What cardiac output is
Every cell in the body needs a continuous supply of oxygen, and the only way it gets one is if blood keeps moving. Cardiac output is the measure of that movement: the volume of blood the heart pumps out in one minute. It’s a flow, so it’s measured in litres per minute.
The heart works in cycles, not in a continuous stream. It fills, it squeezes, it fills again. So the amount it shifts in a minute comes down to two things only: how many times it squeezes, and how much it moves with each squeeze.
CO = HR × SV
where:
CO (L/min) = cardiac output
HR (beats/min) = heart rate
SV (mL/beat) = stroke volume
A stroke volume of about 70 mL at a rate of about 75 beats per minute gives roughly 5.25 L/min (a resting cardiac output of 5–6 L/min is the usual quoted figure for a 70 kg adult). In maximal exercise a normal person can push this up around fivefold, and in elite athletes peak output may reach 40 L/min.
Cardiac index — comparing across body sizes
Five litres a minute is generous for a small elderly woman and nowhere near enough for a large young man. To compare people fairly, cardiac output can be scaled to body surface area.
CI = CO / BSA
where:
CI (L/min/m²) = cardiac index
CO (L/min) = cardiac output
BSA (m²) = body surface area
Normal resting cardiac index is 3.0–3.5 L/min/m². Stroke volume can be scaled the same way to give the stroke volume index, normally 33–47 mL/m² per beat.
Why flow is not the same as pressure
Cardiac output is flow. Blood pressure is what that flow generates when it meets the resistance of the circulation. The two are linked by Darcy’s law — pressure equals flow multiplied by resistance — which for the systemic circulation gives:
MAP = CO × SVR
where:
MAP (mmHg) = mean arterial pressure
CO (L/min) = cardiac output
SVR = systemic vascular resistance, the resistance of the systemic circulation, set mostly by arteriolar radius
Stroke volume — how much each beat moves
What is left in, what goes out
The ventricle never empties completely. At the end of filling it holds a certain volume, and at the end of ejection it still holds some. Stroke volume is simply the difference between the two.
SV = EDV − ESV
where:
SV (mL) = stroke volume
EDV (mL) = end-diastolic volume, the volume in the ventricle at the end of filling
ESV (mL) = end-systolic volume, the volume left behind at the end of ejection
The proportion of the filled ventricle that actually gets ejected is the ejection fraction:
EF (%) = (SV / EDV) × 100 = ((EDV − ESV) / EDV) × 100
Ejection fraction is the number echocardiography reports, and it’s the most common clinical index of how well the ventricle is contracting.
The three things that set stroke volume
Stroke volume is a function of three factors:
- Preload — how stretched the muscle is before it contracts. Strictly it’s the myocardial sarcomere length just prior to contraction. You can’t measure sarcomere length in a living patient, so we use surrogates: end-diastolic volume on echocardiography, or end-diastolic pressure from a catheter. Central venous pressure approximates right ventricular end-diastolic pressure, and pulmonary capillary wedge pressure approximates left ventricular end-diastolic pressure. More stretch means a more forceful contraction and a bigger stroke volume — that relationship is the Frank–Starling mechanism.
- Afterload — what the ventricle has to push against. It is the sum of the forces, elastic and kinetic, opposing ejection, and it equates to the wall stress developed in the ventricle during systole. Its surrogates are mean arterial pressure and systemic vascular resistance for the left ventricle, and pulmonary vascular resistance for the right. As afterload rises it becomes harder to pump against increased resistance, so stroke volume falls.
- Contractility — how hard the ventricle can squeeze, regardless of how full the ventricle is or what it’s pushing against. Formally it’s the ability of the myocytes to generate mechanical power at a given preload and afterload. It’s mainly influenced by availability of intracellular calcium. Sympathetic stimulation, parasympathetic blockade and inotropic drugs raise it; β-blockers, calcium channel blockers, volatile and intravenous anaesthetic agents, ischaemia and sepsis lower it.
So: heart rate acts directly on cardiac output, while preload, afterload and contractility act on it indirectly, through stroke volume. Every intervention that changes cardiac output — a fluid bolus, a vasopressor, a volatile agent, a pacing wire — is working through one of those four handles.
Importantly: the atria contribute to filling too. Atrial contraction tops up the ventricle at the end of diastole, the “atrial kick”. In a normal compliant ventricle its loss is tolerated, but in a stiff, hypertrophied ventricle — as in aortic stenosis — atrial contraction can contribute up to 40% of left ventricular end-diastolic volume. That is why the onset of atrial fibrillation in a patient with aortic stenosis can be catastrophic: stroke volume falls off a cliff because the top-up disappears.
Where the blood comes from — venous return
A pump can only move what arrives at it. Venous return is the rate at which blood comes back to the heart, and over any sustained period it has to equal cardiac output — otherwise blood would pile up on one side of the circulation. The veins hold around 65% of the total blood volume, so they act as the reservoir the heart draws on.
What drives blood back is a pressure gradient, exactly as flow anywhere else in the circulation is:
VR = (MSFP − RAP) / RVR
where:
VR (L/min) = venous return
MSFP (mmHg) = mean systemic filling pressure, the pressure everywhere in the circulation if the heart stopped and flow ceased; normally about 7 mmHg
RAP (mmHg) = right atrial pressure
RVR = resistance to venous return
Mean systemic filling pressure sounds abstract, so it helps to picture it. Imagine stopping the heart and letting the pressures equalise. What’s left is the pressure produced purely by the volume of blood sitting inside vessels of a given compliance — like the pressure in a hosepipe with the tap shut, set only by how much water is in it and how stretchy the pipe is. That residual pressure pushes blood back to the right atrium.
Anything that changes one of the three terms changes venous return:
- Mean systemic filling pressure: circulating volume (haemorrhage lowers it, fluid resuscitation raises it) and venous compliance (venoconstriction raises it, venodilatation lowers it).
- Right atrial pressure: spontaneous inspiration makes intrathoracic pressure negative, drops right atrial pressure and improves venous return — the respiratory pump. Positive pressure ventilation does the reverse. Pericardial constriction and tamponade raise right atrial pressure and choke filling.
- Resistance to venous return: posture, and anything compressing the great veins — obesity, the gravid uterus, a laparoscopic pneumoperitoneum.
The skeletal muscle pump helps too: leg muscles contracting against competent venous valves milk blood back towards the heart.
Most of what we do to a patient’s cardiac output in theatre acts on venous return first. Spinal anaesthesia venodilates and drops mean systemic filling pressure. Positive pressure ventilation and pneumoperitoneum raise right atrial pressure and resistance to venous return. Head-up positioning pools blood in the legs. None of these touches the heart itself — they’re just changing preload.
Heart rate — how often the pump fires
The sinoatrial node sets the pace
The heart doesn’t need a nerve supply to beat. Several parts of its conducting system generate their own rhythm, and the fastest of them wins — that’s the sinoatrial node, which is why it’s the normal pacemaker. Below it, the atrioventricular node and the bundle of His have their own slower intrinsic rates, which only surface if the structures above them fail.
The reason for this intrinsic rate is that between beats, the sinoatrial node membrane potential drifts steadily upwards from around −60 mV (sources give the starting point anywhere from −50 to −70 mV) until it reaches threshold at about −50 mV, at which point an action potential fires. That upward drift is phase 4 of the pacemaker potential, and its steepness is what sets the rate.

How the autonomic nervous system moves the slope
Both arms of the autonomic nervous system act on the pacemaker cells, and both work by changing the phase 4 slope.
Sympathetic stimulation
- Noradrenaline from cardiac sympathetic nerves, plus circulating adrenaline, acts on β₁-adrenoceptors.
- Membrane permeability to sodium and calcium rises.
- The membrane potential drifts up faster — phase 4 steepens.
- Threshold is reached sooner, so the interval between beats shortens.
- Rate rises, up to around 200 bpm in an adult.
Sympathetic stimulation also does two other things:
- it increases contractility (a positive inotropic effect, by loading the sarcoplasmic reticulum with calcium), and
- it speeds relaxation (a positive lusitropic effect). Faster relaxation shortens systole, which protects the diastolic interval the ventricle needs for filling.

Parasympathetic stimulation
- Acetylcholine from vagal terminals acts on muscarinic M₂ receptors.
- Membrane permeability to potassium rises.
- The cell hyperpolarises — it sits more negative, so it has further to travel — and phase 4 flattens.
- Threshold is reached later, so the interval between beats lengthens.
- Rate falls.
Sympathetic: ↑Na⁺ and Ca²⁺ permeability → steeper phase 4 → threshold sooner → faster rate.
Parasympathetic: ↑K⁺ permeability → hyperpolarisation and flatter phase 4 → threshold later → slower rate.
The wiring
- Parasympathetic: preganglionic fibres leave the nucleus ambiguus in the medulla and travel in the vagus nerve. The right vagus supplies mainly the sinoatrial node, the left mainly the atrioventricular node. Vagal effects are largely confined to the atria.
- Sympathetic: preganglionic fibres arise from the intermediolateral columns of the upper thoracic cord, T1–T5 (the cardioaccelerator fibres), under the control of the rostral ventrolateral medulla. The right paravertebral ganglia mainly influence rate; the left mainly influence contractility. Postganglionic fibres reach the whole heart, not just the atria.
An easy way to keep the sides straight: the right-sided supply, both vagal and sympathetic, sits closest to the sinoatrial node and therefore drives rate. The left-sided supply reaches the left ventricle and drives force.
The two arms are not balanced. At rest, the parasympathetic system is tonically active in the heart. Acetylcholine is being released continuously onto the sinoatrial node the whole time.
This is why the resting heart rate is not idling naturally at 60–80. It’s being actively held back. Take the vagus away and the rate climbs: a denervated heart, such as a transplanted one, runs at roughly 100–120 bpm, and the isolated sinoatrial node’s own intrinsic rate is quoted at anywhere from about 80 to 110 bpm depending on the source. A fit athlete’s resting bradycardia is a heavier vagal brake at work.
Why? Parasympathetic effects are quicker in onset and offset than sympathetic ones, so vagal withdrawal is how the body achieves beat-to-beat rate control — at the onset of exercise, rate rises first by removing the brake, and only afterwards by sympathetic drive.
Moment to moment, the main input to the cardiac parasympathetic system is the baroreflex. A rise in arterial pressure stretches the baroreceptors in the carotid sinus and aortic arch, which inhibit sympathetic outflow and activate the vagus, so heart rate, stroke volume and systemic vascular resistance all fall and pressure comes back down. A fall in pressure does the opposite.
What happens if the vagus wins outright
Strong vagal stimulation can silence the sinoatrial and atrioventricular nodes completely. The ventricles then stop for 5–20 seconds until the bundle of His fibres take over at their own intrinsic rate of 15–40 bpm. This is ventricular escape, and it’s the reason a profound vagal event in theatre — traction on the eye muscles, peritoneal stretch, laryngoscopy in a lightly anaesthetised patient — produces a pause followed by a slow rhythm rather than asystole.
Vagal stimulation also has a modest negative inotropic effect, reducing ventricular contraction strength by around 20–30%. It’s smaller than the sympathetic effect because there are relatively few vagal fibres reaching the ventricles.
Rate control at the bedside
Treating a bradycardia
- Stop the surgical stimulus first if there is one — release the eye retraction, desufflate the abdomen.
- Atropine 300-600 micrograms IV, repeated every 3–5 minutes to a maximum of 3 mg.
- Glycopyrronium 50–100 microgram increments as a slightly milder alternative.
- If that fails your consultant should know about it and might consider: adrenaline 2–10 micrograms/min or isoprenaline 5 micrograms/min, or transcutaneous pacing.
- Look for reversible causes — hypoxaemia, electrolyte disturbance, hypothermia, myocardial infarction, drugs and toxins.
The transplanted heart
Cardiac transplantation removes both sympathetic and parasympathetic innervation, so the resting rate sits high from the loss of vagal tone. The practical consequences are as a result of this:
- Atropine and glycopyrronium are ineffective — there’s no vagal tone left for an antimuscarinic to block.
- Indirectly acting sympathomimetics such as ephedrine work poorly, because they rely on releasing noradrenaline from nerve terminals that aren’t there. Only their direct component has any effect.
- Directly acting agents — adrenaline, isoprenaline — do work, and are the sensible choice.
- Neostigmine may still slow the heart and should be used with caution.
Children
In neonates and small children the sympathetic system is relatively underdeveloped while the parasympathetic supply is comparatively mature. Despite their high resting rate, an insult tips them into profound bradycardia easily, and the most important insult is hypoxia. Bradycardia in a child is hypoxia until proven otherwise, and the treatment is oxygen and ventilation before it is atropine.
β-blockers
β-blockers are competitive antagonists at β-adrenoceptors, and they hit cardiac output from both directions at once: lower heart rate and lower contractility, so lower output, lower blood pressure and lower myocardial oxygen consumption. Non-selective agents also block β₂, which cause raised systemic vascular resistance and can provoke bronchospasm.
The cardioselective (β₁) agents are bisoprolol, esmolol, atenolol and metoprolol. Esmolol is the useful one in theatre: given by infusion at 50–200 micrograms/kg/min, it’s broken down by red cell esterases with a half-life of about 10 minutes, so an unwanted effect wears off quickly. In β-blocker overdose, glucagon 3–10 mg IV followed by an infusion of 3–5 mg/h bypasses the blocked receptor.
The two levers pull on each other
Treating heart rate and stroke volume as independent would make the maths easy: double the rate, double the output. Physiology doesn’t work like that as doubling the rate reduces the time available for everything else in the cardiac cycle.
Faster rate, stronger beat — the Bowditch effect
Raise the heart rate and the myocardium contracts more forcefully, entirely on its own, without any change in nerve traffic or drugs. This is the Bowditch effect, also called the Treppe or staircase effect, and it operates over roughly 40–140 bpm.
The mechanism is myocardial calcium retention:
- As rate rises, each cardiac cycle shortens, and diastole shortens more than systole.
- More beats per minute means more systolic calcium entry through L-type calcium channels.
- Sodium also enters with each beat, and the shortened diastole leaves the Na⁺/K⁺-ATPase too little time to pump it all back out, so intracellular sodium rises.
- The Na⁺/Ca²⁺ exchanger normally uses the inward sodium gradient to expel calcium. With intracellular sodium raised, that gradient is smaller and the exchanger works less well.
- Calcium therefore accumulates inside the cell — and more available calcium means more forceful contraction.
Rate rises → diastole shortens → Na⁺ accumulates → Na⁺/Ca²⁺ exchange falters → intracellular Ca²⁺ rises → contractility rises
Digoxin produces the same end result by a different route: it blocks the Na⁺/K⁺-ATPase directly, raising intracellular sodium and, through the same exchanger, intracellular calcium.
The sympathetic lusitropic effect described earlier adds a second protection: by speeding relaxation it shortens systole, leaving a little more of the shortened cycle for filling. Tachycardia driven by sympathetic activity is therefore better tolerated than tachycardia of the same rate driven by an arrhythmia.
Faster rate, less filling time
Filling only happens in diastole, and diastole is what gets shortened when the rate climbs.
- Up to around 140 bpm, cardiac output rises as heart rate rises. Filling time is still adequate, and the Bowditch effect is helping.
- Above about 150 bpm, diastolic filling time falls to roughly 0.15 seconds. End-diastolic volume falls because the ventricle simply can’t fill, so stroke volume falls faster than rate is rising.
- Cardiac output therefore falls, despite the higher rate.
This is a key anaesthetic point: tachycardia attacks the myocardium from both sides at once. Coronary perfusion of the left ventricle happens almost entirely during diastole, so a shortened diastole cuts oxygen supply. At the same time the faster, more forceful heart consumes more oxygen. Supply falls, demand rises, and in a patient with coronary disease that gap is where ischaemia appears. Ischaemic myocardium contracts poorly, which lowers stroke volume further and lowers cardiac output further — a self-worsening loop, and the reason a fast ventricular tachycardia becomes a peri-arrest rhythm.
Rate rises above the ceiling → diastole shortens → filling falls and coronary perfusion falls → stroke volume falls → cardiac output falls → ischaemia → contractility falls
Slower rate — how far stroke volume can compensate
Going the other way, a slow heart has a long diastole, so the ventricle fills more completely. A bigger end-diastolic volume stretches the muscle further and, by the Frank–Starling mechanism, produces a bigger stroke volume. Up to a point, output is defended.
The compensation has a limit, because the ventricle can only fill so far and the pericardium restrains it. Once stroke volume has maxed out, any further fall in rate is a straight fall in cardiac output. That’s the physiology behind the bradycardia algorithm’s adverse features — shock, syncope, myocardial ischaemia, heart failure. They are the clinical signs of a cardiac output that the slow rate can no longer sustain.
Which lever does the body actually use?
Both, but not equally. Heart rate can treble — 60 to 180 bpm is an ordinary physiological range. Stroke volume typically rises by only about 50%, say 70 mL to 105 mL. Rate has far more room to move, so under normal circumstances heart rate is the dominant influence on cardiac output.
That asymmetry explains several familiar situations:
- Neonates and infants combine an underdeveloped sympathetic system with a well-formed vagal supply, so they drop their rate easily. With rate the dominant determinant, a bradycardia in a small child is a low-output state until it’s corrected.
- Severe aortic stenosis (valve area below 1.0 cm², vs normal of 2.6–3.5 cm²) fixes stroke volume mechanically. When systemic vascular resistance falls — on exertion, after a spinal, at induction — the usual compensation is to raise cardiac output. That patient can’t, so mean arterial pressure falls unopposed. Exertional syncope in aortic stenosis and the devastating hypotension after a central neuraxial block are the same failure.
- Pregnancy raises resting cardiac output by up to 50% at term, through both rate and stroke volume.
In summary
- Cardiac output is the volume of blood the heart ejects per minute, and it has exactly two components: CO = HR × SV. Roughly 5 L/min at rest; 3.0–3.5 L/min/m² when indexed to body surface area.
- Stroke volume is end-diastolic volume minus end-systolic volume, and it’s set by preload, afterload and contractility.
- The heart can only eject what reaches it, so venous return — driven by mean systemic filling pressure minus right atrial pressure, over the resistance to venous return — sets the ceiling on stroke volume.
- Heart rate is set by the phase 4 slope of the sinoatrial node. Sympathetic activity steepens it; vagal activity flattens it. At rest, vagal tone dominates and holds the rate well below the node’s intrinsic rate.
- Rate and stroke volume interact. Faster rates raise contractility (Bowditch) but eat diastolic filling time; output peaks around 140 bpm and falls above roughly 150 bpm.
- Rate has the greater range, so it’s usually the dominant determinant — which is why patients who can’t move their rate, or can’t move their stroke volume, tolerate haemodynamic insults so badly.