of low cardiac output when clinical and echocardiographic evaluations are inconclusive or therapeutic trials of intravascular volume expansion and vasoactive agents have failed or are associated with high risk. However, reasonable differences of opinion exist as to whether RHC should be used for guidance of inotropic and/or vasopressor therapy after hemodynamic stability has been achieved in patients with significant cardiac dysfunction.
Primary pulmonary hypertension
RHC is warranted for establishing the diagnosis of precapillary (normal PAOP) pulmonary hypertension and to accurately assess its severity. Although echocardiography/Doppler imaging can be used to estimate pulmonary artery systolic pressure, measurement of PAOP, which is essential to distinguish between postcapillary and precapillary pulmonary hypertension and to determine pulmonary vascular resistance, cannot presently be accomplished without RHC. RHC is warranted before institution of long-term pharmacologic therapy of primary pulmonary hypertension, in particular when calcium channel blocking agents are used, to evaluate the hemodynamic response to vasodilators. The pattern of the acute hemodynamic response has been shown to correlate with long-term prognosis. Adverse hemodynamic responses, such as a reduction in cardiac output or systemic arterial pressure, or an increase in right atrial pressure, regarded as contraindications to long-term vasodilator therapy, can only be determined by hemodynamic monitoring. There is considerable difference of opinion as to whether RHC should be used for evaluation of long-term efficacy of vasodilator therapy.
Complications
Complications related to RHC are associated with the establishment of central venous access, the catheterization procedure itself and catheter residence.
Adverse events related to central venous cannulation include arterial puncture, bleeding at the site of insertion, nerve injury, pneumothorax and air embolism. The primary complications related to the RHC procedure itself are arrhythmias, which are usually clinically insignificant. Sustained ventricular arrhythmias are uncommon and occur primarily in patients with myocardial ischemia or infarction or preexisting ventricular arrhythmias. The most serious complications of RHC are related to catheter residence. They include pulmonary artery rupture, thrombophlebitis, venous or intracardiac thrombus formation, pulmonary infarction and endocarditis. The risk of thrombotic and infectious complications increases significantly when the catheter remains in place >3 to 4 days.
Alternative or complementary procedures
Echocardiography with Doppler imaging is the most important alternative or complementary procedure to RHC in the critically ill cardiac patient. It elucidates the likely mechanism of hypotension and shock, aiding the distinction between cardiogenic and noncardiogenic shock. Echocardiography/Doppler imaging has a primary role in the diagnosis of mechanical complications after acute myocardial infarction, including acute mitral regurgitation with and without valve disruption, ventricular septal defect and ventricular free wall rupture. Transesophageal echocardiography (TEE) is superior to transthoracic echocardiography (TTE) in certain patients, facilitating data acquisition and enhancing diagnostic information. Estimation of left ventricular filling pressure and stroke volume for hemodynamic monitoring has not yet been validated for widespread clinical use as an alternative to RHC. These techniques remain the subject of continued clinical investigation. TEE plays an important role in the perioperative period for evaluation of left ventricular function and results of valve repair. It is of particular value in the patient with hypotension and low cardiac output during separation from cardiopulmonary bypass. A variety of promising developments, including miniaturized probes and techniques for real time on-line monitoring of ventricular volumes, ejection fraction and compliance, are under clinical investigation.
Proposals for randomized trials
The Expert Consensus Committee considered carefully the role of future research for the evaluation of the safety and efficacy of RHC in patients with cardiac disease. It identified two areas, namely heart failure and cardiac surgery, in which RHC is frequently used and randomized trials may be considered. The Committee came to the following conclusions:
1. Randomized clinical trials in patients with heart failure are ethical in those subjects (such as patients with severe heart failure who require titration of diuretic, vasodilator and inotropic therapy) for which reasonable differences of opinion exist regarding the indication for RHC. For considerations of sample size and power, end points other than mortality would be necessary and should include morbidity, length of intensive care unit (ICU) and total hospital stay and cost. For the benefit of RHC to be assessed, it must be deployed in clinical conditions for which beneficial treatment is available.
2. The efficacy and safety of RHC in low risk patients undergoing cardiac surgery could be established by a clinical trial, but such a trial would be difficult to perform. RHC would be expected to have minimal effect on mortality and morbidity, and the required sample size would be very large. In the absence of sufficient power for examining a clinical end point, a cost-effectiveness study would not be feasible. In the case of an equivalent outcome whether or not RHC was used, a pure cost study (cost minimization study) would be possible. A clinical trial in high risk patients undergoing cardiac surgery is more likely to be revealing because the incidence of end points will be higher, and the potential of RHC to affect outcome would appear to be greater than in a trial in low risk patients. However, the Expert Consensus Committee is concerned that such a trial would be difficult to perform because of unwillingness of physicians to randomize patients.
Conclusions
The Expert Consensus Committee was convened to formulate recommendations for the use of bedside RHC in patients with known or suspected cardiac disease. Because few data are available from well controlled clinical trials, the recommendations are based primarily on inferences from published reports and expert opinion. The role of RHC is currently in transition because of the availability of noninvasive diagnostic techniques, in particular, echocardiography with Doppler, which in some cases complement and in others are alternatives to RHC. RHC has a primary role in the management of heart failure in certain patients in whom the diagnosis of heart failure is uncertain, heart failure is of sufficient severity to require intensive pharmacologic management or heart transplantation is contemplated. In patients with acute myocardial infarction, echocardiography has a primary role in the diagnosis of mechanical complications, and RHC is helpful in certain instances for assessment of the severity of hemodynamic compromise and the response to therapy. The Committee does not recommend the routine use of RHC in patients undergoing uncomplicated cardiac surgery who have good ventricular function and hemodynamic stability.
The Committee is not optimistic about the feasibility of randomized clinical trials to establish the safety and efficacy of RHC, with the possible exception of a trial in patients with refractory decompensated heart failure. The Committee questions whether major resources should be devoted to reevaluation of the role of RHC or whether such resources would be better directed toward future development and evaluation of semi-invasive and noninvasive diagnostic techniques.
| Present use of bedside right heart catheterization in patients with cardiac disease |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
A recent case-control study (1) reported excess mortality and length of stay in a diverse group of cardiac and noncardiac patients who had undergone RHC, and an accompanying editorial called for a moratorium on RHC (2). In response to these publications, the Technology and Practice Executive Committee (TPEC) appointed this Expert Consensus Committee and charged it with two related tasks: 1) to provide peer comments on the study by Connors et al. (1); and 2) to reassess the role of RHC in patients with cardiac disease, providing recommendations for current use and future research. The Consensus Committee included a variety of experts, encompassing invasive and noninvasive cardiology, critical care medicine, cardiac anesthesia, epidemiology and biostatistics. The recommendations of the Committee are based on published data and, in the absence of conclusive data, on a consensus among its members. In accordance with ACC policy, the recommendations are prefaced by background information regarding the various clinical situations considered by the Committee.
Several societies have established guidelines for the use of RHC, including recommendations for quality control (510). Each of the guidelines recognized the paucity of objective information, particularly outcome data, derived from randomized trials. Although prospective, randomized trials do form a basis for clinical practice, any valid study of RHC requires a definable measure of competency in both the acquisition and interpretation of data from RHC. There is considerable variation among physicians (11,12) and nurses (13) in the quality of acquisition and interpretation of the data obtained by RHC and subsequent integration into therapeutic strategies. There is also considerable variation in the experience of physicians caring for critically ill patients with cardiac disease and in the perceived need for RHC. These differences create certain difficulties regarding recommendations for the use of RHC as well as for the planning of multicenter trials. Programs for education of physicians and other health professionals in the use of RHC are of considerable importance.
The design of future research on the role of bedside RHC should consider alternative monitoring options that are less invasive. Echocardiographic acquisition of anatomic and physiologic data (14), Doppler estimation of cardiac pressures (15,16) and continuous TEE in the critical care setting (17,18) are presently under intense clinical investigation and may evolve to promising tools for the provision of comprehensive information. Careful consideration is required to determine whether resources should be directed mainly toward the reevaluation of bedside RHC or should be directed preferentially toward the development and investigation of newer noninvasive or less invasive monitoring modalities.
| Evaluation of a case-control study of RHC by Connors et al |
|---|
|
|
|---|
The right heart catheter is a monitoring device and as such is a diagnostic rather than a treatment modality. In life-threatening settings, the overriding benefit of a diagnostic or monitoring device is the intervention that it triggers. In this respect, the adverse impact of a diagnostic intervention, including the potential morbidity and, in rare instances, mortality stemming from the use of the device itself, cannot be evaluated except to the extent that therapeutic interventions are controlled. For example, request of computed tomographic scans for patients with closed head trauma may well select a population at greater risk, but it does not imply a higher mortality for them because they undergo risks of transport and positioning for imaging. The objective documentation in the report of Connors et al. (1) therefore does not allow for clear-cut separation of any adverse effect of the diagnostic intervention, represented by RHC, and the severity of the disorder together with the effects of a diversity of therapeutic interventions. The observations reported by Connors et al. (1) are alternatively explained by selection of more serious disorders and unsuccessful treatment strategies (19).
Unexplained clinical factors that prompt clinicians to perform RHC may contribute significantly to the risk of mortality and increased resource consumption. The decision to perform RHC is itself a marker that identifies severely ill patients. We recognize the attempts of Connors et al. (1) to eliminate what they term treatment selection bias. Yet, if this diagnostic intervention selects patients for whom there is greater likelihood of morbidity and mortality, we cannot conceive of any method, with the possible exception of randomization, by which such selection might be eliminated. Because the study by Connors et al. (1) was not randomized, comparisons between patients who did and those who did not receive RHC depended on a propensity score intended to predict whether RHC would be utilized. Among the clinical variables included in the computation of the propensity, responses to therapy were excluded. To emphasize this point, the clinical features of two patients presenting with pulmonary edema might be considered. The patient responding to initial therapy with diuretic drugs would probably not become a candidate for RHC, whereas the patient in whom more extensive pharmacologic therapy failed would be more likely to undergo RHC, thus inevitably facing a higher mortality risk (20,21). Yet the propensity score as presented would fail to distinguish between these patients.
The effect of unmeasured confounders was miscalculated. Connors et al. (1) estimated that to account for the 20% increased risk of death in the RHC group, a missing covariate would have to increase the risk of death and the probability of RHC threefold. To change the 20% increased risk to a 20% decreased risk, a missing covariate would have to increase the risk of death and the probability of RHC sixfold. In calculating these estimates, Connors et al. (1) have used their estimated odds ratios as if they were probability ratios. An increase in the probability of RHC increases the odds ratio more than it increases the probability ratio. Thus, the effect of unmeasured confounders needed to reduce the observed association was overestimated. The effect that would change the observed 20% increased risk in the RHC groups to no increased risk is actually only a twofold increase in propensity and mortality. A threefold increase in the probability of RHC and the risk of death would change the relative risk to a 20% decrease in risk (22). Thus, an unmeasured confounder with only a moderately independent effect could account completely for the observed association.
The development of the propensity score and the sensitivity analysis apply primarily to patients with acute respiratory and multiple organ system failure. Patients with acute respiratory and multiple organ system failure account for 80% of the patients analyzed by Connors et al. (1). No patients with cardiac diagnoses, possibly the largest group undergoing RHC in clinical practice, were included in these groups. In the 11% of patients who had congestive heart failure (CHF), the relative odds of death was 1.02, indicating neither harm nor benefit. The addition of an unmeasured covariate with only a modest effect on propensity and mortality could change the estimated odds in this group to significant benefit for RHC. Possible benefit for other patient subgroups included in the largest groups may also have been overlooked.
None of these specific criticisms vitiates the possibility that there are adverse effects of RHC. However, a detailed examination of the data does not provide justification for a moratorium on RHC. More objective information is needed before any such drastic action can be recommended (9,11). The Expert Consensus Committee is concerned that multicenter, randomized studies are unlikely to resolve these issues because of the difficulty of standardizing protocols sufficiently to control effects of therapeutic interventions and to accommodate the very large number of patients that would have to be enrolled. Because critically ill patients are a heterogeneous group, subgroups would have to be defined. The challenge of achieving adequate statistical power in such subgroups is likely to be very great. Furthermore and perhaps more fundamentally, the use of RHC is not in equipoise (23,24).
Although the Committee cautions against complacency, it believes that it is crucial, in the absence of scientifically secured data, to maintain objectivity and secure appropriate consensus with respect to the questions posed by Connors et al. (1).
| Acquisition and interpretation of data |
|---|
|
|
|---|
RHC must be performed with strict sterile technique. If the catheter is placed through a protective sleeve with meticulous attention to maintaining its sterility, the catheter may be advanced, if necessary, within 24 h of placement. After 24 h, the catheter should preferably not be advanced. To minimize the risk of infection, the catheter should be left in place for only as long as it provides information essential for patient management. In general, this period should not exceed 3 days; when catheters are left in place for >3 days, justification should be provided. The strategy of routine replacement of catheters over guide wires or through repeat venipuncture has not been demonstrated to reduce the risk of infection (25).
When the catheter tip is in the pulmonary artery, the balloon should be inflated slowly, using tactile assessment of balloon pressure and intravascular pressure waveform monitoring; this will minimize the risk of pulmonary artery rupture and associated hemorrhage. If concordance between pulmonary artery occlusive pressure (PAOP) and pulmonary artery diastolic pressure is established, frequent measurements of the former may not be necessary.
In most cases, RHC through the internal jugular, subclavian or antecubital approach may be performed at the bedside, without fluoroscopy. Fluoroscopic guidance should be considered in the presence of a temporary or recently placed permanent pacemaker or implantable cardioverter-defibrillator; right atrial and/or ventricular dilation, severe tricuspid regurgitation; or left bundle branch block. In the latter instance, fluoroscopy offers the advantages of minimizing catheter manipulation in the right heart, lessening the likelihood of concurrent right bundle branch block leading to complete block, and of facilitating the rapid positioning of a temporary pacemaker, should complete heart block nevertheless occur. Alternatively, the availability of standby transcutaneous pacing allows for RHC in the presence of left bundle branch block without the use of fluoroscopy.
RHC allows for measurement of 1) central venous or right atrial pressure; 2) pulmonary artery systolic, diastolic and mean pressures; 3) PAOP or "wedge" pressure; 4) thermodilution cardiac output; and 5) oxygen saturation.
Current practice is to measure right heart pressures relative to zero pressure defined at the midaxillary line. It has been suggested that pressure is more correctly referenced from the upper border of the left ventricle (26) and that zero reference pressure is best estimated in a plane 5 cm below the sternal angle (27). Meticulous flushing of catheters, transducer devices and tubing to remove blood and air reduces the risk or errors in measurement caused by damping of pressure tracings.
According to principles of cardiopulmonary physiology, current ICU practice is to record PAOP and other pressures at end-exhalation. This practice differs from that of cardiac catheterization laboratories (and in a large body of published reports on which much of our understanding of the clinical significance of pressure measurements is based), which is to record mean pressures averaged throughout the respiratory cycle. These disparate practices create a discrepancy between pressures measured in the ICU and those measured in the cardiac catheterization laboratory. In a patient breathing without the aid of positive pressure, end-exhalatory pressure is higher than mean pressure; in a patient receiving positive pressure ventilation, end-exhalatory pressure is lower than mean pressure. Until the practices of ICUs and cardiac catheterization laboratories converge, the clinician must be aware of the differences in measurement technique and must interpret the data accordingly.
For the PAOP to reflect left atrial pressure accurately, there must be a patent fluid column between the catheter tip and the left atrium. If alveolar pressure exceeds hydrostatic pressure in the pulmonary capillaries, they collapse; in this instance, the measured pressure may be grossly misleading. Thus, the pressure recorded when a pulmonary artery branch is occluded by a balloon (PAOP) reflects pulmonary venous (and thus left atrial) pressure only when pulmonary venous (Pv) and pulmonary artery (Pa) pressures exceed pulmonary alveolar (PA) pressure (i.e., Pa > Pv > PA, defined as West zone 3 [28]). In other regions of the lung (West zones 1 and 2), balloon inflation results in measurement of alveolar pressure (29). Fortunately, with the patient supine, most of the lung consists of zone 3 units (30).
In the presence of positive end-expiratory pressure (PEEP), alveolar pressure is higher, so that less of the lung consists of zone 3 units, and the likelihood of catheter placement in zones 1 or 2 increases. Inaccurate estimation of left atrial pressure is likely when the catheter is above the level of the left atrium (31) or when the catheter is at the level of the left atrium but the left atrial pressure is low. In one study (32), 43% of catheters placed through the internal jugular approach lodged at or above the level of the left atrium, yielding inaccurate assessment of left atrial pressure in the presence of PEEP. Even in the absence of an increase in alveolar pressure to a level that exceeds pulmonary venous pressure, the use of PEEP may affect the measurement of intravascular pressures because the positive airway pressure may be transmitted to the central vessels. The problem is particularly significant when >10 cm H2O of PEEP is used (33,34). Although a formula for correction has been offered (35), the extent to which an increase in airway pressure is transmitted to the pulmonary vasculature is variable and not accurately predictable.
Although PAOP is often used as an estimate of both pulmonary capillary pressure, the driving force for shifting fluid from pulmonary capillaries into the interstitium and alveoli, and left ventricular diastolic pressure, it is an imperfect measure of both. Although the pulmonary capillary pressure exceeds PAOP by only a few mm Hg in the normal lung, it may exceed PAOP by 10 to 15 mm Hg in sepsis and other inflammatory disorders, resulting in pulmonary edema despite an "acceptable" PAOP (36). PAOP reflects left atrial pressure, which is in turn indicative of left ventricular diastolic pressure only in the absence of mitral stenosis or more than mild mitral regurgitation. Furthermore, the relation between left ventricular pressure and volume in diastole depends critically on ventricular compliance, which is often abnormal in critically ill patients.
The concept of an "optimal" PAOP for maximizing left ventricular output was introduced along with the pulmonary artery catheter. Traditional values for "optimal" PAOP, typically 14 to 18 mm Hg, were based on early data from patients with acute myocardial infarction (37). Since then there have been changes both in the therapy of myocardial infarction and in the techniques for measuring PAOP. Despite these changes, the concept of optimal PAOP has remained and has been extrapolated to other patient groups. However, it has been shown (38) that effective vasodilator treatment of patients with heart failure results in higher output and lower PAOP, with no "lower limit" or optimal PAOP. Optimal PAOP varies greatly among patients and requires an empiric approach that assesses clinical status and cardiac output at various filling pressures in an individual patient. In assessing the effects of therapeutic interventions, directional changes in PAOP may be more useful than absolute values of PAOP.
Pressures and thermodilution cardiac output measurements should be interpreted, when appropriate, together with measurements of pulmonary artery oxygen saturation and arteriovenous oxygen difference. Thermodilution cardiac output may be inaccurate in the presence of arrhythmias, tricuspid regurgitation (39) and intracardiac shunting. Data obtained from RHC should be interpreted in the context of clinical assessment, taking into account weight, input/output tallies, mental status, skin temperature, lung examination, urine output, blood urea nitrogen, creatinine, arterial PO2 and pH and chest radiographic and, when appropriate, echocardiographic findings.
Technology of right heart catheters. The routinely used balloon flotation catheters have four lumens, including two for transmission of pressure signals from the pulmonary artery and the right atrium, one for balloon inflation and one for a thermostat located near the catheter tip. Multipurpose electrode catheters incorporate atrial and ventricular electrodes for recording of intraatrial and intraventricular electrocardiograms, facilitating the diagnosis of complex arrhythmias (40) and, in rare instances, temporary pacing. Newer catheters have a fifth lumen containing fiberoptic bundles for measurement of mixed venous oxygen saturation. Balloon flotation catheters with the additional capability of measuring right ventricular stroke volume, end-systolic and end-diastolic volumes and ejection fraction have been introduced (41). Continuous determination of cardiac output is feasible by the thermodilution principle with the use of a specially designed cardiac output computer without requiring delivery of indicator solution (42). Alternatively, Doppler techniques may be used for determination of continuous cardiac output (43). Heparin-bond catheters designed to decrease catheter thrombosis and catheters impregnated with antiseptic agents are also available.
| Method of data collection |
|---|
|
|
|---|
Grading of recommendations. The recommendations were graded according to the guidelines for Expert Consensus Documents provided by the TPEC of the ACC, as follows:
| Indications and recommendations for use of bedside RHC |
|---|
|
|
|---|
Accuracy in distinguishing between cardiogenic and noncardiogenic shock and between hemodynamic and permeability pulmonary edema is critical because the management strategies for the two conditions are quite distinct. The information provided by RHC often results in a change in therapy (47,48).
Pericardial tamponade
Patients with suspected pericardial tamponade constitute a special case of the diagnosis of heart failure. Tamponade is suspected on clinical grounds (tachycardia, pulsus paradoxus and jugular venous distention) and is best confirmed by echocardiography (pericardial effusion with Doppler echocardiographic signs of tamponade). In some cases, echocardiography may be unavailable or technically suboptimal. The findings of "equalization" of right atrial, right ventricular diastolic, pulmonary artery diastolic pressures and PAOP and of absence or blunting of the y descent in the right atrial tracing by RHC support the diagnosis of pericardial tamponade. Equalization of pressures may occur in other conditions, such as right ventricular infarction and pericardial constriction, and may be absent in pericardial tamponade, for example, if there is an independent cause of elevation of left atrial pressure. In selected patients, RHC may be useful for documenting the hemodynamic response (fall in filling pressures, increase in cardiac output) that accompanies removal of pericardial fluid and for detecting persistently abnormal hemodynamic variables (elevated right atrial pressure with prominent x and y descents) in patients with effusiveconstrictive pericardial disease (49). In hemodynamically unstable patients with suspected pericardial tamponade, timely therapy should not be delayed by the performance of RHC.
Management of severe heart failure
Invasive hemodynamic measurements may be useful for effectively and efficiently titrating the dosages of diuretic, vasodilator and inotropic medications, particularly during acute exacerbations of chronic heart failure or hemodynamic instability (50). Stevenson et al. (51) have introduced the concept of "tailored therapy" for heart failure. In their initial report, 50 patients who were candidates for urgent heart transplantation underwent RHC. Nitroprusside infusion and intermittent intravenous furosemide were administered in an attempt to achieve a PAOP of 15 to 20 mm Hg and a systemic vascular resistance
1,200 dynes·s/cm5 while maintaining systolic blood pressure
80 mm Hg. Oral vasodilator agents were subsequently added as intravenous medications were discontinued. Of the 50 patients, 40 were discharged without transplantation (51). In 152 patients undergoing such tailored therapy, mortality was predicted by failure of PAOP to improve during therapy (p = 0.005) (52). The authors acknowledged that it is not known whether the achievement of low filling pressures improved survival or merely identified those patients with potential for survival (52). A study of survivors demonstrated continued hemodynamic efficacy of therapy after a mean of 8 months of follow-up (53).
Management of patients with heart failure undergoing noncardiac surgery
The special case of management of patients with heart failure undergoing noncardiac surgery was addressed in the 1996 report of the ACC/AHA Task Force on Practice Guidelines (Committee on Perioperative Cardiovascular Evaluation for Noncardiac Surgery) (3). The Committee suggested that "patients with signs and symptoms of heart failure preoperatively, who have a very high (35%) postoperative incidence of heart failure, may benefit from invasive hemodynamic monitoring." It recommended that "preoperative preparation in an intensive care unit may benefit certain high risk patients, particularly those with decompensated CHF" and those undergoing high risk (emergency, prolonged or certain vascular) operations. The introduction of TEE has decreased the intraoperative use of RHC in some institutions. However, the hemodynamic information obtained from RHC is complementary to that derived from TEE, and RHC provides serial data to guide management in the postoperative period, when information from TEE is less readily available.
Evaluation of patients for heart transplantation
Evaluation for heart transplantation constitutes another special case of the management of patients with heart failure. The outcome of heart transplantation is poor in patients with substantial fixed preoperative elevation of pulmonary vascular resistance; such patients have a high incidence of postoperative right heart failure (54). Patients with reversible pulmonary vasoconstriction have a more favorable outcome after heart transplantation (55). Preoperative RHC identifies patients with high pulmonary vascular resistance and determines reversibility in response to vasodilator agents, such as oxygen, nitroprusside and nitric oxide (56). This assessment may be carried out in the ICU or cardiac catheterization laboratory.
Expert opinion of the Committee
Some guidelines for RHC in patients with heart failure were provided by the 1995 report of the ACC/AHA Task Force on Practice Guidelines (Committee on Evaluation and Management of Heart Failure) (57). In some patients with dyspnea and/or pulmonary edema, a trial of diuretic and/or vasodilator therapy may be attempted without previous RHC; in others (e.g., some patients with normal left ventricular ejection fraction and suspected diastolic dysfunction), a strategy that would reduce preload may have deleterious consequences (58). Similarly, in some patients with shock, a therapeutic trial of volume infusion may be indicated; in others (especially those with respiratory distress who have not undergone endotracheal intubation and mechanical ventilation), such empiric therapy may also be associated with substantial risk. Right heart catheterization is warranted when these initial strategies are contraindicated or fail or when there are coexisting manifestations of "forward" (low output with hypotension, oliguria and/or azotemia) and "backward" (pulmonary venous hypertension with dyspnea and/or hypoxemia) heart failure. Similarly, RHC is indicated to determine the hemodynamic contribution to respiratory failure in patients with concurrent pulmonary and cardiac disease. In contrast, RHC is not warranted for the routine management of pulmonary edema, even if endotracheal intubation and mechanical ventilation have been necessary, or for the institution or titration of diuretic and vasodilator therapy in patients with mild or moderate heart failure.
RHC may be helpful to guide inotropic and pressor therapy in patients with severe decompensated heart failure and cardiogenic shock, respectively. It is warranted in the special cases of some patients with suspected pericardial tamponade, patients with decompensated heart failure undergoing noncardiac surgery and patients undergoing evaluation for heart transplantation. It should not be performed before pericardiocentesis in patients with hemodynamic instability when the diagnosis of pericardial tamponade is certain or probable. The use of RHC to guide diuretic and vasodilator therapy in patients with heart failure constitutes an intermediate indication over which there is substantial disagreement. Similarly, there is disagreement regarding the need for RHC in patients with compensated heart failure undergoing noncardiac surgery; use of RHC should be reserved for patients undergoing high risk operations.
Recommendations: heart failure.
Conditions in which there is general agreement that RHC is warranted
Conditions in which reasonable differences of opinion exist regarding RHC
Conditions in which RHC is not warranted
Conditions in which a relative contraindication to RHC exists
Conditions in which an absolute contraindication to RHC exists
Acute myocardial infarction.
Background
The routine use of RHC in uncomplicated acute myocardial infarction has never been recommended. Gore et al. (20) and Yarzebski et al. (59) examined the use of RHC over time in patients with acute myocardial infarction in Worcester, Massachusetts. They found an increase in the use of RHC from 1975 to 1984, followed by a progressive decline until 1990, the last year examined. Even before restrictions imposed by cost containment, managed care and published guidelines, the cardiology community had already reduced the use of RHC in acute myocardial infarction on the basis of clinical judgment and experience.
Indications for the use of RHC are related to the diagnosis and management of specific complications of myocardial infarction. These include 1) hypotension, low cardiac output and cardiogenic shock as a consequence of predominant left ventricular failure; 2) acute mechanical complications (mitral regurgitation from papillary muscle rupture or ischemia, ventricular septal rupture or ventricular free wall rupture); 3) complicated right ventricular infarction; and 4) heart failure or pulmonary edema unresponsive to routine management.
The role of RHC has changed because of the availability of noninvasive diagnostic techniques, in particular, echocardiography. Echocardiography with Doppler imaging is the procedure of choice for the evaluation of left and right ventricular function and the diagnosis of complications of acute myocardial infarction (see later ["Alternative or Complementary Procedures"]). The major role of RHC is related to guiding the management of complications, such as therapy with vasoactive agents, intraaortic balloon pumping (IABP) and other interventions. If echocardiography with Doppler imaging is not available, RHC remains an important modality for the diagnosis of complications of myocardial infarction.
Hypotension, low cardiac output state and cardiogenic shock
RHC is most commonly used in the setting of acute myocardial infarction to guide the management of low cardiac output, hypotension and cardiogenic shock. Published guidelines consistently recommend the use of RHC in cardiogenic shock (57,10). Although the diagnosis of shock is dependent on the presence of hypotension and clinical signs of organ hypoperfusion (oliguria, cold skin, depressed mental status) (60), RHC is recommended to confirm the presence of shock by hemodynamic criteria and to assess filling pressures, thereby distinguishing cardiogenic from other forms of shock, in particular, hypovolemic shock. However, there are no randomized studies that address the benefit of RHC in this patient population. A nonrandomized study by Gore et al. (20) demonstrated excess mortality in all patients with acute myocardial infarction treated with RHC and no benefit in the subset of patients with cardiogenic shock. In that retrospective study, physician selection of the sicker patients to undergo RHC was likely. Similarly, Zion et al. (21) observed increased mortality in all patients with acute myocardial infarction managed with RHC but no difference in patients with cardiogenic shock. The study of Blumberg et al. (61), although not limited to patients with hypotension or cardiogenic shock, reported increased mortality in Medicare patients with acute myocardial infarction treated in hospitals with a higher rate of RHC.
The lack of observed benefit of RHC in the setting of cardiogenic shock may be in part related to the poor prognosis of these patients and the relative lack of effective treatment options, such as vasoactive agents, IABP (62) and thrombolytic agents (63) at the time these studies were carried out. Experimental and clinical studies suggest that aggressive use of IABP, often combined with thrombolysis, may improve prognosis (64,65). More definitive forms of coronary revascularization with percutaneous transluminal coronary angioplasty (PTCA) (6668) or CABG (60,6971) have revealed some encouraging results. The importance of the form of revascularization and the timing of the procedure in patients in cardiogenic shock are currently being studied by the international Should We Emergently Revascularize Occluded Coronaries for Cardiogenic Shock (SHOCK) trial (60). Although a preliminary report from the registry of the SHOCK study (71) revealed a reduced mortality associated with the use of RHC in patients with suspected cardiogenic shock, there was substantial selection bias in that patients receiving RHC were younger and were also more likely to receive more aggressive management with IABP, cardiac catheterization, PTCA and CABG.
Acute mechanical complications
Mitral regurgitation secondary to papillary muscle rupture
The presence of a flail mitral valve leaflet may be suspected on the basis of clinical presentation and physical examination, although in severe cases characterized by low cardiac output, the systolic murmur may not be heard because of rapid equalization of left ventricular and left atrial pressures (72,73). Acute mitral regurgitation can be accurately diagnosed by either transthoracic echocardiography or TEE; the latter is superior for detection of papillary muscle rupture (74). Severe mitral regurgitation associated with papillary muscle rupture usually prompts immediate cardiac catheterization (including RHC) to evaluate coronary anatomy before emergent surgical intervention. Bedside RHC is useful for guiding vasodilator, inotropic and intraaortic balloon therapy (75). On the basis of available published reports and experience, short-term RHC is useful for guiding therapy in patients who are not surgical candidates because of comorbidity or other reasons.
Mitral regurgitation secondary to papillary muscle ischemia
Papillary muscle dysfunction without mechanical disruption of the valve leads to mitral regurgitation, which may be transient. The clinical diagnosis may be confirmed by echocardiography with Doppler imaging. Treatment with anti-ischemic, diuretic and vasodilator agents can reduce ventricular size and alter ventricular geometry (76) and may diminish or eliminate mitral regurgitation. Bedside RHC is not required for diagnosis and may in fact be misleading. Although the presence of a "regurgitant" v wave on the PAOP tracing supports the diagnosis of mitral regurgitation, the height of the v wave often does not give an accurate assessment of the severity of the mitral regurgitation, and large v waves are not diagnostic of mitral regurgitation (77,78). Short-term hemodynamic monitoring is helpful in assessing therapeutic interventions, but the utility of RHC over clinical and echocardiographic assessment has not been studied in this setting.
Ventricular septal rupture
Echocardiography with Doppler imaging is highly sensitive and specific for the diagnosis of ventricular septal defect and provides information regarding the magnitude of the shunt, pulmonary artery pressure and right ventricular function. RHC, although useful for the diagnosis of postinfarction septal rupture (79), is in general not required. However, RHC is helpful for the assessment of the magnitude of the shunt (when not measurable by echocardiography), evaluation of the response of the shunt to pharmacologic and IABP therapy and monitoring of patients in the postoperative period.
Ventricular free wall rupture
Free wall rupture may present as (1) electromechanical dissociation; (2) subacute rupture with tamponade and a low output/shock state; or (3) pseudoaneurysm with or without hemodynamic instability. Rupture with electromechanical dissociation usually results in death before any attempt at premortem diagnosis. After subacute cardiac rupture, echocardiography with Doppler imaging is the primary modality for the diagnosis of pericardial effusion and tamponade and may guide pericardiocentesis. Echocardiography may also detect the presence of a pseudoaneurysm. RHC can confirm the diagnosis of pericardial tamponade in patients with suspected subacute cardiac rupture (80), in whom clinical, electrocardiographic (81) and echocardiographic features are not diagnostic. However, RHC should not delay potentially life-saving treatment.
Right ventricular infarction
Right ventricular dysfunction is a common complication of inferior wall myocardial infarction that can affect treatment and prognosis. The diagnosis of right ventricular infarction can be made by physical examination and electrocardiography, using right precordial leads (82); it can be confirmed by echocardiography or radionuclide imaging (83). Although several hemodynamic criteria have been proposed for the diagnosis of right ventricular infarction (84), RHC is usually not required for the diagnosis. Most patients with right ventricular infarction remain hemodynamically stable, although caution in the administration of nitrates and other vasodilators is recommended to avoid reduced filling of the impaired, volume-sensitive right ventricle (85). The degree of hemodynamic compromise depends on the extent of left and right ventricular as well as right atrial involvement (86,87). Furthermore, ventricular interaction through septal shift and pericardial restraint can markedly affect the systolic and diastolic function of both ventricles.
For patients with right ventricular infarction with hypotension and signs of low cardiac output, treatment includes cautious volume loading, low doses of inotropic agents and, if required, temporary (usually atrioventricular sequential) pacing (88). RHC is not required in the majority of these patients. Reperfusion of the infarct-related artery by either thrombolysis or catheter-based intervention can lead to rapid hemodynamic improvement (89,90). The degree of volume loading needed is controversial because excessive volume may result in a decline in cardiac output (88,91). Therefore, in patients with persistent hypotension or with low cardiac output despite initial treatment, RHC is indicated with particular attention to cardiac output measurements. Because of the alteration of the diastolic properties of the left ventricle caused by ventricular interaction and pericardial restraint, PAOP is not an accurate indicator of left ventricular preload (86). Careful monitoring of the effect of interventions on cardiac output is necessary.
Heart failure
RHC has not