Heart Murmurs in Children

Heart murmurs are a reasonably common finding in children.  Excluding the newborn and early infancy assessments, most of the murmurs that we hear are benign, physiological or flow murmurs.  All of those terms basically mean the same thing – the anatomy is normal, the child is healthy and yet there is an extra noise heard on auscultation of the heart.  The reason that this usually occurs is that children’s physiology responds vigorously to illness and stress.  The heart pumps faster and harder, often making a structurally normal heart into a noisy heart.

In short, heart murmurs are almost always nothing to worry about.  Almost.

The trouble is that everything in paediatrics has an evil twin.  So what are the possible more significant  pathologies and how can these be recognised amongst the far more common benign diagnoses?  When children present with an acute illness, a cardiac cause for their symptoms is not usually top of our list of differential diagnoses because respiratory and musculoskeletal cases are far more common reasons for chest pain or dyspnoea.  This coupled with the way that cardiac problems present (with vague symptoms easily attributable to more common illnesses) make these rare clinical scenarios into the stuff of our worst fears.


Let’s look at a few made up cases to illustrate how to approach the “Oooh-I’ve-heard-a-murmur-what-now?” scenario.

Case 1

A 3 year old child presents with a febrile illness.  Let’s say that they have an otitis media.  They have a heart rate at the top of the reference range for their age and you hear a murmur.  The murmur is soft, easily heard (but not loud), systolic and heard best at the left sternal edge.  There is no radiation.

What now?

In primary care, we are all about the focussed history and examination.  We have to be in order to make time for our other patients.  That ergonomic approach works well, but when we find something that we weren’t expecting, we need to go back and get more information.  In this case, we want to know if the child is known to have a heart murmur.  Have they had an echo done in the past?  We also need to feel the precordium for heaves and thrills and be happy that the pulses have a normal character and volume.  Make sure that the femoral pulses are palpable, with no brachio-femoral delay.  Check for hepatomegaly.  Most importantly, make sure that the child does not have increased work of breathing and does not seem unexpectedly unwell.  Ideally, we need to get a blood pressure checked.

What next?

We need to refer a child with a murmur acutely (usually to general paediatrics) if

  • they are more unwell than expected
  • they have respiratory symptoms which are otherwise unexplained
  • there are concerning clinical examination finding (e.g. hepatomegaly)

A well child with none of the above ‘red flags’ probably needs an outpatient echo if

  • The murmur is loud
  • The murmur is diastolic
  • The murmur radiates outside of the precordial area

If the child is well, there are no red flags and the murmur sounds benign (as in the original description for this case) then common practice is to follow the child up when they are well again.  The presumption is that this is a physiological murmur which has been heard because the illness is causing increased cardiac output and therefore turbulent blood flow.

It is likely that when they are seen, the murmur will no longer be heard.  In these circumstances: case closed.  If the murmur persists, they can be referred then (either for an echo or to paediatrics if they don’t have direct access to this), assuming that the child remains red-flag-free.

Case 2

A two month old baby presents with a runny nose, cough and a low grade temperature.  The parent has noticed that their baby has started to have slightly fast breathing and has not been feeding as well as they normally do.  On examination, the baby has a slight wheeze, mild recession and basically looks like the other babies with bronchiolitis that you have seen that month.  All except for one thing: they have a soft systolic murmur that you can just about hear over the wheeze.

What now?

In short, refer to the acute paediatric medical team.  This child might have bronchiolitis and a flow murmur, but there is every chance that they are just pretending to have bronchiolitis.  Babies who have a ventricular septal defect (VSD) might not be picked up on screening (newborn baby checks etc.) and may have no overt symptoms, until they get their first cold.  Then, shortly after becoming snotty, the illness tips them into heart failure.  What does heart failure look like in a baby?  Well, they have increased work of breathing, a bit of a wheeze and struggle to feed.  Does that sound like bronchiolitis or does that sound like bronchiolitis?

Other clues that give these mimics away are:

  • Excessive tachycardia
  • Hepatomegaly
  • More pale or mottled
  • The course of the illness is different – they continue to get worse, while bronchiolitis symptoms peak at day 3-4 of the illness.

Case 3

A fourteen year old presents a few days into a flu-like illness.  They have been seen by a couple of doctors in the past few days, and have been told that they have a viral illness.  On both occasions they were advised about symptomatic treatment.  The parents are concerned that the young person is not getting better.  On examination, there is a barely audible systolic murmur.

What now?

Don’t presume this to be a flow murmur.  Firstly, in the context of viral illness, flow murmurs are most commonly heard in younger children. Also, the the murmur was not heard on previous visits.   One possible diagnosis here is viral myocarditis.  Other possibilities include bacterial endocarditis.

What next?

Refer this child acutely to the paediatric medical team.


One general rule which works quite well for symptoms in children (in the context of heart problems) is this: An isolated sign or symptom is rarely indicative of significant pathology.  Two signs or symptoms is always worth taking very seriously.  Take chest pain in children.  I don’t expect chest pain to be due to a cardiac cause in children (because it almost never is).  I know that the vast majority of children who have a syncopal episode will be having straightforward vasovagal events.  However, chest pain in combination with syncope, dyspnoea or palpitations is presumed by me to be pathological until said pathology is excluded.


So, as with many presentations in children, everything is normal except when it isn’t.

Edward Snelson
Postcordiologist
@sailordoctor



Many thanks to Dr. Carrie Mackenzie (Consultant paediatrician) who helped with the original version of this piece which is on the excellent RCEM learning site.


The Everest-Lifeboat Test (Easter Egg – Investigating apparent faints in children)

Paediatric guidelines always have to take into account one of the most important maxims of Child Health - the test or treatment must always be in the best interest of the child.  This means a particular discernment about the value of a test or treatment.  Guidelines rarely discriminate between which treatments and tests are essential and which are simply recommended.  When writing a guideline, it is relatively easy to put in a recommendation, often giving the impression that it is a must-do when this is simply not the case.


One of the things that I am most often asked (whether at work or giving and educational session) is "Do I have to do X?  The guideline says to but..."  So how does one discriminate between the must-do's and the would-be-good-if we-could recommendations?  That's easy - apply the Everest-Lifeboat Test.

The Everest-Lifeboat Test is simply a two part thought exercise.  The first part is to ask the following question:  If the person involved was at an advanced stage of climbing Everest, should they turn back to get this test or treatment or could they reasonably continue to the summit?  This part of the test asks if we are doing something just because we can rather than because we have to.  We are purely focusing on the immediate need at this point.

The second part is the lifeboat question:  Imagine that this patient was in a lifeboat, drifting across the ocean.  When they are rescued several weeks later, would they most likely be fine despite their current clinical situation?  This part of the Everest-Lifeboat Test forces us to look forward and consider the medium and long term consequences of action versus inaction.

Let's try this on a fictional patient.

A 10 year old child is brought to you having had an unexplained collapse.  The history given to you is vague and there is no first hand account of the episode as it happened in front of classmates at school.  However, in your further history taking, you find that this child has been having headaches that are worse in the morning.  Also, teachers have noticed that there has been a deterioration in performance at school over the past four weeks.  The child is slightly ataxic and has nystagmus on examination.

Concerned, you discuss the need for further investigations.  In response, the parents say that they are happy to get the tests done but could it please wait a few weeks?  They were just coming to get a medical opinion before they catch a plane to go away for a couple of weeks for their holiday.  They were assuming that this was just a fainting episode and really only came to get their child checked because grandma told them to.

So, what do you think?  This seems like a clear fail of the Everest-Lifeboat Test to me.  I wouldn't be happy to put investigations on hold, despite the obvious inconvenience to the family.

How about this scenario?:

A 12 year old girl comes to be assessed.  She was in school and had been stood in the heat when she began to feel sweaty and nauseous.  Her vision went black and she slumped to the floor.  She was reported to look pale and floppy.  She was unresponsive for a few seconds and then came around slowly over a few minutes.  A few hours later, she feels fine.  There is no history of unexplained deaths in her family.  When you examine her, all is normal.

You check a relevant guideline and see that it recommends that you perform a 12 lead ECG. She flatly refuses to have this test done and will not be persuaded.  What should you do?

Applying the Everest-Lifeboat test would go like this:

Would you advise abandoning the final attempt on the summit?  Well, she had an obvious precipitant and prodrome for her apparent faint.  We can advise how to avoid precipitants and what to do if a prodrome is recognised.  The event appears to be a classical vasovagal syncope without red flags in the history or examination.  So, forcing the issue seems to be unnecessary.

Would a few weeks in a lifeboat be an issue?  This brings us back to the guidelines that recommend investigation.  What are they trying to protect us from?  Much of the practice of ECG screening comes from adult medicine, where pathology is much more likely.  In paediatrics, there are a few arrhythmias that we need to worry about, but a standard 12 lead ECG is not the perfect screening test that we might hope it is.  The sensitivity and specificity of 12 lead ECGs in children is poor (1).  Ask yourself why the guidelines don't say, "Don't bother with history and examination.  The ECG is the crucial bit of information."

A quick look at the guidelines gives some useful insights to help us with the Everest-Lifeboat Test.  The NICE guideline "Transient loss of consciousness (‘blackouts’) management in adults and young people" (2) actually only relates to the over 16 year-olds.  This in itself acknowledges that an adult approach cannot be extrapolated to the child who has had a collapse.

Then there is the European Society of Cardiology's guideline for the diagnosis and management of syncope (version 2009). (3) It does take the view that children and adults can be investigated similarly and recommends ECG for all children who have had a faint.  However, this recommendation seems to be based on the assumption that an ECG is clearly useful additional information and fails to consider the possibility that a thorough history and examination gets you to a point where and ECG would not add value.

The American College of Cardiology/American Heart Association guideline (4) does seem to consider this possibility.  There is a strong emphasis on the value of a good history and examination.  Regarding ECG, it points out that ECG is a simple and available test that might identify a tendency to arrhythmia.  However it also states: "Despite the benefit of identifying a likely cause or potential clue about the cause of syncope from the ECG, prospective studies did not conclude that ECG findings significantly affected subsequent management.  The prognostic value of an abnormal ECG in patients with syncope has been questioned as well."

So there it is.  An honest declaration that, while the experts would recommend that we all do a test, it is unclear what the value of the test is.

I know that it might seem as though I just want to avoid doing work here, but there are genuine risks with tests.  The first risk is that they stop us from thinking.  If the sensitivity and specificity of history and examination is excellent, while that of ECG is poor, why introduce a deceptive piece of information?  The second risk is that of getting information that I don't want.  If I do an ECG on a child, it is almost always to look at the rate, rhythm and QT interval.  While those things are usually fine, the diagnostic report usually sports a bit of LVH and right atrial enlargement.  Of course the child has neither of these things, but the machine is just trying to make sense of the voltage it has been given.  If I were to take these things seriously, I might cause unnecessary anxiety for the child and parents.

So, what does my patient really need?  I need to take a good history and establish that the episode that sounds like a faint truly sounds like a faint.  This means asking about the three P's of vasovagal syncope.
If it sounds like a classic faint, I still need to make sure that I consider my red flags.
If the history given is of a straightforward faint, without red flags, I think that allowing the child to refuse the ECG passes the Everest-Lifeboat Test.

When we are forced out of our normal process, it is a good time to evaluate our routine practice.  If a deviation from the norm passes the Everest-Lifeboat Test, I would question the norm.  You may just have discovered that you are doing a test or a treatment that you don't believe in.  Here is a little list of things that have passed the Everest-Lifeboat Test for me at various times in the past (i.e. I was going to treat, something got in the way of that and I went with the the no treatment option):
That's a short list of times when the option of doing nothing became the right thing despite what was routine practice.  In the case of umbilical granuloma, I am pleased to say that doing nothing is now becoming the norm.

I hope that you find the Everest-Lifeboat Test useful at some point.  As to whether every child who has had a faint needs an ECG, versus it is good to get one or it is simply not needed unless there is a specific reason - this is a debate that is lacking input from the good people of the primary care and emergency medicine communities.  My solution?  Stick a cardiologist, a paediatrician, an emergency medicine doctor and a general practitioner in a lifeboat and leave them there till they've sorted it out.  I'm fairly sure they'd be fine...

Edward Snelson
Precordiologist
@sailordoctor

Disclaimer: The Everest-Lifeboat Test was originally described in 1055 by Egbert the Uncertain, a monk who died at the Battle of Hastings before writing down his idea.  I therefore take full credit for inventing the test myself.


  1. Kapoor WN, Evaluation and outcome of patients with syncope, Medicine, 1990 May;69(3):160-75.
  2. NICE guideline "Transient loss of consciousness (‘blackouts’) management in adults and young people"
  3. Diagnosis and management of syncope, European Society of Cardiology, European Heart Journal (2009) 30, 2631–2671
  4. Guideline for the Evaluation andManagement of Patients With Syncope,  A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines, and the Heart Rhythm Society



The Truth about Lactose Intolerance

Sometimes, when I'm checking my facts, I'm surprised to find conflicting information.  Sometimes this is because there are no facts and this information void is filled with opinion or speculation.   Sometimes there are myths that confuse the picture and sometimes the information is being interpreted differently by different people.  Lactose intolerance in children is one paediatric condition that is described very differently in various places.  So what is the truth of it?

Let's start with the basics.  Lactose Intolerance (LI) and Cow's Milk Protein Allergy (CMPA) are very different problems but are often confused with each other.  The simple difference is that you can't be allergic to a sugar.  In CMPA there are two different types of immune response possible.  The one that usually gets confused with LI is the non-IgE CMPA because it causes similarly vague and non-specific symptoms.

Lactose intolerance is essentially a budgeting problem.  Breast milk and formula milk contain the sugar lactose.  This provides ready energy and makes the milk taste nice.  To break down lactose, you need an amount of lactase which is proportionate to the amount of milk you consume.  This lactase comes from two sources.  Breast milk contains some lactase, but the majority of lactase that is required is produced in the villi of the small bowel.  

The maths is simple.  If you have enough lactase to deal with the lactose that you consume, all is well.  If you don't, then undigested lactase passes through to the large bowel.  In the colon, bacteria ferment the sugar- causing diarrhoea, abdominal pain and bloating.  (Interestingly, this is also how lactulose works, despite what you may have been told - details about that here)

How lactose intolerance works is well established.  What is far more unclear is how common it is in children and what the underlying cause is when it is present.  In order to understand all of that, it's worth knowing the evolutionary history of lactose intolerance.  Since man has only been farming animals for the past few thousand years, the need to digest lactose into adulthood is a relatively recent phenomenon.  The default physiological norm was for lactase production to diminish and stop after infancy.  Most people now have a lactase persistence gene which allows the small bowel to continue to produce lactase.

There are three types of Lactose Intolerance (1)

1- Primary hypolactasia

This is caused by the absence of the gene that causes ongoing lactase production.  When lactase production wanes, symptoms will come on gradually.

2 - Secondary hypolactasia

This is caused by damage to the small bowel.  In many cases this is the result of a gastroenteritis infection.  Secondary hypolactasia can also be the result of damage caused by non-IgE food allergies.  There is nothing inherently wrong with the child's ability to make lactase, but the small bowel wall that does that has been stripped.  The symptoms of secondary hypolactasia resolve when the damage to the small bowel is repaired.  In the case of post gastroenteritis LI, a few weeks lactose free will help that considerably.

3 - Primary congenital alactasia

This is an autosomal recessive genetic defect resulting in the complete absence of lactase production.  This is extremely rare and according to the experts, there have been less than 100 case reported worldwide. (2)

So here's my problem - sometimes, none of the three known entities fit.  I see lots of parents who report symptoms which are being associated with the child's (usually formula) feed.  They are then absolutely convinced that a lactose free formula has suddenly resolved their infant's problem.  Discuss...

Of course the type of lactose intolerance only matters in terms of deciding whether the child will need long term or short term avoidance.  Secondary hypolactasia should resolve, leaving the child able to resume business as usual once the damaged small intestine has repaired itself.  In general terms, the treatment for LI is the same - remove lactose from the diet.  Nevertheless, the fact that children seem to fit the bill for lactose intolerance without fitting any of the known entities is perplexing.

(Screenshot taken from Wikipaedia on 27/06/2017)

One possible explanation for this is that these children do not have lactose intolerance at all.  Instead the diagnosis has been suspected based on non-specific symptoms, and the resolution of these symptoms following the start of a lactose free milk was coincidence.


As well as the children who have no pathology, there is the group who are labelled as lactose intolerant when what is really wrong with them is something else such as reflux disease (GORD) or Cow's Milk Protein Allergy (CMPA).  Wouldn't it be nice to be able to be sure which condition was affecting the child?

There are tests available for lactose intolerance but they are not readily available in primary care, whereas a therapeutic trial is always a tempting option. The tests used in secondary care seem to vary around the world.  Hydrogen breath testing is a pain free way of testing but impractical  in infants making it unhelpful in most cases.  A lactose tolerance test involves giving a large dose of lactose and then doing serial blood glucose measurements.  A lactose challenge does the same but relies on the witnessing of symptoms following the administration of lactose to the child.

In the UK, the vast majority of LI diagnoses are made on clinical suspicion and treated by an exclusion diet.  The difficulty with this is the risk of overdiagnosis when using this strategy.  Symptoms often change with time and so any apparent improvement following an exclusion of lactose from the diet should probably be challenged in every case.

So what is the truth about Lactose Intolerance?  Is is frequently being diagnosed in contradiction of sound medical understanding of the condition?  Is there a significant number of infants who don't fit into the three categories described above but still have genuine LI?

I think that the presumption should be that a baby who has significant symptoms (other than the classic LI symptoms following a gastroenteritis) should have another explanation sought.  They may have GORD or CMPA but it doesn't make sense that they should have lactose intolerance unless there is a form of this condition that is as yet not understood or well described.

Edward Snelson
Congenitally Tolerant Clinician
@sailordoctor

Hot tip - if a lactose intolerant baby is breast fed, there is no point in excluding milk from the mother's diet in an attempt to resolve this.  Breast milk produces its own lactose.  If the diagnosis is convincing, try supplementing the lactase.  If you want to know where to find it, lactase is the 'active ingredient' in some colic treatments.  It doesn't actually treat colic but that is another story.  At least it has a use here.



References
  1. Heyman B et al, Lactose Intolerance in Infants, Children, and Adolescents, Pediatrics, Vol 118, Number 3, p 1279-1286, September 2006
  2. Deng, Y et al, Lactose Intolerance in Adults: Biological Mechanism and Dietary Management, Nutrients. 2015 Sep; 7(9): 8020–8035.