The Sounds of Winter: An audio-visual review of Paediatric Respiratory Disease

The leaves have turned brown, the temperature has dropped and Emergency Departments and wards are beginning to fill with wheezy coughing children. In the Northern hemisphere winter is with us…

Emergency and Urgent Care are often over-burdened by the effects of this seasonal change, which is particularly extreme in children. The predictable increase in respiratory disease this time of year should mean we are experts in its management but the combination of frequent staff turnover and challenging disease phenotypes militates against this.

Why is acute paediatric respiratory disease so challenging?

Different diseases are not only difficult to describe but may co-exist in the same patient

Asthma, Bronchiolitis, Croup and Pneumonia are 4 seemingly distinct clinical entities in children. However there is also this odd condition that exists between bronchiolitis and asthma in the UK termed viral wheeze (or as some family doctors call it: wheezy bronchitis). Finally the most common respiratory ‘condition’ that children acquire is a viral upper respiratory tract infection.

The snotty infant, off feeds with sub-costal recession and a cough, could have any one of these except asthma (although even the British Thoracic Guidance doesn’t use age as an exclusion). Ask a medical student to define asthma and I suspect you’ll get a potentially more correct answer than a doctor in the early years of training. The student may quote the patho-physiological diagnosis of reversible airways disease and bronchial mucus secretion. The junior doctor will be honed in on wheeze and prior history i.e what they see in practice. In the <1 group where bronchiolitis (breathing difficulties, cough, poor feeding in the context of wheeze and/or crepitations on auscultation [1]) is a common differential the diagnostic conundrum versus viral wheeze can cause confusion.  Practical definitions are often worthless until you see enough cases to be able to apply pattern recognition.

Croup is technically an easier diagnosis – a pattern of characteristic cough and stridor differentiates it from other respiratory conditions.

via mommy hood

However it’s not uncommon to see a child with stridor and a history of barking cough but wheeze on auscultation and a prolonged expiratory phase.  There is an association between croup and asthma and certainly a proportion of children can have mixed signs. This isn’t unsurprising as parainfluenza or any of the other viruses that can cause croup can set off the inflammatory cascade that typifies viral wheeze and asthma.

Finally pneumonia is one of those terms which is frequently used in different ways by health care professionals and the public. Lower Respiratory Tract Infections may be bacterial or viral in origin and although pneumonia (an infection or inflammation of the lung caused by nearly any class of organism) can also be either it is often used to imply a bacterial cause. Diagnosis using clinical signs is fraught with challenges (regardless of how clear the crackles are after having seen the x-ray). There is often an (unnecessary) tension to exclude pneumonia as the cause of the severity of a child’s condition in asthma or bronchiolitis. Which leads us nicely onto judgements of illness severity – a real but often unrecognised health system challenge.

Mild, Moderate and Severe Acuity Descriptors have inherent face validity but their assessment is more complex than the clinical features alone. 

Both professionals and members of the public would probably have a similar opinion on the severity of respiratory distress in this child

However we know there is large inter-observer variability in assessment. My own work has shown that experienced paediatricians differ in their interpretation of the severity of specific clinical signs.

Acuity Table from PVC Table

To be fair, interpretation of respiratory signs was probably the least variable (compared to alertness, hydration status etc.) but 40% of participants still differed in their assessment in this category. There are numerous reasons for this (blog pending on the role of gut instinct in interpretation of clinical signs) but one explanation is that specific features of respiratory distress don’t take into account previous illness trajectory. Typically a clinical feature (moderate recession etc.) will be linked to a degree of severity or a specific score. This is done in isolation of the other clinical features and doesn’t contribute to an overall impression of whether the child may be tiring or not.

How “severe” is the respiratory distress in this 8 month old child?

The child is grunting, has subcostal recession, and a raised respiratory rate (saturations were 94% in air). The underlying diagnosis could be anything from viral wheeze to pneumonia but you’ll have to take my word for it that this little guy had viral wheeze.  On a published score  relevant to the child’s age he would be at least moderate to severe. In practice an overall impression is also made on how long you think he will be able to sustain that level of work of breathing for. Part of this decision is based on experience and hence the variability of response when I show this video to different clinicians; from “wow, he is sick” to  “yes he is working hard but he is not too bad


Trajectory of illness is important. A child at point x may have been observed to be getting better (A) or getting worse (B). To an external person arriving to review the child there will be no difference between A and B but to the observing clinician their judgment on illness has been altered by how the child’s acuity has changed over time. There is a complicated third arm as well describing the ‘stable’ child who may suddenly improve or deteriorate (C1 or C2). Bronchiolitis, especially in neonates, often has this pattern. An experienced clinician develops a feel for the potential for sudden deterioration (although may not be able to explain why) and hence modulates their judgement on severity accordingly.

Acute Paediatric Respiratory Disease can be challenging

Every winter throws up a different severity of viral antigen producing varying intensities and complexities of respiratory presentations. There is not always a definitive best evidence approach. Considering all possible diagnoses, being aware of trajectory of illness and always listening to the parents will at least ensure you don’t get caught out.

I finish with a video of our grunting child taken 6 minutes after the original footage was taken (he received a single salbutamol nebuliser.) Please note the initial noise is from the child in the next cubicle 🙂

Thank you to the parents of these children who kindly consented to the filming and display of the video on an openly accessible site.

For those interested there are vine versions of the grunting and croupy child. Both parents were keen other health care professionals should be able to learn from them.

[1] This is taken from the SIGN Bronchiolitis 2006 guidance – although out of date the definition itself I think is pragmatic and remains credible

3 thoughts on “The Sounds of Winter: An audio-visual review of Paediatric Respiratory Disease”

  1. Another insightful review of a challenging and important topic (& I’m not a paediatrician or GP!).

    Two points:

    1. Having seen a patient oneself previously can be much more useful than any test at identifying severity *and* trajectory. Sadly, the system has done just about everything to maximise discontinuity of care at all points. As you point out, the relatives (parents in your case, children in mine!) often represent the only continuity in a case. Their observations are often gold dust, if they can be carefully elicited.

    2. As Gigerenzer has written about in his book Risk Savvy, algorithms are poor at guiding decision-making in complex systems – even with masses of data to inform their generation, whereas gut feelings can be surprisingly useful. This represents the accumulation of thousands of hours’ experience and observation over time, subconsciously curated into heuristic judgement systems.
    That’s why the consultant often ‘cuts to the chase’ – and may well not be able to explain how. I suspect in the future artifical neural networks may well be devised which can mimic this heuristic process, but don’t hold your breath!

  2. Another extremely interesting, thought provoking blog and one which is very topical. Just this week on nights I had an adult trained colleague come to me extremely worried after assessing a 6 year old with DIB. My assessment of them was mildly increased WOB and theirs was severe. Child was alert, speaking in short sentences, mild suprasternal and subcostal recession with O2 saturations of 97% on air. 10 puffs of salbutamol later the widespread expiratory wheeze I heard on auscultation was completely gone and stayed that way for 90 minutes. The situation was a perfect demonstration of experience being a variable in classification of perceived clinical severity and therefore risk stratification.

Leave a Reply