Congenital Heart Disease: The Catheterization Manual

Part 1

The Basics

Lisa Bergersen, Susan Foerster, Audrey C. Marshall and Jeffery Meadows (eds.)Congenital Heart DiseaseThe Catheterization Manual10.1007/978-0-387-77292-9_1© Springer Science+Business Media, LLC 2009

Hemodynamics

Lisa Bergersen¹ , Susan Foerster², Audrey C. Marshall¹ and Jeffery Meadows³

(1)

Department of Cardiology, Children’s Hospital Boston, Boston, MA 02115, USA

(2)

Washington University in St. Louis, St. Louis, MO 63110, USA

(3)

Division of Pediatric Cardiology, University of California at San Francisco, San Francisco, CA 94143, USA

Abstract

When starting in the cath lab it is essential that you have an understanding of the findings as you collect them. This requires an understanding of normal intracardiac saturations and pressures (the only way to know when something is abnormal) and the calculations used for flow and resistance. All cardiologists must clearly understand not only the basis of the calculations, but also the limitations intrinsic to them.

When starting in the cath lab it is essential that you have an understanding of the findings as you collect them. This requires an understanding of normal intracardiac saturations and pressures (the only way to know when something is abnormal) and the calculations used for flow and resistance. All cardiologists must clearly understand not only the basis of the calculations, but also the limitations intrinsic to them.

In practice, most of the calculations for cardiac index, Qp:Qs, and pulmonary vascular resistance (PVR) can be estimated in one’s head during the case as the saturations and pressures are being recorded. These estimates should then be compared to results generated by a computer, or more frequently by a calculator, later in the day after the examinations. If there is a large discrepancy, then it is likely the computer is wrong because you have selected incorrect data (e.g., Hematocrit put in as Hemoglobin).

Pressure Measurements

Accurate recording and interpretation of intracardiac pressure waveforms is paramount in a complete hemodynamic assessment. Before you record a pressure tracing you should be sure that it is of good quality and makes sense with what you know, or suspect, about the patient. It is, at best, a waste of time to record poor tracings. At their worst, poor tracings can lead to misinterpretation, inaccurrate, or missed diagnoses.

Optimizing pressure waveforms for analysis requires some knowledge of the means by which they are transmitted and recorded. Complete understanding requires consideration of the sensitivity, natural frequency, and frequency response of the system. Fortunately, detailed reviews of pressure measurement systems can be found in standard cardiac catheterization textbooks. Hi-fidelity pressure catheters provide the most accurate tracings, but these can be quite expensive and fragile.

Practically speaking, most catheterization laboratory setups rely upon fluid-filled catheters for pressure transduction. These (and most types of systems) must be first zeroed to ambient air pressure. The standard reference point for this is the midpoint of the LA. With the transducer at this height, the transducer membrane is exposed to atmospheric pressure, and the monitor is then adjusted to zero. Then the monitoring system must be calibrated, a process that in most systems is performed automatically. Finally, if you are using more than one transducer, the transducers must be equilibrated. This is typically done by simultaneous pressure measurement at the same intracardiac/vascular location, or by measuring two pressures of similar magnitude, switching ducers to check that when the transducers are switched between catheters they indicate the same pressures.

When using a fluid filled catheter, a well-flushed (no air bubbles), large bore, stiff and short catheter will provide the most reliable waveforms. Of course, enthusiasm for a good waveform should be tempered by practical considerations of the patient’s size and which catheters are needed for the procedure. Finally, remember, recognition of artifact is important. There are a host of artifacts that can mask or mimic pathophysiology.

Pressures and Waveforms

Right Atrial Pressure (RAp)

Normal right atrial pressure varies widely in response to myriad factors, including volume and respiratory status, heart rhythm, structure and function. Normal mean right atrial pressures for a child can be considered anywhere in the range 3–6 mmHg, although lower and occasionally negative values, particularly with inspiration or airway obstruction (snoring), are not infrequent. Many children with congenital heart disease (CHD) have elevated RAp, the long-term effects of which can be attested to by any physician caring for adults with CHD, Figure 1.

Fig. 1

Normal RA pressure waveform with a dominant A wave

The normal right atrial pressure waveform consists of two, and sometimes three, positive waves (a, c, and v) followed by three negative deflections (x, x′, and y, respectively). The a wave results from atrial contraction (and is therefore absent in situations such as atrial fibrillation). Increased a waves are most frequently seen in situations of AV valve (AVV) stenosis, and AV dissociation when the atrium contracts against a closed AVV. Relatively increased a waves can occur with atrial contraction into noncompliant ventricles. The x decent follows the a wave and denotes atrial relaxation followed by AVV closure. The c wave is sometimes evident and occurs during ventricular systole as the closed AVV bulges into the atrium. While it is of physiologic interest, the clinical significance of the c wave is limited. The subsequent x′ decent reflects the combined effects of continued atrial relaxation and descent of the AVV during continued ventricular contraction. The v wave follows, denoting atrial filling while the AVV is still closed. The opening of the AVV followed by atrial emptying is represented by the y descent. The y descent is characteristically slowed in AVV stenosis. The subsequent period of slow or minimal ventricular filling with little change in atrial pressure (open AVV) is termed diastasis and is quickly followed by atrial contraction (a wave). RA pressures should drop with inspiration during spontaneous respiration, the absence of which is the hemodynamic equivalent of Kussmaul’s sign.

Right Ventricular Pressure (RVp)

Normal (subpulmonary) right ventricular pressure also varies considerably with age, respiratory status, heart rhythm, structure, and function. Peak systolic pressure is typically 20–30 mmHg. End-diastolic pressure is typically equal or just slightly less than the right atrial a wave at 3–6 mmHg, Figure 2.

Fig. 2

Normal RV pressure waveform, simultaneous with LV pressure waveform

The right ventricular waveform is marked by a rapid rise during isovolumic contraction, followed by the peak systolic pressure before isovolumetric relaxation and a fall to minimum diastolic pressure (often near zero). There is a slow rise during diastolic filling during which a small RV a wave inflection may be seen as a result of atrial contraction just prior to end-diastole and subsequent isovolumic contraction.

This a wave, sometimes referred to as the atrial kick, frequently is accentuated in patients with first-degree AV bock. Peak RV systolic pressure is elevated in the presence of any downstream obstruction including right ventricular outflow obstruction (subPS or valvar PS), main or branch pulmonary artery stenosis, elevated pulmonary vascular resistance (pulmonary hypertension), or any lesion causing significant pulmonary venous or left atrial hypertension.

Pulmonary Artery Pressure (PAp)

The mean pulmonary artery pressure is usually less than 20 mmHg, with a systolic peak equal to or slightly less than that of the RVp, Figure 3. The pressure pulse is characterized by a relatively slow upstroke, peak systolic pressure, a small dicrotic notch, and slow fall to end diastole. The pulmonary artery pressure tracing provides in a single waveform significant insight into both right and left heart hemodynamics.

Fig. 3

Normal PA pressure waveform with simultaneous LV pressure waveform

If you arrived at the PA through the right atrium and ventricle, a peak systolic pressure significantly lower (>10 mmHg) than RV pressure denotes right ventricular outflow tract obstruction, which deserves further characterization to distinguish subvalvar, valvar, or supravalvar stenosis. Normally, there is a demonstrable diastolic pressure gradient between the PA and RV, the absence of which suggests truly free pulmonary regurgitation. Otherwise, the PA diastolic pressure more typically approaches the LA pressure.

Elevated PA pressures can occur with either increased flow (e.g., VSD), increased resistance (e.g., pulmonary vascular occlusive disease or PVOD), or downstream obstruction (e.g., left atrial hypertension). As such, it is important to use precise and unambiguous terminology. Remember that pulmonary artery hypertension, that is, high pulmonary artery pressure, is not synonymous with high pulmonary vascular resistance. For example, elevated PA pressures can occur with an unrestrictive VSD and normal pulmonary resistance. Full hemodynamic assessment is critical to distinguish between these, as important management decisions are based on the ability to manipulate the underlying process (e.g., you can replace/dilate a stenotic mitral valve, but you cannot as easily replace end-stage PVOD-affected lungs).

Pulmonary Capillary Wedge Pressure (PCWp)

A good pulmonary capillary wedge pressure (PCWp) resembles the left atrial pressure and waveform (Fig. 4) with a time delay of somewhere between 0.02 and 0.08 sec (Figure 4), unless there are significant collaterals or pulmonary vein stenosis. As such this waveform should have interpretable a and v waves and normal respiratory variation. An underwedged tracing usually has exaggerated systolic peaks as PAp is transmitted around the catheter. An overwedged PCWp typically lacks identifiable waveform morphology with a high drifting mean pressure.

Fig. 4

Normal PCWP and LV waveform

Left Atrial Pressure (LAp)

In the normal heart, LA pressure is higher than RA pressure, with mean pressures in the range 6–9 mmHg. Even with respiratory variation, LAp is never normally below atmospheric pressure.

The right and left atrial pressure waveforms are similar (Fig 5), but the v wave is usually dominant in the left atrium, ostensibly because of pulmonary venous contraction (e.g., the left atrial a wave is dominant in TAPVC). Increased a waves may be seen in mitral stenosis or in situations of poor LV compliance. Prosthetic mitral valves in the supra-annular position characteristically result in an increased v wave, probably due to the combined effects of a small, noncompliant LA and pulmonary venous contraction. However, increased v waves are more classically seen in mitral regurgitation. Overall, increased LAp can result from any of the above situations, significant left-to-right shunts, or LV dysfunction, Figure 5.

Fig. 5

Simultaneous RA and LA pressures

Pulmonary Vein Wedge Pressure (PVWp)

The pulmonary venous wedge pressure (PVWp) operates under the same principle as the PCWp, but in the opposite direction, and provides a reasonable estimate of PAp (albeit often slightly underestimating), when the mean pressure obtained is less than 15 mmHg (above this, it is imprecise). When the PAp is a major reason for the catheterization and the patient is potentially unstable or access to the PAs may make them so, this can provide a quick estimate in case things go awry.

Left Ventricular Pressure (LVp)

The left ventricular pressure (LVp) varies with age and a host of structural and hemodynamic factors. Obviously, the peak systolic pressure should equal the ascending aortic pressure, failing which there must be is subvalvar, valvar, or supravalvar obstruction that warrants further characterization Figure 6. The LV end-diastolic pressure (LVEDp) is a crude but valuable marker for LV diastolic health, in that elevated LVEDp (>10–12 mmHg in children) suggests poor diastolic ventricular properties and/or LV failure. Similarly, a steep slope of the diastolic portion of the LV waveform suggests poor ventricular compliance (see Figure 6 vs. Figure 4). Normal pressures vary according to age, with a progressive increase in the average LVEDp as patients progress to old age.

Fig. 6

LV and AO waveforms

Aortic Pressure

The aortic pressure pulse varies uniquely in morphology in health and disease, much of which is due to timing and magnitude of reflected waves. Generally, there is systolic rise, peak aortic pressure, and a variable dicrotic notch on the downstroke Figure 6. A widened pulse pressure (systolic minus diastolic pressure) is characteristic of run-off lesions, including significant aortic (or neoaortic) regurgitation, PDA, surgical shunts or significant aorto-pulmonary collaterals. More commonly in adults than children, a widened pulse pressure may be seen in the setting of arterial stiffening and bradycardia. In contrast, a decreased or narrow pulse pressure may be seen in low cardiac output states and/or tamponade.

Pitfalls and Tips

As noted by Dr. Keane in the so-called Green Book (due to the color of its cover): A catheterization is like a puzzle: everything must fit with everything else. If things do not fit, your case is probably not complete. The following are some tips about common problems:

At the start of every case, ensure that you have a correct baseline and that the transducers are calibrated and equalized. This is easy to forget to do when things move fast. Remember that re-zeroing and re-equilibration between transducers may be required during the case.

Know what to do with respiratory variation on a wave tracing. It is conventional to use the end-expiratory numbers, which are at the height of the tracing in spontaneously breathing patients and at the nadir in patients on positive pressure ventilation.

Poor waveform morphology can result from either an underdamped or overdamped system. Underdamped waveforms, demonstrated by fling or ringing, most commonly result from air in the catheter or transducer line, or the catheter position in a turbulent flow jet. The waveform has wide oscillation, high systolic spikes, and large negative overshoot waves. Try flushing the catheter or filling the catheter with blood or half-contrast. A similar appearance can occur with a catheter bouncing against the wall of the heart or vessel, in which case simple repositioning should work. The opposite, overdamped waveforms (sine wave), usually are the result of a loose connection or kinked catheter, and have a rounded out appearance with blunted and inaccurate limits. Checking the catheter and its connections or flushing the catheter will usually solve the problem.

Remember to consider the patient’s rhythm when you examine your waveforms. Nonsinus rhythm can drastically alter not only atrial waveform morphology but also absolute ventricular pressures and systolic flows.

Finally, if the waveform you are getting does not make sense, consider that you may not be where you think you are in the heart, or your assumptions about the patient’s physiology may be wrong. It is critically important to have the ability to rethink the patient’s condition, and thereby the catheterization procedure, continuously throughout the case.

Assessment of Flows and Resistance

Calculated systemic and pulmonary flows (and their ratio) are important components of almost all catheterizations. The most common methods for obtaining these include calculations utilizing the methods originally conceived by Adolph Fick, and thermodilution (dye dilution being abandoned some time ago). Both of these have their assumptions and limitations that are important to remember.

The Fick Method

In 1870, Adolph Fick described a method for calculation of blood flow. He never tested his theory, but subsequent physiologists have, and the Fick method remains an important means of determining cardiac output. Indeed, the Fick method probably is the most common means by which cardiac flows are calculated in the pediatric catheterization lab.

Derivation of the equation is simple, and if the concept rather than equation is understood, the strengths and limitations of this method will never be forgotten. Stated simply, the total uptake (or release) of a substance by an organ is the product of blood flow to that organ and the concentration difference of the substance in the arteries and veins leading into and out of that organ. So, using arterial and venous oxygen content and oxygen consumption, one can easily calculate flow. Therefore, if oxygen content is:

1.36 mL is the oxygen carrying capacity of 1 gm of hemoglobin.

The relatively small amount of dissolved oxygen in plasma is 0.003 mL O2/dL plasma/mmHg PaO2.

If the saturation is 98%, then use 0.98 in the formula, not 98.

then the important equation becomes:

* Note: Multiplication of the denominator by 10 is necessary to convert dL to L, if the hemoglobin is measured as g/dL.

Dissolved Oxygen

At 37°C, the solubility coefficient of O2 is ∼0.003 mL O2/dL plasma/mmHg PaO2, or 0.3 ml O2/dL plasma when PaO2 is 100 mmHg. Normally, in room air, this represents approximately 1.5% of the total O2 content and is usually ignored. If the patient is anemic, the proportion of dissolved O2 increases. The addition of supplemental O2 increases the contribution of the dissolved O2 and must be included in calculations.

Oxygen Consumption (VO2)

Oxygen consumption (VO2) is often assumed, and is readily available in tables (see the appendix). There is variation based on sex, age, and heart rate. Unfortunately, studies comparing assumed and measured VO2 show poor correlation and there are essentially no data published for the very small infant. Despite the limitations with assumed VO2, this method remains widely used.

Oxygen consumption can be measured, using a hood. The process involves a gas pump that extracts all exhaled air and passes it through a mixing system and then measures oxygen content. The difference between inhaled oxygen content and exhaled oxygen content, with known flow by the pump, enables estimation of VO2 (again based on the Fick principle). This assumes that no exhaled air is lost, that mixing is effective, and that volume of exhaled air equals volume of inhaled.

Common pitfalls and sources of inaccurate saturation measurements:

Use of inappropriate or inaccurate mixed venous saturation.

Samples not obtained as closely timed together as possible.

Wedged instead of free PA sample with end-hole catheters (artificially high saturation).

IVC saturations are inconsistent [both hepatic (low) and renal (high) are present].

SVC saturation can be unusually high if there is reflux from ASD, PAPVR, arterio-venous malformations, or tricuspid regurgitation in the setting of a VSD.

Failure to appreciate sedation-related hypoventilation during saturation assessment.

Inappropriate pulmonary vein saturation assumption.

Assumed oxygen consumption can be unreliable.

If VO2 is measured, inaccuracies can result if staff is unfamiliar with equipment.

A catheter course mistaken for LA is actually coronary sinus (sat is usually 40–50%)

Failure to include dissolved O2 when in supplemental O2.

Small saturation differences can exist without a shunt (and vice versa).

Using these principles, both systemic and pulmonary flows can be calculated using the appropriate arterial and venous parameters. For pediatric catheterization, we most often express the indexed flow (to BSA), distinguishing between cardiac output (CO; L/min) and cardiac index (Qs; L/min/m²). Because the VO2 tables are usually indexed…

Some general rules are the following:

For mixed venous saturation with no shunts, use the PA saturation.

If there is intracardiac shunting, use the SVC saturation for the mixed venous saturation. The SVC saturation used for calculations should never be higher than the PA saturation (unless there is anomalous pulmonary venous return or an oxygen-consuming tumor in the right ventricle!)

If you do not directly obtain a pulmonary vein or LA saturation, assume something appropriate to the clinical status of the patient, ~95–100% in the absence of lung disease.

In practice, when a patient is breathing room air (~21% FiO2) we neglect the relatively small contribution of dissolved oxygen to the oxygen content. This greatly simplifies the math, but should not be done when the patient is receiving supplemental oxygen.

Always document all assumptions in your catheterization report.

Make sure your numbers make sense.

Tip: The Hgb and the oxygen-carrying capacity of Hgb (the 1.36) are essentially the same in both the arterial and venous samples. If you do not have to include dissolved oxygen you can make approximate calculations during the case easier by doing half the math ahead of time. One way is the following: Divide the VO2 by the product of the Hgb and 0.136 (includes correction factors). This gives you a number which when divided by the relevant AVO2 difference quickly gives you the indexed flow. For example:

Your predetermined number derived as described above is 125/(12.8*0.136) = 72. If the MVO2 saturation is 79% and the aortic saturation is 99%, then 72/(99 – 79) = ~3.6 L/min/m². Always double check and confirm your math.

Qp:Qs Calculation

In patients with structural heart disease, mixing is common (and often necessary), making important the concept of the ratio of pulmonary to systemic blood flow (also known as Qp:Qs). Conveniently, much of the above formulae cancel out, leaving a simple equation:

For example, take a situation where the only structural problem is an atrial septal (ASD), as shown in Figure 7. In performing a hemodynamic run you find an SVC (mixed venous) saturation of 75% with PA saturations of 84% and equal pulmonary vein and aortic saturations of 99%. The net Qp:Qs in this patient is (99 – 75)/ (99 – 84) = 1.6.

Fig. 7

ASD with saturation measurements

Now consider a patient with hypoplastic left heart syndrome (HLHS) with shunt physiology Figure 8.

Fig. 8

HLHS with saturation measurements

In this patient you only need to know the SVC, pulmonary vein (sometimes assumed) and aortic saturations in order to calculate the net Qp:Qs because the only source of pulmonary blood flow is the shunt; therefore, the pulmonary artery saturation is equal to the aortic saturation. So, with an SVC saturation of 69%, pulmonary vein saturations of 99% and an aortic saturation of 87%, the net Qp:Qs is ~1.7. This is a similar net Qp:Qs as the ASD above but obviously the physiology is completely different. To give you an idea of how changes in mixed venous and arterial saturations affect Qp:Qs in a patient with this kind of shunted physiology consider Table 1 (a pulmonary venous saturation of 100% has been assumed).

Table 1.

Variation in Qp:Qs with shunted single ventricle physiology

Thermodilution

First introduced in the 1950s,