Chapter 107: Pulmonary Artery Catheterization

The PAC, also called a right heart catheter or named after its creators a Swan-Ganz catheter, was introduced in 1970.¹ Made of flexible plastic (polyvinyl chloride) with an inflatable balloon at the tip, the PAC is directed at the bedside, without the need for fluoroscopy, via a central vein, using the flow of blood through the right heart, into the pulmonary artery, guided by pressure waveforms displayed on the bedside monitor. Harry Swan developed the concept of using a flotation device while observing sailboats and initially planned to use a sail-like device, but a balloon proved to be a safer interface within blood vessels and cardiac chambers. Norman Ganz developed the thermodilution CO capability of the PAC.

Use of the PAC peaked during the 1980s and 1990s, but usage has since trailed off due to its invasive nature and subsequent randomized controlled studies demonstrating no outcome benefit (ie, reduced mortality, organ failure, length of stay, or costs).^2,3 Despite this, there remains a core of accurate and reliable hemodynamic information, not readily available from any other single device, which can be obtained from the PAC.

The lack of demonstrable improvement in patient outcome when using the PAC is multifactorial. The main factors include reliance on inaccurate data (eg, pulmonary artery occlusion pressure [PAOP]) to guide management; the fact that the PAC is a monitoring technology, not a therapeutic technology; and that the critical illnesses the PAC is used to manage such as septic shock, acute respiratory distress syndrome (ARDS), and multiple organ failure are syndromes that have no specific treatment and outcomes are not directly dependent on hemodynamic variables.

This chapter describes standard PAC anatomy, function, indications for insertion, insertion technique, complications, and obtaining reliable data while guiding bedside hemodynamic management.

A standard triple port VIP 7.5-Fr PAC is described herein. A double-lumen port PAC has identical components, other than lacking the additional infusion lumen.

The PAC insertion length (Figure 107–1) is 110 cm with an outside diameter of 2.3 mm (7.5 Fr). The catheter is yellow color and constructed of polyvinyl chloride extrusions. There are 3 main internal lumens (Figure 107–2): the distal PA lumen opens at the catheter tip, the proximal injectate lumen (also called the right atrial [RA] or central venous pressure [CVP] lumen) opens at 27 cm proximal to the catheter tip, and an infusion lumen opens adjacent to the proximal lumen, at 29 cm. In addition to the 3 main lumens, the tip of the catheter has a small 1.5-mL capacity latex balloon that can be inflated (11-mm diameter) and deflated (7-mm diameter) via an inflation lumen. A 3-mL “blood gas” syringe, modified to withdraw a maximum of 1.5 mL of air is connected to the inflation lumen. The inflation lumen has a lock that can prevent unintentional balloon inflation or air embolism in the event that both the syringe becomes disconnected and the balloon has a defect (Figure 107–3A and B). The thermistor is located about 4 cm proximal to the PAC tip. At the proximal end of the catheter, a thermistor connector, protected by a red removable cap allows connection to an electrical cable that transmits core body temperature information from the thermistor to the monitor (Figure 107–4).

The introducer kit (see later under PAC Insertion) includes a catheter sheath that is placed over the PAC prior to insertion to allow sterile manipulation of PAC depth after insertion. Some introducers can lock the PAC to prevent migration, in those that do not, a piece of tape is provided to secure the proximal end of the catheter sheath to the PAC that helps resist PAC migration.

The functional components of the PAC are outlined in Table 107–1. Direct parameters obtained from the PAC are outlined in Table 107–2.

Pressure Data

Pressure information derived from the PAC includes CVP, right ventricular (RV) systolic pressure, RV diastolic pressure, PA systolic and PA diastolic pressure and the PAOP (also called the pulmonary capillary wedge pressure [PCWP] or pulmonary artery wedge pressure [PAWP]). CVP measurement does not require a PAC, only a central venous catheter.

Before the PAC was developed, CVP was recommended as a simple measure of intravascular volume and RV filling or preload. Ventricular preload is the ventricular end-diastolic volume. Either absolute CVP or changes in CVP were proposed to differentiate decreased, normal, or increased intravascular volume or RV preload. However, there is no correlation between CVP and blood volume or cardiac preload in critically ill patients or normal volunteers.^4,5,6

CVP does not reflect left ventricular end-diastolic pressure (LVEDP) and the major reason the PAC was designed was to provide information about left heart function. The PAC was proposed to provide information about the function of the left ventricle (LV) by occluding a segment of the pulmonary artery via the balloon at the PAC tip, creating a no-flow situation in the vessel, thus the pressure at the PAC tip, the PAOP, would be equal to the pulmonary capillary pressure, pulmonary venous pressure, left atrial pressure, and the LVEDP. In both published research and at the bedside, the PAOP often became equated with LV preload. However, subsequent radionuclide and echocardiography studies have shown that the PAOP does not correlate with LV preload or intravascular volume.^7,8 Pressures do not correlate with volume due to alterations in cardiac compliance induced by positive pressure ventilation, positive end-expiratory pressure (PEEP), intra-abdominal pressure, pulmonary vascular disease, cardiac hypertrophy, ischemia, inopressors, and that the heart functions as a suction pump, that is, diastole is active and can create negative pressures within the chambers.⁹ In fact the PAOP may not even reflect the LVEDP due to pathophysiologic factors (eg, mitral valve disease, pulmonary parenchymal and vascular disease) or that PAOP is a very difficult parameter to measure accurately.^8,10 It has also been demonstrated that in stable patients on mechanical ventilation, the PAOP can vary as much as 7 mm Hg in 40% of patients.¹¹

The acronym “PCWP” perhaps best describes its role in hemodynamic monitoring: Parameter that Commonly gives Wrong information about the Patient. The most reliable pressure measurements provided by the PAC are the PA artery systolic and PA diastolic pressures.

Cardiac Output

PAC CO measurement utilizes the indicator-dilution principle using a thermal load as the indicator, called right heart thermodilution. The thermistor located near the tip of the PAC continuously monitors core body temperature in the PA. There is also an external temperature probe measuring the injectate temperature. To obtain a CO measurement, a thermal load (commonly 10 mL of room temperature fluid [saline or D₅W]) is injected rapidly via the proximal (CVP) infusion port and cools the blood as it travels through the right heart, into the PA and as this cooled blood (indicator) passes the thermistor, the temperature change over time is recorded. The monitor displays the curve of temperature over time and the area under the curve is inversely proportional to the rate of blood flow in the pulmonary artery. These data are converted by the monitor into the CO in liters per minute using the Stewart-Hamilton equation:

Where, V₁ is the injectate volume, T_B is the blood temperature, T_I is the injectate temperature, K₁ is the injectate density factor, K₂ is a computation constant accounting for the catheter dead space and the heat exchange during transit, and ∫ΔT_B(t)dt is the area under the time-temperature curve.

A single CO measurement is approximately within 30% of the actual CO. When the average of the 3 closest CO measurements are made, accuracy is within 10% to 20% of the actual CO.¹² Average at least 3 similar measurements, for example, within 10% to 15% of the median value discarding those outside this range.¹³

Right heart thermodilution CO is considered the practical gold standard for CO measurement, but there are conditions that can affect the accuracy of thermodilution CO. These include the following:

1. Tricuspid regurgitation (TR). Either falsely low (commonly) or falsely high CO as the regurgitant flow causes changes in the temperature curve due to recirculation or loss of indicator. Echocardiography can assess for TR. If there is significant TR another CO method must be used to determine the CO.

2. Rapid infusion of intravenous (IV) fluids via the introducer port and proximal infusion or VIP port can reduce the thermistor temperature leading to a smaller temperature change and a falsely low CO.

3. Intermittent SCD. Not in use at the time the PAC was introduced, but very common today, SCD—used to reduce the incidence of deep venous thrombosis (DVT)—increase the flow of relatively colder blood from the lower extremities when they inflate. If thermodilution CO is performed during inflation of the SCD, the thermistor (reading a lower temperature) will sense a smaller temperature change leading to a falsely low CO determination.¹⁴

4. Injectate volume. If less than the expected (eg, 10 mL) injectate volume is injected, a falsely high CO will occur due to a more rapid than expected temperature change.

5. Respiratory variation can change the temperature of PA blood and alter the CO.

6. Intracardiac left-to-right shunts, for example, ventricular septal defect (VSD).

7. The wrong CO computation constant setting in the monitor. The injectate temperature range and the PAC model determine the computation constant.

Thus holding rapid fluid infusion via the introducer, proximal infusion port, and VIP port of the PAC for at least 30 seconds, turning off the inflation of SCD, and timing the start of injection to the same point of the respiratory cycle, for example, end expiration, improves the accuracy of thermodilution CO. Room temperature injectate is used since ice-cooled injectate does not increase the accuracy¹³ and is more burdensome. However, in the setting of hypothermia (low T_B ie, 29°C-30°C) ice-cooled injectate should be used to improve the accuracy by increasing the temperature difference between blood and injectate.

Hemodynamic Waveforms

Hemodynamic waveforms from the PAC consist of “a,” “c,”, and “v” waves in the right atrial pressure (RAP) tracing and “a” and “v” waves in the PAOP tracing (Figure 107–5). The “a” wave due to atrial contraction occurs after the P wave of the electrocardiogram (ECG). The “c” wave in the RAP waveform is caused by tricuspid valve closure, the “c” wave caused by mitral valve closure is too small to be reflected across the pulmonary vascular bed, and hence the “c” wave is absent from the PAOP waveform. The “v” wave is caused by filling of the atria against a closed atrioventricular valve during ventricular contraction occurs after the ECG T wave. The downslope after an “a” wave is called the “x” descent and similarly for the “v” wave is called the “y” descent (not labeled in Figure 107–5).

Large (cannon) “a” waves may be seen with atrioventricular dissociation. Large “v” waves may be seen with TR or mitral valve regurgitation (MR), decreased ventricular compliance or (in the RAP waveform) with a VSD. Enlarged “v” waves may be absent in TR or MR if the right or left atrium is enlarged. An echocardiogram is more specific and sensitive for the detection of TR and MR. In cardiac tamponade early diastolic compression can lead to a prominent “x” descent. In constrictive pericarditis, late diastolic limitation of ventricular expansion can lead to a prominent “y” descent. However in practice, echocardiography is more sensitive and specific for diagnosis if these conditions are suspected.

In reports about hemodynamic waveforms in patients with these and other cardiac disorders, the waveforms are usually obtained in the cardiac catheterization laboratory or coronary care unit setting using equipment that is more sensitive than that used in the general noncoronary intensive care unit (ICU) setting. It is notable that using standard PACs and recording systems in the general ICU setting, no hemodynamic waveform components can be identified in 52% of PAOP tracings and 20% of RAP tracings.¹⁵ Increased monitor/recorder sensitivity can enhance the detection of waveform components.

Mixed Venous Blood Gas

Mixed venous oxygen and carbon dioxide levels provide further information about hemodynamics. The most accurate global venous blood gas, the mixed venous blood gas (MVBG), is obtained by drawing PA blood from the distal port of the PAC. The PA is distal to the mixing of superior vena cava (SVC), inferior vena cava (IVC) blood in the RV. The SVC carries blood returning from the head and upper extremities. Due to the relatively high cerebral metabolic rate of oxygen consumption, SVC oxygen levels are normally lower than in the IVC or mixed venous blood. Depending on tip location, a central line may preferentially sample SVC or IVC blood, thus central venous blood gases may differ from true MVBGs.

The normal mixed venous Po₂ (Pvo₂) is 40 mm Hg with a mixed venous hemoglobin-oxygen saturation (Svo₂) of 75%; it is a mixture of venous return from the SVC and IVC. The normal SVC hemoglobin-oxygen saturation (Ssvco₂) is 70% and the normal IVC hemoglobin-oxygen saturation (Sivco₂) is 80%. The other important component of the MVBG is the mixed venous carbon dioxide level (Pvco₂), which normally is within 5 mm Hg higher than the arterial carbon dioxide level obtained on a simultaneous arterial blood gas (Paco₂).

Interpretation of Svo₂

Low or decreased Svo₂ levels can result from decreased CO, increased oxygen consumption (vo₂), or anemia. A low Svo₂ value (eg, 40%) does not necessarily mean that the tissues are not getting enough oxygen or that the CO is decreased or inadequate. For example, if the CO is low and the tissues extract more oxygen per unit of blood flow, the Svo₂ will fall, but the tissues may be extracting enough oxygen to meet the metabolic needs. Anemia or increased vo₂ (fever, hypermetabolism) can reduce Svo₂ despite a normal or increased cardiac output.

High or increased Svo₂ levels can result from anatomic systemic left-to-right shunting (arteriovenous [AV] grafts, liver disease) or inability of the cells to utilize oxygen as is commonly seen in sepsis and liver dysfunction as part of the multiple organ failure syndrome—conditions that are the most common in critically ill patients in general medical and surgical ICUs.

Normal Svo₂ levels may not indicate that all is well. Normal Svo₂ may result from the mixing of venous blood returning from both high and low Svo₂ regions, as Svo₂ is a global measure.

From the above information, it should be no surprise that Svo₂ does not correlate with CO in critically ill patients. Further, looking at Svo₂ levels as an indicator of the adequacy of oxygen delivery is similar to looking at the amount of garbage your neighbor throws out to determine if they are getting enough food.

Attempts to increase CO to raise levels of oxygen delivery to arbitrary levels via inotropes or blood transfusions have failed to positively impact patient outcome and may be harmful particularly in older patients.^16,17,18

Svo₂ levels should be interpreted with regard to hemoglobin level, CO, vo₂ and lactate level. Lactate levels correlate more closely with the adequacy of tissue oxygen than Svo₂.

Oxygen Content, Delivery, and Consumption

The mixed venous oxygen can be used to calculate oxygen consumption (vo₂), which is the difference in AV oxygen content multiplied by the CO. Table 107–3, contains the formulas for oxygen content (CO₂), delivery (Do₂), and consumption (vo₂).

Fick Cardiac Output

The blood flow to an organ (eg, CO) is equal to the AV difference in the concentration of a substance across the organ multiplied by the release or uptake of that substance by the organ. Using oxygen as the substance and the lung as the organ, CO is equal to the total body oxygen uptake (consumption) as measured by indirect calorimetry divided by the AV oxygen content difference. This is an alternative method to thermodilution indicator-dilution method. The accuracy of Fick-derived CO is limited by considerable inaccuracies in total body oxygen consumption measurement and shunting.

Intracardiac Left-to-Right Shunts

The oxygen levels in the right atrium (RA), RV, and PA are normally similar, for example, Svo₂ approximately 75%. An intracardiac left-to-right shunt, for example, VSD, results in step-up in venous oxygen moving from RA to RV to PA from oxygenated LV blood directly ejected into the RV and pulmonary circulation. This can be assessed by sampling blood gases from the distal port of the PAC as it passes each level.

Interpretation of Pvco₂

Although Svo₂ has garnered the most attention to date, Pvco₂ levels correlate more closely with blood flow and CO. A consistent finding is the dissociation between venous and arterial Pco₂ (increased P[v-a]co₂ gradient) during reductions in blood flow, that is, decreased CO.^19,20,21 The normal P[v-a]co₂ gradient is less than or equal to 5. The P[v-a]co₂ gradient is considered to be increased if it is greater than or equal to 7. Increases in the P[v-a]co₂ gradient are inversely proportional to the CO. However, an increased P[v-a]co₂ gradient may also be due to an increased metabolic rate. Hypermetabolic rates are usually associated with a fever. An extreme example is malignant hyperthermia which is associated with very high fever and both a hyperdynamic and hypermetabolic state with increased CO and an increased P[v-a]co₂ gradient.

Derived Parameters

Parameters that are derived from PAC data are outlined in Table 107–3. It is important to understand that systemic vascular resistance (SVR) is not equivalent to LV afterload and the pulmonary vascular resistance (PVR) contains an unreliable parameter in its calculation—the PAOP. A more direct approach is to analyze the arterial blood pressures and the pulmonary artery pressure (PAP) in relation to the CO. See Table 107–3 for additional explanation.

The decision to insert a PAC should be based on individual clinical assessment where frequent hemodynamic assessments using information obtained from the PAC may allow more effective hemodynamic management. Inability to transport a patient for further diagnostic testing (eg, computed tomography [CT] angiogram) or inadequate echocardiographic windows may influence these decisions as well. The clinical scenarios include the following:

1. Shock (nonhemorrhagic) or hemodynamic instability unresponsive to treatment and when cardiac output determination and trending is deemed valuable.

2. Pulmonary hypertension (PHTN) assessment, trending and treatment in diseases resulting in PHTN (eg, suspected or proven pulmonary embolism, primary or secondary forms of PHTN, severe ARDS).

3. Respiratory failure including severe hypoxemia of undetermined etiology.

The PAC is not utilized as often as in the past. It is especially important to have a physician skilled in PAC insertion be present at the bedside during the insertion when less experienced physicians are involved. The same is true for the nurses in having an experienced nurse skilled in PAC setup and monitoring be available and present.

1. Introducer insertion. The first step is obtaining venous access with a large-bore (8.5 Fr) introducer catheter (Figure 107–6). The right internal jugular or left subclavian veins are preferred as these are the least likely to result in kinking of the PAC compared to the left internal jugular or right subclavian vein sites. The femoral vein can be used if the neck veins are not accessible, but it can be more difficult to direct the PAC into proper position. As with any central venous assess, the access site is prepared in sterile fashion with drapes covering the entire patient; venous cannulation, guidewire placement, and confirmation should be done under ultrasound guidance.

2. Transducer preparation. The transducers are connected to the monitor via electrical cables and connected to tubing that will be attached to the proximal and distal PAC ports during insertion. Transducers are also connected to a saline flush apparatus.

3. PAC preparation. Before insertion of the PAC place the catheter sheath over the PAC and inflate the balloon with 1.5 mL of air via the modified blood gas syringe. Check for symmetric expansion of the balloon and for leaks (Figure 107–7).

   • Place stopcocks on PAC the distal and proximal ports and then isolate them by using a clamp to pinch the sterile sheets around the proximal catheter while retaining the balloon inflation syringe within the sterile field and hand off the ports to the assistant who connects the transducer tubing to the ports (Figure 107–8). A saline-filled syringe may be connected to the VIP port and that is also flushed. The VIP port is connected to IV fluid after PAC insertion.

   • The ports are then flushed. The flush should be cleared of any air bubbles. During flushing, the flow out from the distal port should be brisk; if it only trickles, the pressure bag compressing the saline flush is not inflated properly and needs more inflation. The display is checked for waveforms created by shaking the PAC, noting that both the CVP and PAP waveform displays are moving. Place a sterile-gloved finger to occlude the distal port and the distal port waveform should rise on the monitor (Figure 107–9). If the waveforms are not seen, make sure the monitor is set up properly, check the transducer connections, and electrical cables that may need to be reconnected or changed. If the CVP waveform increases when the distal port is occluded, the transducer tubing or the electrical cables are reversed and need to be switched. If the waveforms are small check the scale—if it is set to arterial pressure ranges, for example, 0 to 120 or 0 to 160, the PAC waveforms will appear very small. The scale should be set to 0 to 30 or, if elevated PA pressures are expected or encountered, 0 to 60.

   • Finally the bed should be leveled with the transducer at the midaxial level and the transducer zeroed to atmosphere. Then with the PAC connected to the transducer, the distal port is positioned at the midaxillary line and the pressure reading on the monitor should be zero. If not, the level of the bed/transducer should be readjusted.

4. PAC insertion. The ECG is closely observed for arrhythmias during PAC insertion (see 4a later). The PAC is initially inserted until the 15- to 20-cm mark is at the entry of the introducer; this places the balloon tip outside of the introducer, near the SVC or RA level with the nonpulsatile small-amplitude oscillations CVP or RAP waveform displayed on the distal port waveform (Figure 107–10). The normal CVP or RAP is 0 to 8 mm Hg. The balloon is then inflated with 1.5 mL of air and advanced quickly passing the tricuspid valve until the pulsatile RV pressure waveform is obtained, usually at 25 to 30 cm of PAC insertion. The normal systolic/diastolic RV pressure is 15 to 30 or 0 to 8 mm Hg. If the RV pressure waveform is not obtained after insertion to more than 40 cm, the balloon is deflated and the PAC is withdrawn and reinserted as earlier until the RV waveform is obtained. Once the RV pressure waveform is obtained, continue advancing the PAC until a step-up in the diastolic pressure is visible on the waveform, confirming passage across the pulmonic valve and placement into the proximal PA (see Figure 107–10). This usually occurs at 35 to 40 cm of insertion, but can be longer, for example, 40 to 60 cm in tall patients and/or with cardiomegaly. A dicrotic notch due to closure of the pulmonic valve may be visible. The normal PA systolic/diastolic pressure is 15 to 30/4 to 12 mm Hg. With further advancement of the PAC into the PA, the pulsatile waveform eventually disappears leaving a nonpulsatile waveform similar to the RAP. This is the PAOP, at the same pressure level as the pulmonary artery diastolic pressure (usually < 12 mm Hg). When the wedge pressure first becomes evident, the balloon at the tip of the catheter should be deflated (see point 5 later).

   • Both atrial and ventricular arrhythmias can occur during passage of the PAC tip through the RA and RV to the PA. Ventricular are more common. Although inflation of the balloon at the tip of the PAC protects the endocardium from tip irritation, arrhythmias can still occur. These arrhythmias usually resolve as the PAC tip passes into the PA. However, if arrhythmias do not immediately resolve with passage into the PA or if there is difficulty in passing the catheter into the PA, the PAC balloon should be deflated and the PAC immediately removed. Occasionally arrhythmias may persist even after the PAC is removed. After treatment of the arrhythmia, if PAC insertion was felt to be the cause, PAC reinsertion should not be done.

5. PAC placement. Once a PA tracing or PAOP is obtained, deflate the balloon and observe the waveform. If the tracing is in the PA (PA waveform), then after balloon deflation, the PA waveform should remain. If a RV waveform results, the balloon should be reinflated; if the PA tracing is obtained then the PAC should be advanced 1 to 2 cm and the balloon deflated; repeat until a PA tracing remains after balloon deflation. If a PAOP waveform is obtained, deflate the balloon and observe for a PA tracing. If the PAC remains wedged, withdraw the PAC about 1 to 2 cm until a PA tracing is obtained.

6. PAC fine adjustment. To ensure that the PAC tip is not further out in the PA than necessary, observe the distal PAC waveform—it should be a PA waveform and not wedged (ie, without a PAOP waveform). If it is wedged, see 5 earlier. If a PA waveform is present, then inflate the balloon to 1 mL (it takes 1 mL to open the balloon, 1.5 mL to fully inflate)—if a PAOP waveform is seen, then deflate the balloon and withdraw about 1 to 2 cm and repeat until the balloon can be filled with the full 1.5 mL of air either without wedging (PA waveform remains) or a PAOP waveform occurs—with the full 1.5 mL. This will ensure that the tip of the PAC is not too distal. Note that it is safer if the balloon takes the full 1.5 mL remaining in the PA without wedging. Recall that the PAOP is the least reliable number obtained from the PAC, so better to err on the side of a more proximal PAC tip location, as long as it is not in the RV.

7. PAC postinsertion. Once the PAC is in proper position with the balloon deflated and displaying a PA waveform via the distal port, the insertion depth at the entry point into the introducer should be noted (eg, 45 cm) and then documented in the procedure note. A chext x-ray (CXR) is obtained to view the course of the PAC and the tip position. Nearly 80% of the time the tip is in the right PA (Figure 107–11), about 20% of the time it is in the left PA. If the tip of the PA catheter is nearer the periphery of the lung, for example, beyond the midclavicular line, the PAC is likely overinserted and needs to be checked (see point 8). In smaller patients, the PAC may wedge at 40 to 42 cm and the PAC may need to be pulled back into the PA to less than or equal to 40 cm. If this occurs it should be appreciated that the CVP port will lie within the introducer, that is, if the CVP port is at 29 cm and the introducer length is 13 cm and the PAC is inserted to 40 cm (or less), both the VIP and CVP ports will lie within the introducer sheath. This may affect the accuracy of CO measurements.

8. Post-PAC insertion monitoring. The PA waveform should be repeatedly assessed by nursing and physician staff, to display a PA waveform and not the wedging (PAOP) or the RV. If a PAOP or an RV waveform is observed, check the insertion depth for migration and adjust the catheter as described in points 5 and 6 earlier. When a PAC is in place, if any atrial or ventricular arrhythmias occur, PAC tip migration should be ruled out as a possible cause by ensuring that the waveforms indicate proper position in the PA (see later).

The following are additional points essential to maximize safety when using the PAC.

1. The main principle of balloon deployment is that when inserting (advancing) the PAC, the balloon should always be inflated; when pulling out (withdrawing) the PAC, the balloon should always be deflated. (Except for when you are first inserting the PAC 15 to 20 cm through the introducer—obviously with the balloon deflated.)

2. Usually ventricular (but occasionally atrial) arrhythmias may result from PAC migration or occasionally only from the PAC itself passing through the right heart chambers. If arrhythmias are clearly due to migration and are readily terminated with proper PAC adjustment, then the PAC can remain; however, if they are felt to be due to the PAC, the PAC should be removed.

3. Right bundle branch block (RBBB) may rarely occur with insertion of the PAC. In the setting of left bundle branch block (LBBB) there is the potential for complete heart block if RBBB occurs during PAC insertion. Therefore, in the presence of a LBBB it is prudent to have a transcutaneous pacemaker immediately accessible when inserting a PAC.

4. Care should be taken to note how much of the PAC has been inserted during PAC placement; no more than 40 cm of PAC should be inserted in attempting to advance the PAC into the RV; the PAC may be coiling in the RA and upon withdrawal, the PAC can wrap around the introducer and become knotted.

5. Multiple attempts at passing the PAC due to difficulty getting into the PA from the RV could result in cardiac perforation and tamponade. It is better to discontinue PAC insertion if things are difficult. Remember, the PAC is an invasive device that is diagnostic, not therapeutic.

6. PACs are heparin coated as this helps prevent thrombosis around the catheter. This bonding also has antimicrobial activity. Patients with heparin-induced thrombocytopenia (HIT) or a history of HIT, should not have a heparin-bonded PAC inserted. There are nonheparin-bonded PACs available for these patients.

7. The PAC balloon is made of latex. In patients with known severe (eg, anaphylactic reactions) latex allergies, a PAC insertion should be avoided unless a PAC with a nonlatex balloon is available via the manufacturer.

The PAC is an invasive monitor. A task force in 2003 reported the overall deaths due to the PAC to be between 0.02% and 1.5%.²² Complications can be divided into those resulting from the central venous access versus those resulting from the PAC itself. Complications of central venous access include bleeding, hematoma, arterial puncture, pneumothorax, arrhythmias (from guidewire insertion), air embolism, and hemothorax. Complications of PAC insertion include arrhythmias (greatest risk when the PAC tip is in the RV), RBBB, complete heart block, knotting of the catheter, valvular damage, cardiac perforation, and infection including catheter-related bloodstream infection and endocarditis. Complications of PAC overinsertion or migration if prolonged and unrecognized, and exacerbated by balloon inflation, include pulmonary infarction and pulmonary artery rupture with hemoptysis and pseudoaneurysm formation.

Complications can be reduced by careful insertion technique, the use of ultrasound for venous access, and only using the PAC when the information is likely to be helpful and for limited time periods—as short as possible, for example, 1 to 3 days.

Information from any hemodynamic monitor must be interpreted in light of the clinical data (eg, physical examination, ventilator, laboratory, radiologic), the treatment goal and the response to treatment, both hemodynamic and clinical. Hemodynamic data derived from the PAC is not pathognomonic; it is always subject to interpretation.

In the past, Starling curves (the relationship between contractility vs heart muscle stretch) were generated using PAC data comparing stroke volume (SV) on the Y-axis to PAOP on the X-axis. The problem is that PAOP is not an accurate reflection of preload, that is, ventricular cavity dimension or stretch. Subsequent studies have revealed the lack of Starling curves using PAOP versus generating Starling curves using radionuclide or echocardiography-determined preload.^5,23

The major diagnostic assessments that can be made using PAC data include hyperdynamic circulation, hypodynamic circulation, and pulmonary hypertension. Trends in cardiac output, stroke volume, and PAPs can help to guide hemodynamic interventions.

When using the PAC it is important to use accurate and reliable information:

1. The transducers should be leveled at the midaxillary level and zeroed. No air bubbles should be present within the tubing.

2. The systolic and diastolic PA pressures (PAS and PAD) constitute the reliable pressure data obtained from the PAC. They are more accurate than echocardiography estimates of PA pressure. The PA pressures are used to assess the presence and degree of PHTN and the changes in PA pressure that occur with changes in condition and treatment. The true LVEDP is less than the PAD with the gap between PAD and LVEDP increasing with PHTN.

3. If the PAC waveform goes from a clear RV tracing to suddenly damped and “wedged” appearing without seeing a PA waveform, suspect a proximal pulmonary embolism—the catheter tip may be lodging in a blood clot.

4. Do not use the CVP or PAOP; they do not reflect ventricular preload or intravascular volume better than flipping a coin.

5. Do not rely on CO measurements if there is evidence of significant TR.

6. Optimize the accuracy of CO measurement by holding rapid fluid infusions via the introducer or PAC for at least 30 seconds, turning off SCDs, and timing the start of injection to the same point of the respiratory cycle.

7. A high CO in the setting of an elevated lactate level rules out heart failure as the etiology and raises the suspicion for a local process, for example, mesenteric ischemia versus a systemic process, for example, septic shock–induced mitochondrial dysfunction. A high CO in the setting of pulmonary edema supports the diagnosis of ARDS rather than heart failure. (The PAOP should not be used to make this differentiation and has in fact been dropped from the current criteria for ARDS diagnosis.)

8. A very low CO on high-dose multiple inotropes should raise the suspicion of dynamic left ventricular outflow tract obstruction due to decreased preload and systemic anterior motion of the mitral valve into the left ventricular outflow tract. Despite an empty left ventricle, the PAOP is usually very high. An echocardiogram should be emergently done to rule this out. The treatment is fluid resuscitation, withdrawal of inotropes, and the use of pure noninotropic alpha agents (ie, phenylephrine) for hypotension.

9. When administering a fluid challenge, repeat the CO immediately after the challenge. An increase of 15% or more in the SV (or CO if the heart rate [HR] has not increased) constitutes a response to fluid.

10. If removing fluid by diuretic or renal replacement therapy, decreases in the CO (≥ 15%) may indicate the limits of fluid removal.

11. Repeat the CO and SV after starting an inotropic agent to assess the response.

12. Use both the mixed venous oxygen and carbon dioxide. Interpret the Svo₂ in terms of the hemoglobin (Hb) level and CO. An increased P[v-a]co₂ gradient (> 7) may signify a low CO unless the patient is hypermetabolic. If the patient is afebrile or only has a low-grade fever, they are not hypermetabolic and the increased P[v-a]co₂ gradient is a reflection of lower flow (CO).

13. Use echocardiography to supplement the PAC data that can be interpreted in the light of the direct preload and contractility information provided by echocardiography. Echocardiography can also determine if there is TR and if the CO is accurate.

We have discussed the standard PAC. Modified PACs include the following:

1. Continuous cardiac output PAC. Utilizes a heated coil to increase the temperature of the blood that then passes by a distally located thermistor-generating CO data without the need to inject a bolus of fluid. The CO is not actually continuous in the sense that it is updated every 2 to 3 minutes.

2. RV ejection fraction PAC. Estimates RV end-diastolic volume (preload) by measuring the decay in the thermodilution curve gated to the ECG R-R interval. It tends to overestimate the right ventricular end-diastolic volume (RVEDV). Also, accuracy diminishes with rapid or irregular heart rhythm.

3. Oximetric PAC. Modified with a fiberoptic bundle to measure continuous Svo₂.

Despite the greater cost of modified PACs, significant clinical advantages over the standard PAC have not been demonstrated. Finally there are also PACs modified with an RV port to allow simultaneous temporary cardiac pacing and PAC monitoring capability.