The &accuracy" of the pulmonary artery catheter (PAC) in this summary will be defined as the true representation of pressures and waveforms in the heart and pulmonary artery, and the correct interpretation of their meaning. Cardiac output is discussed in a separate summary. Accuracy of the PAC is a function of proper set-up (de-airing of all lumens; correct levelling, zeroing, and calibration procedures for the transducer), appropriate waveform identification, and physiologically correct waveform interpretation.
In setting up a PA catheter, the most likely source of introduced error is incorrect levelling of the transducer (the air-fluid interface) to the left atrium. Damping of the waveform due to the presence of air bubbles is visually detected; zeroing and calibration are usually accomplished by the push of a button, and may be verified by testing devices or a mercury manometer. The only method to detect incorrect transducer level is direct observation; this must be performed frequently if the transducer is fixed while the patient is moved. A more difficult-to-detect error is inadequate "dynamic response". This refers to the appropriate matching of damping to the natural frequency of the catheter transducer system (1). "Overdamping" or "underdamping" produce characteristic artifact on the pressure tracings and are more likely to affect systolic pressure than diastolic or mean pressures.
While clinicians may be confident they can identify waveforms from the PAC, a recent survey suggested that many can not. Forty seven per cent of 496 physicians could not correctly identify the pulmonary artery occlusion pressure (PAOP) from a clear tracing (2). In order to correctly identify the PAOP, one must be able to determine the phase of respiration (all pressures should be read at end-expiration), and the components of the normal PAOP. This requires the ability to identify the "a" and "v" waves which occur in the left atrium. Either hard copy (i.e. printout of the waveform) or a monitor screen which can be "frozen" is required for adequate visualization. As the PAOP is an estimate of left ventricular end diastolic pressure (LVEDP), the peak "a" wave height is probably most reflective of this pressure.
There can be little doubt that misinterpretation of pressures or waveforms from the PAC occurs commonly; this subject has been covered in extensive reviews (3,4). For example, while the PAOP waveform may be correctly identified, the pressure may not equivalent to LVEDP (table). Or, the presence of a large "v" wave on the PAOP tracing may be attributed to mitral regurgitation when in reality there is simply a noncompliant left atrium (5). Although clinicians use the PAOP to estimate preload, true ventricular preload is fiber stretch before shortening, or end diastolic volume. Acute reductions in ventricular compliance (e.g. myocardial ischemia) may cause acute increases in LVEDP (and PAOP), but the preload may actually decrease. In cardiac surgery patients LV pressures and volumes correlated poorly both before and after bypass (6).
Some clinicians believe that combinations of factors yield more information than the factors themselves. Examples include calculated vascular resistances, and "work" indices (e.g. LV stroke work index). The vascular resistances are computed by dividing the pressure decrease across a circuit (e.g. PA mean pressure minus PAOP) by the cardiac output. A major problem with this concept is that very small individuals (i.e. with lower cardiac output) will have very high resistances. A problem with the concept of pulmonary "resistance" calculated in this way is that the pulmonary circulation is "recruitable": vascular beds may be opened or closed when flow and/or pressure are changed.
The above discussion should lead to the conclusion that with the very many potential areas for inaccuracy or misintepretation, a major concern in PA catheter safety should be mismanagement. While there are no scientific studies of this phenomenon, the study referred to above (2) showed that 55% of the 496 respondents scored less than 70% correct answers to questions regarding interpretation of PAC data, and 43% of the respondents scored less than 70% correct answers to questions about treatment based on PAC data. Some of the reports that have failed to detect a benefit from the use of the PAC may simply be documenting this phenomenon (7). Journals representing the Canadian and American Medical Associations, and the European Society of Intensive Care Medicine have recently published reviews recommending some combination of restricted availability, the development of practice guidelines, utilization management, as well as research directed at identifying patient groups likely to benefit from use of the PAC (8,9,10).
As an invasive procedure, PA catheter insertion has been associated with a host of mechanical complications. Any structure within needle range of the various approaches has been entered, catheters have perforated vessels and cardiac structures, and have knotted with other invasive lines or wires. Most cardiac anesthesiologists will have had the experience of a PAC being inadvertently sewn to the heart at the time of surgery, requiring a return to the operating room for removal (this may be avoided by verifying free movement of a PAC before chest closure). Thrombosis around even heparin bonded PAC's occurs commonly, and may progress to major vein thrombosis; infections may occur at the site of insertion, or may progress to sepsis or endocarditis. Recent studies have shed doubt on the widely practised "scheduled" replacement of PACs (e.g. every 3 days) as a means to reduce infection; it seems the main result of this practice is to increase the incidence of mechanical complications due to insertion (11,12). Pulmonary infarction may occur distal to a permanently wedged PAC, and massive hemoptysis as a result of PA rupture during balloon inflation may also occur. Fortunately most serious complications occur infrequently, but this fact is of little comfort when the patient is your own.
Table: Situations where mean PAOP may misrepresent LVEDP PAOP > LVEDP PAOP < LVEDP Positive Pressure Ventilation +/- PEEP Noncompliant left ventricle Increased intrathoracic pressure Aortic regurgitation PAC not in West zone III Reduced area of pulmonary vessels (e.g. embolus) Obstructive airways disease Tachycardia Increased pulmonary vascular resistance Mitral valve disease Pulmonary venous compression (from Tuman et al3)
Winterlude 94 Program
Winterlude Main Page
Association of Ottawa Anesthesiologists