Monday, 21 August 2017

Everything Wrong With Infantile Colic

One thing that I have enjoyed from time to time is a series of YouTube videos called "Everything wrong with..."   Here is an example of someone slickly dismantling the 1999 version of the film "The Mummy" which gives you an idea of how this works.  Egyoptologists will particularly enjoy this.

There are many things in paediatrics which lend themselves to a similar style of critique, none perhaps more than colic.  Infantile colic is diagnosed in countless numbers of babies every day.  Advice is given and treatments sometimes prescribed.  All of it is nonsense from start to finish, so here is my "Everything wrong with Infantile Colic."

1 - Colic is not a diagnosis

The commonly touted diagnostic criteria for colic is, "Infantile colic is a benign process in which an infant has paroxysms of inconsolable crying for more than three hours per day, more than three days per week, for longer than three weeks" (1)

What you have there is a definition of a lot of crying.  The implication that this excessive crying equals a diagnosis is utterly ridiculous.  Babies cry for all sorts of reasons.  Also, why is this particular pattern (3x3x3) of any significance compared to any other pattern?  What if the baby is crying for 2 hrs and 45 mins, 7 days per week for 2 weeks and 5 days?  Shall we tell the parent to come back in two days time and suggest that the child steps up to the plate because they're just not bringing their A-game?

If this was any other measurable value, we would be plotting it on a centile chart and monitoring the trend.  There is a reason that we don't have centile charts for the normal amount of crying in babies, and if you don't know what is normal, how can you define what is abnormal?

In any case, this is simply an impractical definition.  How much a child is crying and how much it seems like they are crying (to the parent) are likely to be two very different things.  Why have such a precise definition when the information given will be subjective?  I'm not suggesting that we do try to get precise figures, but maybe there is an app for that.

2 - Colic is not a pathology

Pathology is the study of the cause of disease.  Firstly, it is debatable as to whether colic is a disease, mainly because we don't know what it is.  Secondly, when we give a label of colic, we are saying that we have excluded pathology.  There are plenty of reasons why a baby might be crying excessively and these need to be considered.

Gastro-oesophageal reflux disease
Non-IgE allergy (usually CMPA)
Urinary tract infection
Non-accidental injury

Hair tourniquet
Injury (accidental and non-accidental)

Many possibilities have been suggested as causes for infantile colic.  No specific cause has been found that would explain why some otherwise healthy babies seem to cry excessively in the first few weeks of life.  What that leaves us with is supposition.  It is often said that the crying is due to the sensations that a baby is experiencing in their abdomen.  Perhaps they are getting used to normal peristalsis of the bowel (which does not occur in utero).  Nobody knows.

3 - Colic is not treatable

As mentioned before, there have been several propositions made as to the pathology of colic.  None have been substantiated but that doesn't stop the door from having been opened to treatments being offered.  "It might be low-grade lactose intolerance..." - thus one of the popular colic treatments having lactase as the active ingredient.  "It might be bowel gas causing discomfort..." - thus another containing dimethicone.  I have no problems with a  process of hypothesis and experimentation.  This is how we get evidence for effective treatments.  What is then needed is good evidence that a treatment that might (in theory) work, is genuinely effective.  In my opinion, there is no good evidence that any effective treatments available for infantile colic.  Of all the patient groups where we should avoid the use of ineffective medication, surely this is top of the "don't do it" list?

Bear in mind that to prove efficacy in a condition such as colic, the method would have to be a large double blinded randomised controlled trial.  If not, the effect of the natural resolution of symptoms would make any treatment appear to be effective.  For example, if we suggested that the family plant some potatoes and that this will treat the colic, by the time the potatoes are ready to harvest, the condition will have resolved.  How impressed will they be?  D'ya know something?  I  think I might write a book and create a web page about that.  There's money to be made here.

While offering a treatment might seem both benign and helpful, I feel that it is neither of these things.  Previous treatments for colic have been withdrawn from use due to cardiac side effects.  A drug might be used for many years before the harmful effects are known.  Pessimistic possibilities aside, my main issue with any ineffective treatment is that it can make things harder for the parents.  When given a treatment (or suggested remedy), there is an expectation of symptom resolution which simply does not materialise.  This is psychologically quite hard for a parent who is likely to feel quite desperate.

Instead, I prefer to be honest and tell the parents the truth.

The last point in that list is most important.  Having a baby that cries a lot has a huge effect on parents and can potentially be a factor in parental mental health problems.  Also, the frustrations caused by a baby that cries and cries have been implicated in safeguarding cases.  While we can't take away the exhaustion or the stress, we do have the ability to absolve the parents of the feeling that they must always stop their baby from crying.  I tell parents that if the baby has been cleaned, clothed, fed and loved, then there are times when it is necessary to walk away and leave the child to cry.  If possible, family support should be made use of.  Well meaning relatives need to understand that colic is harmless to babies in the long term, but harmful to parents in the short term.

Edward Snelson
Potato therapist

Disclaimer - Potato therapy requires years of training.  Do not attempt to practice potato therapy without the appropriate qualifications.


  1. Infantile Colic: Recognition and Treatment, Johnson J, Cocker K, DO; Chang E, Am Fam Physician. 2015 Oct 1;92(7):577-582

Tuesday, 8 August 2017

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

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.