In order to understand the limitations of this approach, the limitations of the measurements and of the analysis itself must also be understood.

The limitations of the PAC in establishing a picture of the haemodynamic state devolve to three areas:

The use of PAOP as a measure of preload.
The first area in which the use of PAOP as a measure of preload has limitations is in the nature of the measurement itself. For each of the steps involved in the extrapolation from LV preload to PAOP there are potential errors. The steps in this extrapolative chain can be summarised as:

LV end-diastolic fibre length - LVEDV - LVEDP - LAP - PVP - PAOP (Figure 1).

The first extrapolation that LVEDV is a reasonable measure of LV end-diastolic fibre length is generally true.

The second extrapolation makes the assumption that LVEDP is a valid measure of LV end-diastolic volume. Provided that LV compliance (stiffness) does not change acutely, this is a reasonable assumption. However, we now know that reductions in ventricular compliance are an early hallmark of ventricular ischaemia and that therefore under the conditions when the PAC is most likely to be needed, the central assumption is incorrect. This is not, in fact, as important as it appears, as it may often be more appropriate to transfuse a patient to a particular filling pressure rather than to a particular end-diastolic volume.

In practical terms, transoesophageal echocardiography (TOE) is the only practical tool which can identify changes in preload which occur concurrently with changes in ventricular compliance. In one study of graded blood volume reduction ¹, TOE was found to correlate more accurately with the degree of blood loss than either CVP or PAOP measurement.

Unpredictable changes in ventricular compliance can also occur during the reperfusion and cardioplegia 'washout' phases that occur in the early post-bypass period. As a result, the correlation between PAOP and LVEDV following coronary artery bypass surgery may be poor ².

The third extrapolation makes the assumption that LAP is a valid measure of LVEDP. This is not always the case. - In the presence of mitral stenosis or other causes of obstruction to left atrial outflow, LAP (and therefore PAOP) will overestimate LVEDP. Even if the mitral valve is normal, the presence of tachycardia may also lead to overestimation of LVEDP.

The presence of large 'V' waves in the wedged trace may also lead to overestimation of LVEDP if a mean, rather than correctly-timed, end-expiratory value of PAOP is used. 'V' waves can occur either as a result of mitral incompetence or in the presence of acute myocardial ischaemia - particularly if associated with papillary muscle dysfunction. As stated elsewhere in this text, the most appropriate measure of LVEDP is the peak of the 'A' wave at the end of expiration.

'V' waves can also be misinterpreted as an unwedged trace causing the operator to persist in attempting to 'wedge' the already wedged catheter with consequent PA rupture.

The fourth extrapolation makes the assumption that PVP is a valid measure of LAP. - This is generally true except in the unusual circumstance of there being some focal obstruction to pulmonary venous drainage.

The final extrapolation makes the assumption that PAOP is a valid measure of PVP. - The major caveat here is that the catheter tip be situated in a West zone III of the pulmonary vasculature. This caveat is explained in detail in the section describing the effects of intrathoracic pressure.

If, for any reason, PAOP is not available, PAEDP is commonly used as a measure of PAOP. This is often reasonable practice, but is likely to overestimate PAEDP (and therefore preload) if PVR is high, or there is a tachycardia.

The use of cardiac output as a measure of contractile function.
The second area of possible misapprehension relates to the use of cardiac output as an index of contractile function under measured conditions of preload and afterload.

1. It should be remembered that the measurements derived from a PAC provide information about global ventricular function and convey no information at all about regional myocardial conditions.

While it may be legitimate to make deductions about the likely overall myocardial energy balance from indices such as stroke work index and cardiac index, techniques such as electrocardiography, echocardiography, (particularly trans-oesophageal echocardiography) and other imaging techniques offer the clinician far more information about regional cardiac conditions.

2. Even in the absence of any concomitant change in the contractile state, cardiac output may be extremely sensitive to changes in both afterload and pulse rate (Figure 2).

The use of Systemic Vascular Resistance as a measure of afterload.
Probably the most 'correct' measure of left ventricular afterload is systemic input impedance which defines all the external forces which oppose ventricular ejection. However, in a clinical context, impedance contains components which are impossible to measure and for this reason, systemic vascular resistance - which comprises more than 90% of impedance, is generally accepted as a valid measure of afterload. Resistance has the advantages that the measurements required for its calculation are all available from a PAC and that it can then be simply derived from the equation:

SVR = (MAP - RAP * 79.9)/ CO

- where 79.9 is the constant used to convert from mm Hg / L / min to dynes.s.cm-5.

An alternative approach is to use a more chamber-related measure of afterload such as wall tension (Figure 3), but this measurement cannot be made using a PAC.

1. Cheung AT, Savino JS, Weiss SJ, Aukburg SJ, Berlin JA: Echocardiographic and hemodynamic indexes of left ventricular preload in patients with normal and abnormal ventricular function. Anesthesiology 1994;81:376-387

2. Hansen RM, Viquerat CE, Matthay MA et al Poor Correlation between Pulmonary Arterial Wedge Pressure and Left Ventricular End-Diastolic Volume after Coronary Artery Bypass Graft Surgery. Anesthesiology 1986, 64:764-770.

Last edited on: 13/11/2000