Wednesday, 31 January 2018

A genuinely useful test - Blood sugar in acutely unwell children

When your patient is a child, the decision to do investigations is a complex one.  On one hand, it is often the case that at the end of history taking and examination there is some uncertainty which tempts the clinician to gather more information in the form of tests.  Parental anxiety might seem to be best managed through investigations either to demonstrate normality or to show that concerns have been taken seriously.

The reality is that these reasons for doing tests are all dubious and can achieve the opposite to the desired result.  There are few paediatric medical problems for which a test is more valid than clinical assessment.  For example, upper respiratory tract infections - If the symptoms are all well controlled by simple analgesia, what useful information could tests add?

Worse still, many tests are potentially misleading or might cause anxiety.  I like a full blood count as much as the next medic but goodness me don't they have a lot of information in them?  Does it mean something when the lymphocyte count is just below the reference range if the total white cell count is normal?  Sometimes the result can be alarming.  "Toxic lymphocytes" are a good example of something that you don't want to get on the lab report, especially when the child is climbing up the shelves in your consulting room while you try to work out how to answer the question, "Was the blood test result normal?"

But it's not all me, me, me.  The truly difficult part of all of this is that we should not use tests to reassure.  No matter how much another person wants a child to have a test for their own reassurance, we should only be doing tests which are in the best interest of the child.  Although a tough rule to live by, it is important to remember that the child is the patient, not the parent. 

The good news is that tests rarely tell you something of great value once a child has been clinically assessed.  Tests are most useful when they give you a number which has a clear clinical significance and where you could not have made a diagnosis without that test.  This criteria applies very nicely to testing blood sugar for an acutely unwell child, in certain curcumstances.  

Here are two cases where a simple skin prick glucose test has made all the difference:

Case 1

A three year old boy has been vomiting and febrile for 24 hours.  Clinically they are not dehydrated (they have wet mucous membranes, normal skin turgor, their eyes are not sunken and they are passing urine reasonably well) and they are neither septic nor meningitic.  In fact what is odd about them is that they seem subdued and grumpy despite having a normal heart rate and normal temperature.

Blood sugar is tested and found to be significantly low.

Why do unwell children get low blood sugar?  This happens for one of two reasons.  The most common reason is that the child has exhausted their available supplies of glucose.  The child's body has a limited reserve and if consumption exceeds supply then the child will eventually run out of glycogen, the most readily available way of mobilising energy in times of need.

The second possible reason is that the child has a metabolic problem.  Some children have a disorder in their metabolism which is normally masked by the almost constant supply of carbohydrate that they receive.  When this supply is interrupted, it unmasks the fact that one of the chemical pathways (needed in times of crisis) is faulty.  No plan B, no blood sugar.

How do children with low blood sugar look?  The brain is a complicated thing but it runs on two main things: oxygen and sugar (or backup substances but let's keep this simple).  If a child has a low blood glucose then the first signs are usually neurological.  Most commonly the child is not fully alert and is often unhappy.  I believe the term is hangry.  They will be less active but they may also be combative.

How should low blood sugar be treated?  The first thing to mention here is a pitfall.  Do not use glucagon to treat hypoglycaemia in children when the reason for their low blood sugar is that they have run out of reserves.  It simply won't work.

If the child is able to swallow safely, they can be given a sugary drink.  If not then what happens next depends on where you work.  If you are outside of a hospital setting, then a glucose gel rubbed into the oral mucosa will be a good plan.  In hospital, most places will give an intravenous bolus of dextrose 10%.  (ALSG recommends 2ml/kg) When the cannula goes in, there are usually some blood tests that need to be taken, depending on local guidelines.

The response to initial treatment should be assessed and further intervention given as needed.  If there is a good response, it is important to realise that this is not 'job done'.  To become hypoglycaemic, the child had to have exhausted their reserves, so a single bolus will have a very temporary effect.  It is essential to continue the carbohydrate supply, often in the form of maintenance intravenous fluids.

Case 2

A 5 year old girl presents with abdominal pain and vomiting.  They started being non-specifically unwell two days ago.  Yesterday they started having abdominal pains and overnight they started to vomit.  The child is drinking well and passing urine.

The child has no significant past medical history.  On examination the child is subdued and miserable.  The abdomen is soft.  The child is tachycardic but has a normal temperature.

When a blood glucose is checked, it reads three times the upper limit of normal.

Why do children get high blood sugar?  Although any significant illness can cause a moderately raised blood sugar level, hyperglycaemia of this magnitude always means diabetic ketoacidosis. (Well, that's not strictly true.  Occasionally it means that there is sugary residue from a drink or snack on the finger that was tested.  That's why you need to clean the finger first.)  Of course, it is possible (and preferable) to make the new diagnosis of type 1 diabetes before it gets as far as DKA.  If the blood sugar is high, type 1 diabetes should be the presumed diagnosis.

How does diabetic ketoacidosis (DKA) present?  DKA is a difficult diagnosis for all sorts of reasons:
  • Symptoms are somewhat non-specific.  Abdominal pain and vomiting are good examples.
  • DKA is rare and gastroenteritis is common
  • Clinicians tend to check that a child is drinking and passing urine enough, but may not ask if oral intake and urine output is excessive
  • The onset of symptoms can be surprisingly gradual.
  • While less common, type 1 diabetes can occur at very young ages.  Unsurprisingly, it is not the first diagnosis that people think of when seeing an unwell young child.

Most commonly, the illness is initially mistaken for a viral URTI or gastroenteritis.  Polyuria and polydipsia are unusual symptoms in a child with a viral illness who is significantly unwell and should be major clues.  Abnormal breathing is a feature of the acidosis and should also prompt the clinician to think of DKA.  However there is one feature that rather tidily applies to both high and low blood sugar: the child will be subdued and lethargic.

How should DKA or newly diagnosed type 1 diabetes be treated?

This is fairly specialised stuff so these children will all need to be referred.  The acute management of DKA is rather complicated and requires careful fluid management to reduce the risk of developing cerebral oedema.  According to the NICE guidelines (1), fluids should be the priority for the first 1-2 hrs after which insulin can be started for DKA.

As tests go, blood glucose is much more useful that many, and the result and meaning are usually clear.  Blood glucose is a test to go for in children who have atypical symptoms or are significantly subdued.  High or low, an abnormal result is a game changer and early recognition of the problem will make a huge difference.

Edward Snelson
Tricorder trainer

Disclaimer: While it would be wrong to do an unnecessary test on a child, it is perfectly acceptable to Feng Shui them. They then need careful monitoring to see the effect of the intervention.

  1. NICE guideline [NG18], Diabetes (type 1 and type 2) in children and young people: diagnosis and management

Thursday, 18 January 2018

Investigations in Bronchiolitis - Help a Brother Out

One of the things that a crisis should do is to bring people together.  As the UK's National Health Service experiences a winter that is bringing  it to it's knees, we should be more aware than ever that Primary and Secondary Care need to work together as a team so that we all come out the other end of this intact.  There are those in each camp that are so entrenched in the us-and-them mentality that this idea of teamwork is against  their very being.  If you think that General (family) Practitioners (GPs) are lazy idiots or you think that hospital doctors are arrogant and lack communication skills, I'm not sure that I've got anything to help you.  (Perhaps a hug?)  If you are willing to believe that we are all part of a team and that any divisions are both understandable and resolvable, read on.

The issues that cause division are (in my experience) usually due to misunderstanding and a lack of insight into the part that we play in creating these disparities.  The irritations that I cause my colleagues, whether in Primary or Secondary Care, tend to be completely unintentional.  Sometimes it can be what seems like the smallest thing and I have no idea of the impact it has on someone else.  That's why it's so important to talk to each other.

Let me describe an example: the seemingly simple act of requesting a chest X-ray for a baby with bronchiolitis.

Bronchiolitis is usually a set piece.  An infant, usually under 12 months old, gets inflammation in their airways during a viral respiratory tract infection.  This leads to a three to five day period of increasing symptoms that usually consist of cough, wheeze, difficulty feeding and increased work of breathing.  In practical terms, that might play out like this:

Monday - A three month old is snotty and coughing but well and feeding nicely.  Breathing will probably be normal at this stage.

Tuesday - The parents have now noticed some noisy breathing and that feeding isn't as good as it normally is.  like many people in these circumstances, they seek a medical opinion and so go to their GP.  They are seen and the diagnosis of bronchiolitis is explained.  The clinician advises frequent small feeds and gives safety-netting advice for what to look out for in case this child becomes one of the very small minority who develop breathing or feeding difficulties that require support as an inpatient.  In keeping with NICE guidelines (1), they do no tests and give no treatments.

Wednesday - The parents have noticed a slight worsening of the feeding and are unsure about whether this is bad enough to need something to be done.  They hold off doing anything during the day but when their baby refuses an evening feed they take their precious bundle to the local emergency department.  There the child gets a thorough assessment.  Clinically the infant is well hydrated and although they have wheeze and mild recession they are looking happy and have oxygen saturations of 95% in air. Despite the earlier feed refusal, they now take 100 mls of milk. The assessing clinician hears a few crackles in the chest and decides to do a chest X-ray, just to be sure that today's worsening symptoms are not due to a bacterial pneumonia developing.

When is it necessary to do a chest X-ray for babies with bronchiolitis?

There are many possible complications of bronchiolitis.  In theory, these should be apparent on a chest X-ray.  There are also conditions that mimic simple bronchiolitis, most notably congestive cardiac failure.  So, a chest X-ray seems a sensible thing to do to add value to the assessment.  Except that it doesn't, for two main reasons.

These are the reasons why a chest X-ray doesn't add value to the assessment of a typical presentation of bronchiolitis:

  • There is always something to be seen on the X-ray.  If you are looking for normal, you are unlikely to find it.
  • If the infant has bronchiolitis, that is enough explanation for their clinical condition.  If they had a complication such as lobar pneumonia as well, you would't be doing the chest X-ray to see if there is a problem, you'd be doing it because the child is seriously unwell.

Coming back to our little one with bronchiolitis, let's assume that the chest X-ray shows what it always shows in bronchiolitis - something of uncertain significance.  There will be a little fluffy shadow somewhere due to the mucous plugging and atelectasis that is part and parcel of bronchiolitis.  Often, there is something to be seen at the right heart border (or as a wise radiologist once redefined for me: "the area of radiological romance") which may or may not be something.  Once seen, it is very difficult not to feel the need to treat for possible bacterial infection.  This brings me on to the next few reasons for not doing a chest X-ray.

Doing a chest X-ray in bronchiolitis is associated with increased antibiotic prescribing rates.  Bacterial superinfection in bronchiolitis is rare and is usually clinically apparent.  Bronchiolitis is reason enough to have respiratory symptoms.  If a baby also develops pneumonia, they are usually quite unwell and have a marked degree of distress.  So, following that logic, for our baby the antibiotics are simply unnecessary, in which case they may do harm.  Since one of the problems that comes with bronchiolitis is feeding and hydration, why give these babies a gastric irritant?

Finally, there is a second victim here: the clinician who saw the baby on Tuesday.  Despite sterling work by the GP, the parents are likely to believe in retrospect that they were given a poor service.
All the evidence is there:

  • The person that they saw in the community did nothing and their child became worse.  
  • The experts in the hospital did a test which showed that their baby needed antibiotics.
  • Antibiotics were given and eventually the bronchiolitis got better. 

Why didn't the GP do something when we went the first time?

Of course both clinicians in this story were trying to do their best for the patient.  Neither of them had anything on their mind other than trying to ensure the best outcome for the child. There was no disrespect intended and no one wanted to cause anyone any problems.  We are all just trying to get through the day and to keep our patients safe and well.

Thankfully, it turns out that in bronchiolitis, keeping your patient safe and well is probably best done without the use of chest x-ray and antibiotics.  So, we in Secondary Care can help our colleagues in Primary Care out by making the assessment that needs to be made: in or out (of hospital)?  That decision is made on clinical grounds and if they need to be admitted, a chest X-ray may be useful.  Maybe.

Edward Snelson
Vice President of the Guild of Doing as Much Nothing as Possible

Disclaimer:  I did a chest X-ray for bronchiolitis once.  There was a white patch at the right heart border but it turned out to be tippex on the screen.


  1. Bronchiolitis in children: diagnosis and management, NICE guideline [NG9] Published date: June 2015