Physiological basis of heart sounds and murmurs

There are unique problems associated with auscultating these subtle noises in the chest of a mechanically ventilated patient. How can you hear anything in there, one might ask, what with the wind howling through the ventilator tubing, and the rhythmic doof-doof of the balloon pump? I suppose one can make one's life rather easier by turning the ventilator rate to 1 per minute, turning the PEEP down to zero, and pausing the pump for a few seconds.

Anyway. The college places an unfair amount of emphasis on physical examination. The following past paper questions have involved heart sounds, murmurs, and cardiovascular examination findings in a broader sense:

Cardiovascular Examination Findings in CICM SAQs
  • Question 30.3 from the second paper of 2018 - AR severity (4 signs)
  • Question 16 from the first paper of 2016 - AR severity, indications for AVR, causes of systolic murmur.
  • Question 14.2 from the first paper of 2013 – Aortic regurgitation vs. mitral stenosis
  • Question 26.3 from the second paper of 2011 – 4 causes of a diastolic apical murmur
  • Question 5.3 from the second paper of 2010 -  double peak arterial pulse
  • Question 5.4 from the second paper of 2010 – clinical featurs of Tetralogy of Fallot
  • Question 3.2 from the first paper of 2010 - causes of a diastolic apical murmur
  • Question 12.1 from the second paper of 2009 – causes of an irregularly irregular pulse
  • Question 24.2 from the first paper of 2009 – features of pulmonary hypertension
  • Question 24.6 from the first paper of 2009- causes of a diastolic apical murmur
  • Question 7.1 from the second paper of 2008 – signs of ASD, VSD and PDA
  • Question 13.1 from the first paper of 2008 – features of severe aortic stenosis
  • Question 27 from the first paper of 2008 – causes of a pan-systolic murmur

From this table, one may observe certain trends:

  • The college's love for apical diastolic murmurs extends over 3 identical SAQs
  • Praecordial auscultation findings are the main focus of most of these questions
  • Discriminating between murmurs on the basis of physical examination findings and dynamic manoeuvers seems to occupy an eminent position

Without further ado, one should at least know a little about the following physical signs.

The four heart sounds

First heart sound - S1

This is the sound made by the mitral and tricuspid valves closing.

It indicates the beginning of systole. It is a loud, low-pitched LUB sort of sound.

Loud S1

This is the result of mitral valve leaflets remaining open at the end of diastole, rather than drifting back into a closed position as diastolic flow slows down. Such a wide-open mitral valve will slam shut loudly with the onset of systole, making the first heart sound louder than usual.

Conditions which produce slow or incomplete diastolic filling will cause a loud S1.

Some examples:

  • Tachycardia (ie. not enough time to fill the LV in diastole)
  • Mitral stenosis (enough time, but diastolic filling though the stenosed valve is too slow)
  • Short atrioventricular conduction time (ie. diastolic filling cut short by early systole)

Soft S1

Well, its the reverse of the loud S1, and the opposite conditions cause it to develop.

Anything that slows the diastole allows the mitral valve to be almost completely closed by the time systole begins. For example:

  • First degree heart block (ie. ie. diastolic filling unnaturally prolonged by delayed systole)
  • Left bundle branch block (ie. ie. diastolic filling unnaturally prolonged by slow conduction)
  • Mitral regurgitation (the mitral leaflets never actually meet, so there is no mitral component in S1)

Splitting of the first heart sound

  • Right bundle branch block can produce a split first heart sound - because the contraction of the right ventricle is delayed- the conduction occurs via the left ventricle rather than the bundle of His- and thefore the closure of the tricuspid valve occurs after a substantial delay.
  • Atrial septal defect can result in a fixed split of the first heart sound

Second heart sound - S2

This is the sound made by the aortic and pulmonary valves closing.

It indicates the end of systole. It is a quiet, high-pitched DUB sort of sound.

It is composed of A2 and P2, the aortic ad the pulmonary components, which correspond with the closure of the aortic and pulmonary valves.

Loud A2

  • Systemic hypertension places massive back-pressure on the aortic valve, causing it to slam shut with an unusually large amount of force. Systemic hypertension is the main cause of a loud A2.

Soft A2

  • Aortic stenosis (or rather, sclerosis) prevents aortic valve leaflets from moving normally - so they are unable to slam shut with a loud bang at the end of systole.
  • Aortic regurgitation produces a soft A2 beause the valve leaflets never come into any sort of physical contact with one another, at least not enough to make any sort of "slamming" sound.

Loud P2

  • Pulmonary hypertension - In the same way as systemic hypertension causes a loud A2, a loud P2 results from pulmonary hypertension because pulmonary arterial pressure slams the valve shut at the end of systole.

Splitting of the second heart sound

It is normal for this sound to be split. The high pressure in the systemic circulation slams the aortic valve shut rather abruptly, almost angrily. In contrast, low pressure of the pulmonary circulation tends to close the pulmonary valve gently, and therefore the pulmonary component of the second heart sound (P2) is usually delayed by about 20-30 milliseconds.

It is also normal for increased right ventricular filling to cause a widening of the split. The more blood in the RV, the longer it takes to eject, and therefore the greater the delay until pulmonary valve closure.

In the spontaneously breathing patient, the delay is greatest during inspiration. Naturally, in the patient ventilated with positive pressure the delay is greatest during expiration (positive pressure being a barrier to diastolic filling)

Increased normal splittng of S2

Anything that delays the end of right ventricular systole can cause this sort of picture.

  • Right bundle branch block - the delay in conduction via the left ventricle causes a delay in right ventricular contraction, and therefore a delay in pulmonary valve closure. The S1 will also be split.
  • Ventricular septal defect - because the right ventricle receives a large volume load directly from the left ventricle, and therefore takes longer to complete its systolic contraction.
  • Pulmonary valve stenosis - because the right ventricle takes longer to empty though a narrowed valve
  • Mitral regurgitation- not because right ventricular contraction is delayed, but because left ventricular contraction is shortened (as the LV empties in both the aortic and the atrial directuion, systole is over very quickly).

Fixed splitting of S2

  • Atrial septal defect - the atria, joined by a gaping hole in their seput, act as one atrium. The result is a reasonably equal distribution in volume betweent the right and left atrium. This way, both sides of the circulation share the same diastolic filling pressure. Dragging more volume into the right atrium with respiratory activity will not cause an inequality of ventricular filling (between the right and left ventricles) because the venous return will be "shared".

Reversed splitting of S2

In this situation, P2 occurs before A2, and splitting widens during expiration (or inspiration in the mechanically ventilated patient). This only happens if the conduction to the left ventricle is delayed, or if the left ventricle is massively volume overload (and the right ventricle is not).

  • Left bundle branch block - the left ventricle depolarises after the right ventricle, and A2 is delayed
  • Aortic stenosis - the left ventricle empties slowly though a narrow valve
  • Large patent ductus arteriosus - the left ventricle receives a backflow of blood from the aorta, which causes it to become volume-overloaded

Physiological S3

The S3 is a low-pitched mid-diastolic sound. It is generated by tightening of papillary muscles in the venricles and is associated with very rapid diastolic filling. The physiological S3 typically occurs in children and young fit people.

A pathological S3 is typically associated with a decreased ventricular compliance and increased atrial pressure, which could occur in either the right side or the left.

In fact, all the additional heart sounds are vaguely related to decreased ventricular compliance.

Left ventricular S3

  • Thyrotoxicosis - due to an increased cardiac output
  • Pregnancy - also due to an increased cardiac output
  • Left ventricular dilatation and LV failure - a big dilated LV is a non-compliant LV
  • Aortic regurgitation, mitral regurgitation, ventricular septal defect and patent ductus arteriosus are all listed in Talley and O'Connor as causes of S3, but I expect these conditions achieve an S3 by causing severe LV failure.

Right ventricular S3

  • Right ventricular failure - when the right ventricle is dilated and non-compliant
  • Constrictive pericarditis - when the right ventricle may be normal, but the compliance is decreased by the brutal mechanical facts of a fibrosed pericardial wall

The fourth heart sound - S4

The S4 is a sound generated by the atria, and therefore absent in atrial fibrillation. It results from the reflection of an atrial pressure wave off the walls of a stiff non-compliant ventricle. Again, it can be due to right ventricular or left ventricular pathology.

Right ventricular S4

  • Pulmonary valve stenosis - uncommon and usually a consequence of congenital heart disease
  • Pulmonary hypertension - and the resulting RV wall hypertrophy, with a decrease in RV compliance

Left ventricular S4

Talley and O'Connor list a series of conditions associated with decreased left ventricular compliance, which sounds very similar to the list of conditions which cause the left ventricular S3. I will regurgitate the list here, and add some extra material.

  • Aortic stenosis - this tends to result in LV hypertrophy, which in turn causes decreased LV compliance
  • Acute mitral regurgitation - for some reason, only acute MR causes this, after chordae rupture
  • Systemic hypertension - again as the result of a stiff hypertrophied LV wall
  • Advanced age - a fibrotic old ventricle, crusty with age
  • Ischaemic heart disease - a fibrotic ventricle
  • Myocardial infarction - apparently, the LV wall
  • Hypeterophic obstructive cardiomyopathy - again, the LV wall is stiff and non-compliant

Unusual sounds


Systolic ejection click

This is a high-pitched sound occurring after S1, which one would expect to hear somewhere around the aortic or pulmonary auscultation window. Essentially, it is caused by congential stenotic disease of these valves. A congentially stenosed aortic valve is not calcified, it remains elastic and mobile - it is simply too small for its purpose. Consequently, when the hypertrophied left ventricle pushes against it, the valve "domes" into the aorta, and this produces the click.

  • Congenital aortic stenosis
  • Congenital pulmonary stenosis

Non-systolic ejection click

Again, this is a high-pitched sound, but it occurs at some stage in the cardiac cycle which is not associated with systeole. It could happen after S1 or after S2, doest matter where. Itsn origin is the useless flailing of a prolapse valve leaflet, or (even more rarely) the lonely wiggle of randomly misplaced tricuspid valve leaflets in Ebsteins anomaly. The leaflets, as it occasionally comes into contact with other structures, produces all sorts of random noises, and these are all characterised as a "non-systolic ejection click".

  • Mitral regurgitation - with leaflet prolapse
  • Atrial septal defect - which makes this sound by opening
  • Ebsteins anomaly - with misplaced tricuspid valve leaflets

Opening Snap

This is a weird high-pitched sound occuring after S2. It happens when the stenotic mitral valve suddenly opens after a period of isovolumetric relaxation. Typically, it is followed by the diastolic murmur of mitral stenosis.

  • Mitral stenosis - but obviously, analogous pathology of the tricuspid valve could also give rise to this sound; of course with tricuspid stenosis being vanishingly rare, one merely needs to know about this possible pathology, without necessarily expecting to ever witness it.
  • Tricuspid stenosis - erm, as above.

Tumour plop

Though perhaps then coolest of all heart sounds, the pendunculated atrial myxoma plop is rare even among the population of genuine myxoma sufferers - apparently only 10% of them have this finding. Its the sound made by the tumour when it blocks the mitral valve in diastole. The first description of this as a plop comes from a 1960s paper which presents a case history of a 42-year old housewife, who had a 7cm atrial tumour.

  • Pedunculated atrial tumour - sometimes.

Diastolic pericardial knock

Constrictive pericarditis can apparently result in a situation where the filling ventricle abruptly stops filling (when it meets the resistance of the fibrotic pericardium). This abrupt cessation of filling is called the "diastolic plateau" and patients with constrictive pericarditis who dont have a diastolic knock dont seem to have this plateau.

  • Constrictive pericardial disease - is what the textbooks say, and I am somewhat perplexed as to how this differs from the S3 (which is also due to decreased ventricular compliance, and can be associated with pericardial disease). Apparently, it is higher in pitch, and arrives somewhat earlier than the S3.

In addition to the abovementioned, there are a whole symphony of weird sounds produced by artificial heart valves and pacemakers

Pan-systolic murmurs


These tend to be the result of a high pressure chamber draining into a low pressure chamber. For instance, the murmur of mitral regurgitation is a pan-systolic murmur. Non-standard plumbing of the greater vessels tends to also produce such sounds when one of the two vessels is inder higher pressure.

  • Mitral regurgitation
  • Tricuspid regurgitation
  • Ventricular septal defect
  • Aortopulmonary shunts - eg. the aforementioned Blalock-Taussig shunts.

Ejection systolic murmurs

These tend to involve a ventricle ejecting blood through some sort of stenotic orifice, be it a conventional valve or some sort of catastrophically narrow outflow tract. They follow a classical crescendo-decrescendo pattern.

  • Aortic stenosis
  • Pulmonary stenosis
  • Atrial septal defect
  • Hypetrophic cardiomyopathy - with LVOT obstruction, and turbulent flow through this narrowed tract

Late systolic murmurs

These are the outcome of some sort of regurgitant valve, leading backwards out of a ventricle. So by that lame definition, they are represented by only tricuspid and mitral regurgitation. Specifically, the mitral valve needs to be actually prolapsed in order to generate a late systolic murmur. Or, the papillary muscles need to be somehow damaged.

  • Mitral valve prolapse: This one typically commences with a click
  • Pulmonary stenosis: this may be present in the company of an S4, and there may be features of right heart failure. Its a late ejection systolic murmur which does not radiate to the carotids.

Early diastolic murmurs


Aortic regurgitation

This tends to occur together with aortic stenosis.

Blood regurgitating back through the aortic valve forms a high-pitched early diastolic murmur, best heard at the 3rd and 4th intercostal spaces.

There may be an Austin Flint murmur in late diastole.

Features of severity in chronic aortic regurgitation

Question 16 from the first paper of 2016 asked for five clinical signs of severity, whereas Question 30.3 from the second paper of 2018 only asked for four (and gave more marks for them).  UpToDate gives the following list of features associated with chronic AR:

Signs of widened pulse pressure:

  • These were mentioned in Question 14.2 from the first paper of 2013
    • Corrigans sign: a "jerky" carotid pulse: full expansion, followed by complete collapse. You're palpating the pressure of the left ventricle, essentially. It's named after a 19th century Irishman. It indicates a severe aortic incompetence.
    • de Musset's sign which the college answer has spelled incorrectly is  a visible nodding of the head in time with arterial pulsation in patients with severe aortic insufficiency. It is named after an aortically insufficient French poet.
    • Quincke's sign, otherwise known as Quincke's pulse, is a nail sign: it is seen when the nailbed is blanched. The pale nail bed flashed red and white as capillary refill is restored. It can also be seen in the absence of any aortic problems, in patients who have sclerodactily.
    • Duroziez's sign is elicited by listening over the femoral artery with the bell of the stethoscope. It is supposed to be a double murmur. According to some recent evidence, it has almost 100% specificity. There is supposed to be both a systolic and a diastolic bruit, as blood rushes into - and then rapidly out of - the femoral artery.
  • These are mentioned in UpToDate:
    • Traube's sign – A pistol shot pulse (systolic and diastolic sounds) heard over the femoral arteries.
    • Mueller's sign – Systolic pulsations of the uvula.
    • Becker's sign – Visible pulsations of the retinal arteries and pupils.
    • Hill's sign – Popliteal cuff systolic pressure exceeding brachial pressure by more than 20 mmHg with patient in the recumbent position.
    • Mayne's sign – More than a 15 mmHg decrease in diastolic blood pressure with arm elevation from the value obtained with the arm in the standard position.
    • Rosenbach's sign – Systolic pulsations of the liver.
    • Gerhard's sign – Systolic pulsations of the spleen.

Chacteristic auscultatory findings:

  • Soft S1
  • Soft A2
  • An S3 if LV function is severely depressed
  • A systolic ejection sound due to abrupt aortic distension

Indications for aortic valve replacement, as given in the 2014 AHA/ACC guidelines, are as follows:

  • Symptomatic patients:
    • When the AR is Stage D, i.e with a dilated LV, Doppler jet width ≥65% of LVOT and holodiastolic flow reversal in the proximal abdominal aorta
  • Asymptomatic patients:
    • LVEF <50%
    • Normal LVEF, but with significant LV dilatation (end-systolic diameter > 50mm or end-diastolic diameter >65mm)
    • None of the above, but about to undergo cardiac surgery anyway (for some other reason)

Pulmonary regurgitation

Blood regurgitating back through the pulmonic valve forms an early diastolic murmur along the left sternal border. They also call it the Graham Steell murmur. Generally speaking, unlike TR, this murmur is not associated with a presystolic pulsation of the liver.

Mid-diastolic murmurs


Mitral stenosis

This tends to be associated with a loud S1 and an opening snap.

It is best heard in the left lateral position, with the bell of the sthethoscope. It is louder on expiration.

Tricuspid stenosis

This is best heard at the left sternal edge. It is louder on inspiration.

Atrial myxoma

The tumour makes various sounds, but usually one expects to hear a diastolic murmur as the blood finds its way around the mass in diastole.

Carey Coombs diastolic murmur

Named after Dr C.F. Coombs from Bristol, it is a "short mid-diastolic rumble" associated with rheumatic mitral valve disease, and which disappears with resolving valvular disease.

Austin Flint diastolic murmur

This one is named after Austin Flint, whose monumental contribution one can still find scanned into the cavernous archives of Google. Its a murmur of the aortic regurgitation jet hitting the apex of the left ventricle in diastole. Flint said of the murmur,

"“Oftener rough than soft. The roughness is often peculiar. It is a blubbering sound, resembling that produced by throwing the lips or the tongue into vibration with the breath of respiration.” 

This murmur can be distinguished from mitral stenosis by the absence of a loud S1 and the absence of an opening snap.

Presystolic murmurs

Mitral stenosis, tricuspid stenosis, atrial myxoma.

I shant repeat myself.

Continuous murmurs

Patent ductus arteriosus

The persisitng connection between the pulmonary artery and the aorta is a noisy one. This is usually a constant "machinery" murmur, maximal at the first left intercostal space.

Aortopulmonary connection (eg. Blalock-Tausig shunt)

Similarly to the PDA, but of unnatural origin, this is a communication between the pulmonary circulation and the aorta.


AV fistula

This is oe mumrur you might not hear maximally at the praecordium! The fistula itself will be where the sound is maximal.


Venous Hum

This a benign phenomenon.

Ruptured Sinus of Valsalva, emptying into the RV or RA

This is freakishly rare. The sinuses of Valsalva are the anatomical dilatations of the ascending aorta, which either give rise to the coronary arteries, or (posteriorly) are "blind". If one ruptures, it is bad news. The sound it makes will probably be heard best aroiund the parasternal areas.

"Mammary Souffle"

In 1960, Tabatznik Randall and Hersch proudly examined two hundred and fifty seven South African Bantu women and foudn this murmur in about 15% of them. It is though to originate from the turbulent flow around the sites of anastomosis between the internal mammary arteries and the intercostal arteries.

Discriminating between different systolic murmurs

The following table may be found in the SAQ discussion notes , where it is used to answer Question 16 from the first paper of 2016. Specifically, that question asked for five different causes of a systolic murmur, and their "auscultatory characteristics". 

Cause Auscultatory characteristics
Tricuspid regurgitation
  • Right of sternum, or left sternal edge
  • Louder on inspiration
  • Does not radiate to the carotids
Aortic stenosis
  • Radiates to the carotids
  • Louder on expiration
  • Quieter with isometric hand grip
  • Quieter with Valsalva
Mitral regurgitation
  • Loud S3
  • Soft or absent S1
  • Maximal at apex
  • Radiates to axilla
  • Pan-systolic
  • Louder with isometric hand grip
  • Quieter with Valsalva
Atrial septal defect
  • Fixed split P2
  • Louder on inspiration
  • Loudest at left sternal edge
  • No click
  • S4 is present
  • Quieter with isometric hand grip
  • Louder with Valsalva

Discriminating between different diastolic murmurs

The following table may be found in the SAQ discussion notes , where it is used to answer Question 14.2 from the first paper of 2013.

Features which Distinguish Aortic Regurgitation from Mitral Stenosis
Feature Aortic regurgitation Mitral stenosis
Rhythm usually sinus usually AF
First heart sound Normal S1 Loud S1
Second heart sound Soft S2 (A2) Normal S2 (M2)
Additional sounds Third heart sound (S3) Opening snap
Pulse quality Collapsing pulse Small pulse pressure
Where is it loudest 3rd intercostal space, parasternally Apex
When is it loudest Sitting forward, during expiration In a left lateral position
Pitch, quality Decrescendo low-pitched, rumbing
Associated findings

Signs of LV failure,

displaced apex beat

"Malar flush"
Coexisting disease Infective endocarditis, ankylosing spondylitis, other seronegative arthropathies, Marfans. Pulmonary hypertension

For whatever reason, the college frequently ask for four causes of a diastolic murmur loudest at the apex. This has come up three times:

Those causes are:

  • Mitral stenosis
  • Severe mitral regurgitation (flow murmur)
  • Significant left to right shunt (VSD)
  • Austin-Flint murmur of aortic regurgitation
  • Carey-Coombs murmur

Characteristic findings in Tetralogy of Fallot

The college has also previously asked about the characteristic auscultation findings in tetralogy of Fallot (Question 5.4 from the second paper of 2010). Those findings are:

•     An ejection systolic murmur: either pulmonic stenosis or RVOT obstruction
•     Right ventricular heave
•     A loud single second sound


Clinical Examination of the Critically Ill Patient, 3rd edition by L.I.G. Worthley - which can be ordered from our college here.

Clinical Examination: whatever edition, by Talley and O'Connor. Can be acquired any damn where.

Nicholas Joseph Talley, Simon O'Connor; Clinical Examination: A Systematic Guide to Physical Diagnosis (7th ed)

SEGAL, JACK P., W. PROCTOR HARVEY, and MICHAEL A. CORRADO. "The Austin Flint murmur: its differentiation from the murmur of rheumatic mitral stenosis." Circulation 18.5 (1958): 1025-1033.

Flint, Austin. Compendium of Percussion & Auscultation, and of the Physical Diagnosis of Diseases Affecting the Lungs and Heart. W. Wood & Company, 1870.

PITT, AUBREY, et al. "Myxoma of the Left Atrium Hemodynamic and Phonocardiographic Consequences of Sudden Tumor Movement." Circulation36.3 (1967): 408-416. - where they quote the article which first came up with the "plop" description which was ABBOTT, 0. A., WARSHAWSKi, F. E., AND COBBS, B. W., JR.: Primary tumors and pseudotumors of the heart. Ann Surg 155: 855, 1962.

Tyberg, Theodore I., Allan VN Goodyer, and Rene A. Langou. "Genesis of pericardial knock in constrictive pericarditis." The American journal of cardiology46.4 (1980): 570-575.

Edwards, Edward A., and Harold D. Levine. "The murmur of peripheral arteriovenous fistula." New England Journal of Medicine 247.14 (1952): 502-505.

Koshy, George, et al. "Ruptured sinus of Valsalva." Saudi medical journal 23.8 (2002): 995-998.

Tabatznik, Bernard, THOMAS W. RANDALL, and Colin Hersch. "The mammary souffle of pregnancy and lactation." Circulation 22.6 (1960): 1069-1073.

Brest, Albert N., Vasant Udhoji, and William Likoff. "A re-evaluation of the Graham-Steell murmur." New England Journal of Medicine 263.24 (1960): 1229-1231.