Wednesday, 11 October 2017

School Time Safety Netting

Safety netting is one of the most important interventions in acute paediatrics, whether you work in Primary or Secondary Care.  When we give safety netting advice, it is usually in the context of a medical problem that we expect to run a benign course and then resolve.  This is plan A.  Safety netting advice allows us to inform the patient (or responsible carer) when to activate plan B.

Plan B might be needed for a number of reasons:

The principles behind good safety netting are simple.


All of these factors are key to the success of the safety netting process.  Leaflets can be a very useful supplement to explanation and discussion.  The opportunity to ask questions is also essential.  Most importantly the parent should feel empowered to make an assessment and to return without feeling that they will be seen as over-anxious.

Recently, there was a lively discussion online (sparked by Damian Roland's writings about safety netting) about this subject, including some great tips about how to do it well and ways that it can go wrong.  One of the pitfalls that were mentioned was the possibility that discussion about the appropriateness of the attendance might get in the way of the touchy-feely aspect of the safety netting.  The gist was that if you discussed when not to come, this would be a barrier to appropriate attendance.  While I agree that this can happen, I don't believe that by discouraging attendance we are running this risk, as long as it is all done in a positive way.

The idea that discussing appropriate attendance is inherently negative is based on a false assumption: namely that the parent wants to come to the doctor.  Even where healthcare is free at the point of delivery there are many, many reasons why people do not want to see a doctor, with inconvenience being one of the most common.

Another issue is that the anxiety associated with the perceived need for a medical assessment is itself an unpleasant experience.  I think that doctors under-appreciate this because we are made to feel like the hero of the hour.  Someone was worried about a symptom and now we are the person to tell them that everything is almost certainly going to be fine.  Go us!

Probably the most common example that I can think of is the way that parents often think that a cough and fever equals a chest infection.  This is sometimes compounded by the belief that chest infection is often fatal - a rare outcome in a healthy child who is given appropriate treatment.

When I hear someone say that they have come to see me because they believe that their child has a chest infection and the child promptly runs off to play with the toys, I could be forgiven for having a 'why me?' moment.  I could use this opportunity to explain why this is an inappropriate attendance since the child is so very well.  That would be a tad self indulgent since I'm not the one with the worry.  Instead, I should make sure that my consultation finishes on a positive note, with good safety netting advice that encourages re-attendance, right?

Well, I think that the two things (good safety netting and discouraging unnecessary attendances) are far from mutually exclusive.

If anything, the two things work together in beautiful harmony and create the opportunity to take safety netting to a platinum standard.
What might this look like in practice?  For the child with the not-a-chest-infection, my school time safety netting might go like this:


It is absolutely important that parents do not feel criticised.  If fear of criticism leads to a child not being brought for assessment when needed, that is of course a bad thing.  So, we have to have the best of intentions when we talk about when and when not to seek medical advice.

I have no interest in doing myself out of a job by reducing attendances.  If anything, reducing avoidable attendances.   It takes me 2 minutes and zero stress to assess the child who has had a minor bump to the head.  If the parents who bring their child to see me when all that's happened is that they fell over and cried for a few seconds stop coming "because it's always best to get checked isn't it doctor?" then I'll just have to see more patients that take time and challenge my thought processes.  So why do I take the time to explain to every single parent the things that would constitute a reason to seek assessment for the next bumped head?

The answer is that it is in the best interests of the parent and child, so why wouldn't I?

Edward Snelson
Founder of MediLeaks
@sailordoctor

Disclaimer:  Exploring health beliefs can be hazardous.  Always wear a helmet.


I would love it if you would post the things that you educate parents about (in a positive way).  You could do that here, on Twitter or on Facebook.  If you post a comment here, don't worry if it doesn't appear straight away.  I have to check all comments before they are published.  (There are a lot of spammers out there!)

Monday, 2 October 2017

It's Not Easy Being Wheezy - about antibiotics and wheezy kids

In my formative postgaduate years as a doctor, I was told by more than one mentor that antibiotics were a good treatment for children who were wheezing. This advice was given bay various people at different times and whether this was bronchiolitis, viral wheeze* or an exacerbation of asthma, the principles seemed to be the same. The logic is sound - we know that infection triggers all three, and we can never be certain of the infection being a straightforward viral episode.  That was how it was put to me anyway.
*If you are unsure about the difference between viral wheeze and bronchilitis, follow this link for an explanation.

I no longer believe in this strategy as a treatment option for wheezy kids, and this is why:

It's not easy being wheezy.  Children with tight lower airways are up against it but often cope extremely well with their bronchospasm or their bronchiolitis.  I am constantly delighted by the ability of these children to be cheerful despite quite significant breathing difficulty.


It's even worse to have a bacterial lower respiratory tract infection.  As well as the breathing difficulty that comes from the loss of functioning lung, there is the tiring effect of the illness.  Having pneumonia is unpleasant and often exhausting.  It would be unusual to see a child who was cheerful and well despite a bacterial lower respiratory tract infection.




Now imagine combining the two.  Doesn't look good does it?  Children with bronchiolitis and viral wheeze cope with the difficulties of wet or constricted airways because they are systemically well.  Add the lethargy of bacterial infection to this and you go from a child who can compensate to one who cannot.  In short, you won't think "maybe there's a bacterial LRTI as well as the bronchiolitis or viral wheeze."  You'll know it.

Evidence from research backs this up.  The Cochrane review of antibiotics for bronchiolitis concluded that there was no benefit from antibiotics. (1)

You could say, what's the harm in trying antibiotics?  There are many reasons why unnecessary antibiotics might be harmful and none of these are to do with drug resistance.

So, it's time to do away with the idea that antibiotics have a role in treating well children with bronchiolitis or viral wheeze.  I believe that you'll know the children who need antibiotics because they will be properly unwell.

Edward Snelson
Antibiotic Guardian of the Galaxy
@sailordoctor

Disclaimer:  Secretly we all know that antibiotics do treat viruses, but if you tell anyone, you'll be removed from the Magic Circle.

Reference
  1. Farley R et al, Antibiotics for bronchiolitis in children under two years of age, Cochrane database of systematic reviews. 2010

Wednesday, 20 September 2017

The Cheater's Guide to Complex Paediatric Problems


If ill children worry clinicians, then the child with a complex medical background must be terrifying.  Paediatrics is full of syndromes, most of which are so rare that you will probably never encounter them, but these syndromes number so many that the total number ends up being significant.  The end result is that if you work in General Practice or an Emergency Department, you are guaranteed to see a child with an underlying syndrome every so often.

These syndromes are not the only reasons for children to have complex medical histories.  Sometimes a single disease is enough to dramatically change how a child presents and responds to illness.  Intuitively, these children make all of us anxious, and I include paediatricians in that collective of worriers.

So what should you do when faced with a child who comes to you and has a complex medical history?  Most of us would like to practice medicine well within our area of expertise and this scenario is going to be way outside of that?  Well, it doesn't have to be a case of complete panic and run.  No, when faced with unfair odds, if you choose to engage, you just need to know how to balance the playing field.  My approach is simple.  I cheat.

Here's a clinical scenario to illustrate what I am talking about:

A three year old girl presents with a history of pulling her ear. She has had a raised temperature but this has responded well to antipyretic medicine.  Examination is normal apart from a runny nose and a left sided inflamed tympanic membrane.

You see from the child's medical records that she has Aicardi syndrome and has regular followup with a paediatrician for this.


I'm guessing that this all felt safe for the first bit.  There is something particularly unnerving however about finding that a child has a thing that you've never even heard of.  This is a protective instinct that clinicians have (quite rightly) to let them know that they are outside of their area of expertise.

However, you do have options here.

You should of course look at the child's medical records.  This will often give you the diagnosis but not the detail about what the specifics of the thing, how the thing affects this child, or what the plan is for a child with the thing when they become unwell.

You could ask the internet for information about Aicardi syndrome.  Clinicians often feel that this would make them look foolish or lacking in knowledge.  The truth is that the parents do not expect you to have the first clue about their child's rare condition.  If you ask for two minutes to get a bit of information about the thing, they are most likely just relieved that you are not planning to ignore the fact that their child is different.  If anything, they will be pleased that you are taking the time.

The next option is to ask the parents what the deal is with their child's condition.  This approach works really well, assuming that the parents are reasonably in the know.  While the internet can give you an overview of a syndrome, no disorder ever affects two individuals in the same way.  For this reason, I find it useful to ask the following questions:


Note that this is cheating at the highest level.  You can't just expect to be this good straight away, but it's something I feel everyone should be aiming for.

What I've done there is to find out about the underlying medical problem, exactly how it affects this child in front of me and to find out what is expected of me.  This immediately gets me past a few of the most common barriers:
  • Possible gaps in knowledge of the condition
  • The way that a disease/ syndrome manifests is variable
  • There may be specific requirements or expectations when a child with this underlying problem becomes unwell
  • It is very difficult to spot an unwell child if you don't know how they look and behave when they are well.
  • The gut feel that we use when assessing children is less reliable if a child has a complex medical problem.
The reason for going to all of this is all of this trouble is to do with the way that children with complex underlying medical problems develop and respond to illness.


These children may or may not have good immune systems.  What they all have in common is that they will not respond to illness in the same way that a child who has no underlying medical problems would.  The reasons for this may be physiological, neurological or even due to human factors.  These children with complex medical backgrounds experience a normalisation of illness that other children do not.  As a result, they do not necessarily even know how sick they are themselves.

Putting a little bit of extra work into the assessment of these children does make a big difference.  It maximises the chances that the will get seen by the right person, in the right place, at the right time.  Sometimes there are specific things that can be done early on, which will help the child to get better.

Sometimes, being admitted or referred is not the best thing for the child or the parent. I the case of the child above, it turns out that Aicardi Syndrome doesn't affect immunity and that this child is not one of those severely affected.  The end result is that Mum is confident that she can recognise when her child is seriously unwell.  Today she feels that the problem is pain rather than concern about being dangerously ill.

More importantly, Mum has a physiotherapy appointment in an hour, so a referral 'just in case' is going to make her miss that.  After applying the full list of cheats, you decide to manage the child's illness yourself.

Epilogue:  The following week, you find a card and a box of chocolates on your desk.  They are thanking you for taking the time to manage this girl's illness rather than refer to the local hospital, which is what often happens to her.

Edward Snelson
Finding the easy way since 1996
@sailordoctor

Disclaimer: Cheating is only wrong if you get caught.


If you have your own tips and tricks for looking after children with complex medical problems, please post something in the comments box below.


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

Meningitis
Osteomyelitis
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
@sailordoctor

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

Reference

  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
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.


Thursday, 20 July 2017

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



Thursday, 29 June 2017

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.






Tuesday, 13 June 2017

Think Sepsis - What does that mean?

You may have noticed that there are a lot more paediatric sepsis guidelines flying around these days.  When people write guidelines, they are trying to be helpful but it is always worth knowing why they decided to be helpful.  For example, guidelines for diagnosis of asthma in childhood are strongly motivated by the desire to reduce the overdiagnosis of asthma in childhood.  We know this because the guideline writers tell us.  They're quite good like that.

So what about sepsis?  Well there are two genuine problems that keep coming up around sepsis:
  • Early diagnosis of sepsis
  • Early and aggressive treatment of sepsis
There's probably more to it than that but that is the main thrust of what most sepsis guidelines are trying to achieve.

I think that the guidelines that have come out over the past few years have done a good job in guiding our management of sepsis.  Once you have decided a child has enough evidence of being septic to be treated, crack on and don't spare the horses.  There is no doubt that as a profession, we are getting our act together in this respect.

The first part is more tricky. diagnosing sepsis is difficult.  Sorry, let's be honest, it is really, really difficult.  Sepsis is missed all the time, and I am not talking about the overdiagnosis of missed sepsis which goes like this:

There is a two part truth which guideline writers and readers need to accept.  Sepsis is often missed because it is often easy to miss it.

So, back to the guideline writing - in order to help us diagnoses sepsis, guidelines have been written to help us to recognise sepsis.  As a colleague of mine recently pointed out, that only works if you know to look at the sepsis guideline.  If you are already looking at the sepsis guideline then the battle is already won, because if you are worried enough to look at the sepsis guideline, it's usually time to phone a friend.

So why is it easy to miss sepsis?  There are several reasons;
  • The diagnosis of sepsis is subjective.  There is no mathematical equation (Fever + Tachycardia ≠ Sepsis), test or even definition that gives anyone the answer to the question does this child have sepsis. 2016 saw the third meeting of the Society of Critical Care Medicine and the European Society of Intensive Care Medicine in an attempt to achieve a consensus definition of sepsis.  They will need to meet up again in 2018 if they are going to find true consensus about what sepsis is and what it looks like, since they couldn't quite decide the first three times.  Perhaps they just miss each other that much between conferences.
  • Sepsis doesn't appear, it develops.  There is a reason that we don't have an issue with clinicians missing a diagnosis of croup.  Croup announces its arrival most helpfully.  If only sepsis did the same.  
  • There is almost always another diagnosis to distract the clinician.  As mentioned above, before a child is diagnosable with sepsis, they usually have a prodromal illness.  A classic example of this is secondary sepsis in children with chickenpox.  It is completely understandable that when a child presents with fever and being miserable, having the typical chickenpox rash seems to make the diagnosis obvious.  However, some of these children have sepsis, and it is important to know when that is a strong possibility.
  • All of the features of sepsis are non-specific and can belong to another diagnosis.  Features such as tachycardia are frequently difficult to interpret as a fast heart rate may be due to pain, fear or pyrexia - all of which occur in children who do not have sepsis.  In any case, there is no definition of tachycardia, so we rely on guideline figures.  
  • You can't rely on any one feature to be present all of the time.  Even pyrexia may be absent.  Recognising severe sepsis is relatively easy, but we are being asked to recognise sepsis earlier, before it becomes severe.  That is much more of a challenge.

So, in summary, early sepsis is vague and it is easy to miss because it often hides behind a more obvious diagnosis.  The result is that guidelines are inherently too vague or too prescriptive when it comes to the recognition of sepsis.  Furthermore, if sepsis is not considered, the guideline is of no use whatsoever.

There is relatively simplistic way to deal with all of this.  Since the challenges are mainly about awareness and decision making, I think that a sepsis guideline could simply look like this:

Here are some footnotes on this:
  • Recognising the unwell child starts with recognising the well child.  All sorts of factors are taken into consideration.  While many guidelines emphasise physiological values (heart rate etc.), the behaviour and activity of a child are very important.  The gut feel assessment of the parent and the clinician are also valid.
  • The trajectory of the illness is not always treated with the importance that it deserves.  Children who are not septic often have periods of being subdued but then pick up and have a time where they look and behave as though they are much improved.  This is the "I can't believe how well my child looks now doctor!" effect.
  • Recognising sepsis comes with experience but any clinician can think about the possibility of sepsis.  If you are unsure, get a further assessment.

So, thinking about sepsis is the crucial first step.  It's the deciding that takes the most skill.  Then, when it comes to acting, we should be getting on with with doing whatever we need to do without unnecessary delay.  Hopefully that has made it sound a lot more simple than it really is.

Edward Snelson
Simple is what I need
@sailordoctor

Disclaimer - Simplification is a huge cop out for medical writers, but it's also a lot of fun.  Try it sometime.


Thursday, 18 May 2017

Ctrl F and Child Mental Health Problems - Making everything simple

Recently, someone showed me something that has changed my life.  If you press the Ctrl and F keys on your computer (Command +F for Apple) at the same time, a magic box pops up.  This is the 'find' function.  If you don't use it much, you should.  It makes a lot of things much easier.

Here's a non-medical example:  Let's say that you are looking for a payment that you know you made sometime 2-4 years ago from your bank account.  (I don't know, maybe the insurance company want to know how much you bought your laptop for.  Just go with the example please.)  You download the last five years of bank statement from the website and start trawling through for the money that you know that you paid to Amazon for the laptop.  Well stop that.  There is a much easier way.  Press the Ctrl and F keys and then type 'Amazon' in the box.  Hit enter and watch the magic begin.  Using this witchcraft you can find what you are looking for instantly.

Here's a medical example of how I use this function all the time.  Go to the NICE guideline for gastroenteritis in children.  Download the full guideline, not the summary.  Now read it until you find the evidence statement for how the guideline group formulated its decision regarding use of loperamide.  No, don't do that.  The document is over 200 pages long.  Instead use Ctrl and F to start your search, then move on using the arrows (or 'next') until you are where you want to be.

This little trick works for word documents, spreadsheets and anything else.  My favourite trick is to use it on a webpage to find something that I can’t see (like unsubscribe).  Since I was shown how to do this, it has made so many things much easier.  What is amazing to me is that not everyone knows about it.

I don't know what it is like in the rest of the world, but trying to help a child with a mental health problem in the UK can be a lot like trying to find something in a 200 page document.  Primary care clinicians can put a lot of work into trying to help children and young people (CYP) with mental health problems and it can feel like we never get anywhere.  Recently, a child psychiatrist told me a few things that helped to make a lot of sense of these problems and how to help CYP with them, including how to get your referral to the Child and Adolescent Mental Health Services (CAMHS) to get the most appropriate response.

What he told me all made perfect sense.  So I thought you might like to have me share his beautiful and simple insights into child and adolescent mental health problems.

1 - There are usually three factors which lead to children and young people's mental health problems

One of these factors is the child's genetic predisposition.  You can't do anything about that but it is still useful information.

The next factor is the child's environment.  Note that the weight of the domains of a child's environment change as they grow.  For example, the importance of different domains for a 4 year old might look like this:

Then as a child becomes more independent, the importance of each domains changes.
Of course, this is a gross oversimplification, which is exactly what I need in these kinds of circumstances.  By the time you reach adolescence, I suspect the weight that each domain holds over the young person varies greatly, but what the future holds (health, wealth and success or lack of) starts to become much more significant.


The third factor is a trigger.  This brings the intrinsic into contact with the extrinsic factors, precipitating the mental health problem.

2 – Every Child needs an anchor

Children and young people usually have at least one functional and dependable adult in their lives who they can rely on to give them consistency and who will make the CYP feel that they are worthwhile individuals.  A child who never has one of these people in their lives is unlikely to escape mental health disorders.  A child who loses their anchor is at high risk of developing a problem.   Ask, “Who is the most important person in this child’s life?”  If they used to rely on their grandmother who has recently died, this is very important information.

3 – Children and Young People get different mental health problems at different ages

It’s fairly obvious to say that but it does help you when it comes to assessing a problem.  When we are deciding whether something is a mental health problem in the first place, our first question should be, “Is what is happening normal for this age group?”

So what problems do CYP get at different stages?  They get mental health disorders which fit their stage of psychosocial development.  Young children tend to get behavioural problems and neurodevelopmental disorders (oppositional defiance disorder, attention deficit disorder, separation anxiety disorder).  Older children get problems that are related to their transition from child to adolescent (anxiety and self harming).   The top end of the CYP age group (in the UK this goes up to 18 years old) will get the beginnings of adult mental health disorders.

Knowing that something is abnormal doesn’t tell you how significant the problem is.  What tell you the answer to this is the same thing that almost always tells you about how significant a problem is in paediatrics: function.  So, the next question is, “How does this problem impact on the child’s ability to do the things that they want to do or should be able to do?”

And there you have it: your child psychiatry equivalent of the ‘find’ function.  A little understanding goes a long way when it comes to assessing and referring CYP with mental health problems.  Knowing what to ask always brings the best answers.  The answers to these questions just happen to be what a CAMHS consultant needs to see in a referral letter.  By including all of this information, we maximise that consultant’s ability to prioritise you patient.  Sound’s good eh?  So here they are again:


OK, so it’s hardly a keystroke, but considering that we are talking about one of the most complicated problems that we will see in our work, a five question model for getting what you need is pretty good going.

Edward Snelson
Impulsive clinician with a short attention span
@sailordoctor

Acknowledgement - Huge thanks to Girish Vaidya (@DrGirishPsych) who has helped me to understand the core principles of child mental health.  His ability to make the complex simple is a real gift.


Wednesday, 3 May 2017

If you don't like what you hear, change the tune. (What to do when you don't hear a wheeze in a child who should be wheezy)

Every now and then, a clinician will go to see a child who appears to have increased work of breathing and is well (in the way that children with viral induced wheeze usually are) but find no wheeze on auscultation.  How strange!

Making a diagnosis is a complex business.  It is such a complicated process that, most of the time, we don't really think about it at all.  This has been described as type 1 thinking (intuitive) in the context of clinical diagnosis. (1)  Most days, I do very little thinking.  When I come across something unexpected or unfamiliar, I am forced to come out of this unconscious automatic mode and think carefully and consciously (type 2 process) about what is going on.  I have to engage my cortex, and it hurts.

So, going back to this little scenario.  Let's say that the child is 2 years old and has has a cold for 3 days.  They are brought by his parents because they have noticed that his breathing is a little fast.  He looks well and is really quite happy with the toys in your room.  Snot bubbles from his nose as he comes and sits on his mother's knee.  Looking at his chest, he has mild subcostal recession and a mild tachypnoea.  When you listen to his chest, you hear...  breathing.

So, what most people do in this circumstance is to listen some more.  It is traditional to check your stethoscope for gremlins or signs of tampering before pushing the earbuds a bit harder into your ears before listening again.  However, there is no getting away from the fact that there is no wheeze, nor is this the silent chest that is so feared in asthma and viral wheeze.  In a silent chest, no breath sounds can be heard and the patient looks awful.  This child has breath sounds and looks well.

How strange.

The reason that it causes us to have a confused moment is that there are certain combinations of signs and symptoms that indicate a particular illness.  That is particularly useful in children's respiratory illnesses since no one sign or symptom is likely to be specific to an illness.  This is why cough and fever do not equal a lower respiratory tract infection.  We need to look for the presence or absence of other features to form a likely diagnosis.


So what we've got now is a mystery illness.  What causes a well child to have respiratory distress without a wheeze or stridor?
Your next move is simple.  Give the child inhaled (or nebulised if necessary) β-agonists.  I would go with 10 puffs of salbutamol via a spacer device.  Then sit back and watch the magic.

What will probably happen next is a little surprising the first time you experience it: a wheeze appears.  More importantly, the child's breathing improves.  So, what is going on here?

The answer to that would be science.  Science and music are happening and it goes like this:  In order to have a wheeze, there must be the correct conditions for this to occur.  A musical note needs the right amount of air flowing through a tube in the right sort of way.  The size of the tube matters quite a lot.  Ask any wind musician or organ player.

In these cases there is bronchospasm (caused by viral infection) but the conditions are not right to produce a wheeze for you to hear.  Of the parameters that affect the musical note (length of tube, diameter of tube and flow of air) you can change two with β-agonists.  You can't change the length of the tube but the other factors should respond nicely.

So, if you don't like what you hear, change the tune.  When your clinical diagnostic brain tells you that there should be a wheeze, you are probably correct.  If you were expecting a wheeze but don't hear one, by all means rethink your presumption.  If you are left with the same conclusion, then try the β-agonist trick.  It works a treat.

Edward Snelson
Soverynotamusician
@sailordoctor


Disclaimer: This is a very different thing from rechecking a blood pressure until you get the number that you want.  Very different.   Anyway, I would never do that.

Reference
  1. Croskerry P, A universal model of diagnostic reasoning, Acad Med. 2009 Aug;84(8):1022-8.


Wednesday, 12 April 2017

Henoch Shonlein Purpura - Who, What, Where and When?

In Paediatric Emergency Medicine in the UK, it is often the case that a referral requires me to choose between more than one specialty.  For example cuts on the face (the ones that require suturing under general anaesthetic) could be referred to either plastic surgeons or maxillofacial surgeons.  Hand fractures are sometimes looked after by the orhopaedic surgeons (note use of their Sunday name) and sometimes by plastics.  Now that I think about it, the overlap almost always involves plastic surgery.

The decision about who to refer a patient to will depend on various factors.  These include resources, availability and the particulars of the case.  Matching the child and the illness (or injury) to the right person to follow the case up is part of the art of Primary Care.  After much thought and consideration, I've decided to refer children with Henoch Schonlein Purpura (HSP) to the very best.  I refer them to their General Practitioner.  
(Ankle bracelet not included.  See in store for more details)

The reason for this is that General Practice does two things better than any other specialty.  General Practice excels at avoiding the harm done by unnecessary tests and treatments, and (in my opinion) no other specialty in medicine is so good at providing the continuity needed to monitor a condition.  This allows the patient to receive appropriate care while at the same time having an experienced clinician involved who will intervene if needed.  For uncomplicated HSP, that is exactly what is needed.

HSP is the most common vasculitis in children.  Nobody really knows why children get it but the number one suspect is of course 'recent infection'.  These are always getting the blame for mysterious childhood illnesses.  If you ask me, I think it is institutional infectionism.

HSP is best known for causing a purpuric rash on the buttocks and lower limbs,  While this is the typical rash, atypical rashes are pretty common.  The thing that is most consistent about the HSP rash is the following triad:

HSP is a clinical diagnosis.  It is usually identifiable through a typical presentation and because other possible causes of non-blanching rash can be ruled out clinically.  Sepsis, leukaemia and immune thrombocytopaenia are some of the possibilities that should be considered if the presentation is not that of typical HSP, or if the child is unwell, or anaemic or has splenomegally (etc. etc.).

It is also worth mentioning that not all children have the courtesy to present with their symptoms in the correct order.  Everyone know that the rash should come first.  Some children choose to ignore this and complain of aches and pains before the rash appears.  Also, before the rash becomes purpuric it can start out as non-specific erythema or even look a little urticarial.

I am often asked if blood tests are needed for a child when diagnosing HSP.  My answer is this:  
For example, I don't do CRPs to diagnose sepsis.  I use a CRP once I have diagnosed sepsis because that will help in the ongoing management.  In injury, I use X-rays where it is probable that the X-ray will alter my management.  I don't use CXR to diagnose pneumonia.  I use CXR if I suspect complicated or atypical pneumonia.  In HSP diagnosis, tests are really for uncertainty in those cases where the presentation is ambiguous or atypical.  If the presentation is unambiguously HSP, no diagnostic tests are needed.  There are tests to be done but those are screening for complications.

Most cases of HSP follow a benign course.  The rash persists for several weeks and there many children experience aches and pains in their legs and abdomen.  However, symptoms other than the rash will typically settle in the first week or couple of weeks.  Things don't always go this way though.  There are four possible ways for the course of HSP to be problematic.

Problem 1) Extreme symptoms.

Pain is a funny thing and in some cases of HSP, despite a lack of significant complications, some children and young people get severe pain (despite good doses of simple analgesia) in their legs or abdomen.  These children should be referred acutely to be seen by a paediatrician.  This is both to enable management of the pain and to look for more significant complications.

Problem 2) Non-renal complications of HSP

I've said that leg and abdominal pains are fairly common symptoms of HSP.  It is also true that everything in paediatrics has an evil twin.  In the case of leg pains, it is possible for HSP to cause a significant polyarthritis.  If a child has swollen joints, the pain is severe or they develop a significant limp - refer.

Similarly, if a child with HSP shows signs of severe abdominal pain, consider the possibility that they have developed one of the most significant complications of HSP - intussusception.  Other signs might include pale episodes and blood or 'redcurrant jelly' in the stools.  With or without these other signs, assessment by a paediatric surgeon is urgently needed if abdominal pain becomes acutely severe.

Almost anything that can go wrong can go wrong in HSP.  Involvement of other organs than kidneys, skin, bowel and joints is rare.  Atypical symptoms such as headache or chest pain should be taken very seriously and the child should be referred acutely.

At some point, all of these factors will lead to a decision about where the child is managed.  I think that the decision usually looks like this:
There is nothing mandatory about referring a clear cut case of HSP at first presentation.  However, most GPs that I know would prefer an initial assessment in secondary care, even if the child is extremely well and has no symptoms with their non-blanching rash.

Problem 3) Renal complications

While aches and rashes are what cause children and parents to lose sleep over HSP, it is the possibility of renal complications that usually worries the clinician.   A significant chunk of children who have HSP will develop a mild nephritis.  A smaller number go onto get a significant nephritis and a small number develop severe, kidney-threatening nephropathy.  Don't worry too much about the numbers as they don't really help in a case series of one.  What is true is that it is very likely that the child in front of you will either never develop a renal problem, and if urine dip flags something up, it is most likely to be a transient mild haematuria or proteinuria.  In order to identify those cases of silent but significant nephritis, monitoring is needed for all cases of HSP.

Exactly how much monitoring is needed is slightly unclear.  Advice is based on clinical experience, common sense and a bit of logic.  If you put those things together, you might come up with this:
This is followup strategy is based on the well child, who has no significant haematuria or proteinuria at any point.

What this aims to do is detect renal involvement early, while avoiding any OCD type behaviour (such as daily urine dipsticks).  The length of follow-up is debatable but is based on a cohort study that showed that no child with HSP developed renal involvement after 6 months.  Roughly half the guidelines from big centres recommend follow up to 6 months and half say 12 months.

There are those that advocate longer followup (including lifelong annual reviews) for all cases.  However, I suspect that there is a significant effect of bias driving that.  If a child has had a complicated case of HSP, this may well be warranted, but if the course of the illness was benign I doubt that they will develop late complications.  The Nottingham renal unit recommends lifelong monitoring if everything else has been normal but there was at least one occasion where the urine dip showed ++ or more of protein. (1) This stratification allows for the possibility of nephropathy in selected cases rather than ongoing screening for all.

The need for a blood pressure check after the first couple of visits is also debatable (thus the *).  By continuing to check BP after it has been normal on two occasions, are we saying that the child could have had renal involvement severe enough to cause hypertension, yet this never showed up as a positive dipstick?  This seems very unlikely to me and I have never found a paediatrician who has seen such a thing happen.  Despite this, all guidelines and review articles that I have seen recommend that BP is checked at every visit.

Problem 4 - Prolonged or relapsing course of illness

This brings me neatly onto the final thing that might cause a clinician to feel that the child needs more investigation and management than can be given in Primary Care.  In some cases, the course of HSP does not follow the typical one of a few weeks of rash and a few days of other symptoms.  The symptoms might continue for months, or relapse after initial remission.  A mild relapse is actually quite a common occurrence in HSP and worth warning the family about.  If a relapse is severe or prolonged I would suggest that these children also need to be referred.  The paediatrician will need to review the diagnosis (in case the reason for an atypical course is that the diagnosis is wrong) and consider whether to investigate e.g. for an autoimmune cause of the rash.

HSP FAQs

Q. Does the child need blood tests at presentation?

A. Only if there is diagnostic ambiguity or if the urine dip/ BP is abnormal.


Q. Do steroids prevent complications?

A. No, but they may have a role in treating complications of HSP. (2) Other interventions that have found to be unhelpful include antiplatelet treatment or immunosupression.


Q. What counts as hypertension in a child?

A. This is difficult question to give a straight answer to.  It's not a lack of people coming up with definitions that is the issue.  It's more that we don't know much about what levels of hypertension cause what effect in the long term.  Most guidelines go for something like systolic BP >95th centile on three occasions.  It is important to use the correct size of BP cuff for the size of the child and to use an appropriate reference range. (3)


Q. So should I be checking BP at every visit even if the urine has been normal (no more than a trace of blood and no more than + protein) on every test?

A. If you didn't, you would be going against all the guidelines in existence.  I still haven't had a good explanation as to why it is really necessary, but nor have I convinced others that it is unnecessary, so for now you should keep on doing it.


So in summary, managing HSP is relatively straightforward.  After initial assessment, most can be monitored by their GP.  Some will need to stay under the care of a paediatrician and a small number will need to be referred on, usually to a nephrologist.  For the uncomplicated cases, there is no absolute certainty about often or how long to screen for renal involvement, so guidelines have to be mostly based on pragmatic expert opinion.

So, who, what, where and when?  Me, you and maybe them doing as much nothing as possible until we are sure that everything is OK.  Simple really.

Edward Snelson
Dogmatologist
@sailordoctor


I would like to express my deepest gratitude to the young person who consented to have photos of her HSP rash used for this purpose.  Thank you.


References
  1. Henoch-Schönlein Purpura (HSP) Guideline, Nottingham Children's Hospital
  2. W Chartapisak et al., Prevention and treatment of renal disease in Henoch-Schönlein purpura: a systematic review, ADC, Feb 2009
  3. Blood Pressure Levels for Boys by Age and Height Percentile, National Heart, Lung and Blood Institute