Components and determinants of cardiac output

This chapter is relevant to Section G3(ii)  of the 2017 CICM Primary Syllabus, which asks the exam candidate to "define the components and determinants of cardiac output including the effects of positive pressure ventilation". This uncontroversial by-the-numbers topic should hold no unexpected horrors for the exam candidates. Still, even though it has appeared in several past paper SAQs, each appearance seems to have caught the exam candidates by surprise:

• Question 3 from the second paper of 2021
• Question 19 from the first paper of 2014 
• Question 8 from the first paper of 2011. 

Judging by the content of these questions, the examiners mainly wanted us to know the definition of cardiac output and the factors which affect it. Question 19 from the first paper of 2014 also went into the thermodilution measurement of cardiac output, which is discussed elsewhere.

In, for lack of a better word, summary:

"Understanding cardiac output" by Jean-Louis Vincent (2008) is probably the best single revision resource for this specific topic, as it does not go into pointless detail. If pointless detail is in fact what you are after, David Young (2010) has you covered in 97 pages.  For some excellent historical perspectives, Carl Wiggers (yes that Wiggers) has a 1951 paper which summarises the previous century of research, as do Sequeira & van der Velden (2015).

Definition and normal value range of cardiac output

Though there may often be disagreement or alternative definitions for a lot of the terms in physiology, this one seems to be fairly fixed and agreed upon.

"Cardiac output is the volume of blood ejected by the heart per unit time"

That's pretty uncontroversial. The volume of blood ejected is generally expressed in litres, and the unit of time is usually one minute, giving a widely quoted "average" number of 5.0L/min.  From this, it logically follows that this parameter has no clinical utility unless it is somehow related to the patient's body size, as 5.0 L/min would be vastly excessive for a person of small stature, and vastly inadequate for a humongously large person.  For this purpose, cardiac output is usually described in terms of "cardiac index", which by convention is gross cardiac output in litres of blood per minute divided by the surface area of the body. This convention has existed since the early 1950s (Tanner, 1949; Taylor & Tiede, 1952). Prior to that, investigators had a  tendency to present the cardiac output as percentage of body weight pumped per second, which would generate unwieldy numbers, like 0.00283 %/sec. George Stewart (of the Stewart & Hamilton fame) in 1897 summarised some of these studies, many of them variations on the theme of emptying an animal's blood volume into a bucket. 

Anyway, it would be amiss for this digression-prone author not to digress here on the boundaries which frame cardiac output for the human body. As with everything, there are maximums and minimums for this parameter. Åstrand et al (1964) forced a group of healthy volunteers to pedal some exercise ergometers and determined that maximal cardiac output was 18.5 L/min for women and 24.1 L/min for men (that's a cardiac index of 14)  which is a fairly spectacular five-fold increase from the resting value.

These values are probably not a real ceiling. These were relatively normal healthy people (university students studying physical education, i.e. teachers in training), which means that conceivably, some kind of meth-powered gym mutant might significantly exceed this value. Textbooks occasionally bring up 35-40L/min as the maximum cardiac output for the elite athlete subgroup, but it is difficult to find the exact study where this has come from.  The origin of these values is probably an old study by Ekblom & Hermansen (1968), who measured them from Swedish youths involved in professional sport (for example, one of the test subjects was the world champion in the bicycle team race of 1967). That guy (Gösta Pettersson) only got a maximum cardiac output of 39.8L:/min; some nameless silver champion in the sport of orienteering got 42.3 L/min, which appears to be the highest on record.

In the other direction, the minimum cardiac output required to sustain life is probably very individual and dependent on the metabolic rate. With that said,  assuming that the average 70kg body consumes about 200ml of oxygen per minute and the oxygen-carrying capacity of slightly anaemic blood being 134ml/L, one can estimate that the cardiac output would need to drop below about 1.5L/min (CI = 0.9)  before oxygen delivery fails and anaerobic metabolism becomes necessary. Realistically this seems to happen in the territory of 2.0-2.5L/min, probably because the body is composed of tissues which require unequal amounts of oxygen delivery, and which have unequal importance for the survival of the organism as a whole. Perhaps because of this, Kasnitz et al (1976) found that all their cardiogenic shock patients developed severe lactic acidosis and died if their cardiac output fell below 2.5L/min.

Factors which affect cardiac output

The heart being a pump which functions in a pulsatile fashion, it is possible to describe the cardiac output in terms of stroke volume and heart rate.

• Cardiac output = stroke volume × heart rate

Beyond this, stroke volume has several determinants, which are:

• Cardiac contractility
• Preload
• Afterload

However, cardiac output also determines preload and afterload, and preload determines contractility, so all these elements are interconnected and it is impossible to separate them into a perfect clockwork model of the cardiac output. Most textbooks will therefore list all of these variables together. 

Heart rate as a determinant of cardiac output

Though the equation [Cardiac output = stroke volume × heart rate] seems logical, in fact the use of heart rate as a straight multiplier here is a bit facile, as it actually also factors into the stroke volume. Consider that the heart rate determines diastolic filling time, which in turn determines the stroke volume by the Frank-Starling mechanism. Thus, at especially rapid heart rates, diastolic filling might actually be insufficient, and the cardiac output might drop.

This relationship has been investigated by multiple authors. For example, Sugimoto et al (1966) increased the atrial rate of anaesthetised mongrel dogs and recorded their cardiac output. At a certain heart rate (about 200-250) the cardiac output achieved its maximum value, and then began to fall - as can be seen from the (slightly modified) original diagram below.

From this, one can appreciate that for every individual, there will be some maximum heart rate which achieves the best haemodynamic performance. A truistic equation (max HR = 220 minus age) is often trotted out, but it is probably wildly inaccurate (Antonacci et al 2007). All that can be safely said is that each person will have some optimum heart rate at which cardiac output will be maximal, and this value decreases with age, not to mention pathological factors like valvular disease and diastolic function.

Stroke volume as a determinant of cardiac output

The stroke volume is defined as 

"the volume of blood pumped out of the left ventricle of the heart during each systolic cardiac contraction"

Again, that's fairly uncontroversial. It is usually calculated as the difference between the end-diastolic and end-systolic left ventricular volumes. The number of millilitres of blood is the stroke volume, and the ratio of the two volumes is the ejection fraction - which will be explored in detail elsewhere, and here it will suffice to say that it is possible to have a terrible-sounding ejection fraction with still a fairly reasonable stroke volume (eg. if you have dilated cardiomyopathy). The normal value for a normal sized person is about 70ml, 

Stroke volume factors into cardiac output, but cardiac output also factors into stroke volume. And stroke volume is determined by several parameters which are also affected by the stroke volume. These interdependent factors are afterload, preload and contractility. Weber et al (1974) demonstrated this elegantly in the isolated canine heart, by setting up constant conditions and then altering each variable individually. They paced the ventricles at a constant rate, controlled preload and afterload with a servopump, and adjusted contractility by adding or subtracting calcium from the perfusate.  This laboratory study is of course quite unrealistic, as the control of the variables had made it impossible to observe what backward effects might take place owing to the interconnectedness of these parameters, but it did serve an important purpose of demonstrating each effect independently. After much soulsearching, the author had decided that ultimately the exact numbers generated from those disembodied dog ventricles were immaterial, and the relationship itself was the most important thing to depict in a diagram:

In summary, 

• Stroke volume increases with increased preload, up to a plateau, beyond which it begins to decrease again
• Stroke volume decreased with increased afterload, in a fairly linear fashion
• Stroke volume increases with increased contractility, for any given preload and afterload value

These relationships are probably the key takeaway point of this chapter.

Preload as a determinant of stroke volume

Preload is an important determinant of cardiac output. Its definition is discussed elsewhere, and here it will suffice to say that when people study this variable, they usually use end-diastolic volume as a surrogate. When asked to discuss this in a viva, the trainees will usually be expected to draw the Frank-Starling cardiac function curve in a labelled diagram, such as this one:

There's a whole variety of these, and the CICM trainee is advised to uncritically place their trust in one plausible version. It does not matter which one they select, and it is possible to pass the exam without examining where it might have come from in any great detail, or knowing much more about it. The version demonstrated here is basically identical to the one used by Kam in his physiology textbook ("Studies of mechanical performance of the whole heart", p 133 of the 3rd edition from 2015). The numbers used for this diagram were acquired from Kanstrup & Ekblom (1982), but are so variable and meaningless that they could just as easily have arrived in a dream. 

Afterload as a determinant of stroke volume

Afterload is an important influence on stroke volume, and one of the primary determinants of cardiac performance. Specifically, for any given set of stable conditions (stable heart rate preload and contractility) increasing the afterload will usually decrease the stroke volume. Here's a grainy scan of such a relationship from some foetal lamb experiments by Hawkins et al (1989).

Without going into too much detail about what the determinants of afterload are or how you might define it, one can say that afterload also influences preload (by increasing the end-systolic volume) and contractility (by various mechanisms, not the least of which is the Anrep effect), and is in turn influenced by these factors. For example, in the Hawkins diagram here, B C and D are different (increasing) preload values.

Cardiac contractility as a determinant of stroke volume

At any given afterload and preload value, the increase in contractility will increase the stroke volume. In fact that's almost a part of the definition of contractility. As contractility increases, at the same preload, the stroke volume will increase, as is demonstrated in this classic diagram from Sarnoff & Berglund (1954). The investigators produced a decrease in contractility by a crude but effective method (they progressively clamped the coronary arteries).

Similarly, if one increases the contractility while maintaining a stable afterload, the stroke volume will also increase. Unlike the dreary ubiquity of the ventricular function curves depicted above, a real experiment involving afterload and contractility is a rarity, hard to find. Here is a slightly altered one from Weber et al (1974), where contractility was enhanced by the addition of calcium into the perfusate. The coloured lines represent measurements taken with the same preload volume.

Right and left cardiac output

Question 8 from the first paper of 2011 specifically asked for the factors that affect the output of the right ventricle. A suspicious person would interpret that as setting up for a sequel where the LV is the protagonist. Ergo:

Determinants of Cardiac Output,
Separated by Ventricle for Some Reason

Factor	Right ventricle	Left ventricle
Heart rate	The cardiac output of both the RV and the LV are affected by the heart rate in the same way.
Stroke volume	On average, though there might be beat-to-beat variations, the stroke volume both the ventricles is the same
Factors which affect afterload	Pulmonary vascular resistance is much lower (this decreases RV afterload), which is affected by many factors Thin wall (increases afterload) Positive intrathoracic pressure (increases afterload)	Systemic vascular resistance is much higher (this increases LV afterload) Thick wall (decreases afterload) Positive intrathoracic pressure (decreases afterload)
Preload	Determined by right atrial pressure and mean systemic filling pressure Positive intrathoracic pressure (decreases preload)	Determined by left atrial pressure Positive intrathoracic pressure (decreases preload)
	The preload of both ventricles are affected by: Pericardial compliance and pericardial content Total blood volume
Contractility	For both ventricles, contractility is affected by: Heart rate (Bowditch effect) Afterload (Anrep effect) Preload (Frank-Starling mechanism) as well as cellular and extracellular calcium concentrations and temperature
Effect of ventricular interdependence	The compliance of the right ventricle is decreased in systole by the contraction of the interventricular septum	The compliance of the left ventricle is decreased in diastole by dilatation of the right ventricle