Bronchiolitis – The Simple Things in Life….

A 9-month old boy presented to his local emergency department in Australia with a 3 day history of poor feeding and difficulty breathing. His history was otherwise unremarkable. He was fully immunized and following normal growth and developmental trajectories.

His initial observations showed a low grade fever, mild tachycardia, tachypnea, and oxygen saturations of 95%. He had no signs of dehydration, moderate work of breathing with widespread wheeze and localized crepitations in the right lower chest posteriorly. He was also noted to have a soft systolic murmur.

The treating doctor diagnosed bronchiolitis. Because of localized findings in the right lower chest, she ordered a chest x-ray (CXR) to exclude pneumonia. The CXR was performed anterior-posterior and demonstrated borderline cardiomegaly with patchy opacification in the right lower lobe. The working diagnosis was modified to pneumonia.

The child was taken to the treatment room and an intravenous line inserted to commence antibiotics. After 3 attempts at insertion, work of breathing increased and a trial of bronchodilators was given with apparent effect. Regular bronchodilators were subsequently prescribed.

An electrocardiogram (EKG) was ordered based on the murmur and the borderline cardiomegaly seen on CXR. The EKG showed occasional ectopic beats and tachycardia. Given the constellation of murmur, ectopic beats, and CXR findings, a referral was made to a pediatric cardiologist.

The child was admitted for intravenous (IV) antibiotics and transferred to the ward. On the ward, he was monitored continuously and several transient desaturations to 88% were noted overnight. Therefore, he was commenced on low flow nasal prong oxygen with improvement in oxygen saturations to the mid-90s. Over the next 12 hours, his work of breathing increased and following multiple medical reviews, high flow nasal prong oxygen (HFNPO) therapy was commenced. In keeping with protocols for HFNPO, he was made nil by mouth and IV fluids were started. Chloral hydrate was also prescribed for irritability.

He remained in the hospital for 3 days while he was weaned from oxygen and feeds were re-established.

This child had typical bronchiolitis with no signs of bacterial infection, he was maintaining hydration, and not hypoxic. Reassurance, education, and discharge home with clear advice on when to seek further medical care would have been the most appropriate management.

Total cost saved would have been $5030, and as importantly side effects, time in hospital, risk of nosocomial infections, radiation, inconvenience to the family, and discomfort to the infant would have been avoided.

The CXR was performed to exclude a bacterial infection, but this child had no atypical features for bronchiolitis and was not more unwell than expected for this condition. CXRs are poor at discriminating between bacterial and viral infections. This is because viral infections often cause atelectasis because of airway obstruction, which can be difficult to distinguish from focal consolidation caused by bacterial pneumonia (Fig 1).^1,2  Also, a true inspiratory film can be difficult to acquire in a poorly compliant infant and poor inspiratory effort can mimic lung pathology (Fig 2).³  Finally, interpretation of CXRs is also subjective with variation seen even among experienced radiologists.^2,4,5 

Not surprisingly, with CXRs leading to misdiagnosis of pneumonia, antibiotic prescription rates also increase. In a prospective study, clinicians intent to order antibiotics rose from 2.6% to 14.7% of infants after a CXR was performed. As a self-limiting disease, with children often presenting at the peak of their illness, the improvement seen in the days following presentation may be falsely attributed to the effect of antibiotics. Unwarranted prescribing has implications for antibiotic stewardship as well as increasing the potential for unwanted side effects.⁶ 

In this case, the CXR not only incurred direct costs but led to unnecessary intravenous insertion, antibiotics, and a prolonged length of hospital stay. The child also received unnecessary radiation exposure at the beginning of their lifetime cumulative exposure. In addition, a spurious finding of an enlarged heart because of the CXR being taken anterior-posteriorly, contributed to the decision to refer to cardiology for follow up.

Although bronchiolitis presents similarly to asthma, bronchodilators have not been demonstrated to improve any clinically meaningful outcomes, such as duration of hospital admission or oxygen saturations.⁷  On the contrary, bronchodilators are associated with significant adverse effects such as tachycardia, oxygen desaturation, and tremors.⁷  Evidence, based on over 2000 participants in 31 randomized controlled trials, demonstrate limited evidence for the efficacy of bronchodilator use.⁷  Bronchodilators also contribute to environmental harm through their delivery device, metered-dose inhalers, which use potent greenhouse gases, hydroflurocarbons.⁸ 

Although the illness trajectory of bronchiolitis sees a peak in symptoms, the severity of symptoms may wax and wane throughout the day, with symptoms typically worsening after any energy output, such as feeding or crying. In this case, our infant’s work of breathing substantially increased after the trauma of intravenous line insertion. The timing of the bronchodilator administration coincided with the infant naturally calming after being soothed by his mother. This misattribution of causality has been well described by others.⁹  Attributing the improvement of symptoms to the medication led to ongoing prescription of an unnecessary medication, which in turn contributed to tachycardia and spurious results on the EKG and ultimately an unnecessary referral to cardiology.

The introduction of pulse oximetry in the 1980s led to rates of hospitalization for bronchiolitis soaring. With the ability to measure and react to low oxygen saturations, more children were diagnosed with “hypoxia” and “treated” with oxygen therapy. Despite this, no improvement in mortality was seen.¹⁰  The higher the oxygen treatment threshold, the higher the admission rates and length of stay.^11–13  A randomized control trial artificially elevated the oxygen saturations by 3% in the intervention cohort and showed lower rates of admission and shorter stay in the cohort with falsely elevated saturations.¹²  American guidelines now recommend an oxygen saturation threshold of 90% with UK and Australasian guidelines recommending a threshold of 92%.^14–16  The reduction in the threshold in the 2016 American guidelines saw rates of hospitalization fall from 17.9 per 1000 person-years in 2000 to 13.5 in 2016.¹⁷  A prospective study followed 118 children at home postdischarge and found transient desaturations below 90% were common and had no impact on readmission rates or clinical outcomes.¹⁸  A further study showed no difference in clinical outcomes but improved nursing satisfaction with intermittent oxygen saturation monitoring.¹⁹  The lack of certainty regarding what constitutes a safe oxygen level, combined with a growing recognition that transient desaturations are common and safe in infants with bronchiolitis, has led to the recommendation that continuous monitoring be avoided.^18–20 

In this case, the continuous monitoring detected fleeting desaturations and resulted in unnecessary oxygen treatment and a prolonged length of stay.

Once available only in intensive care units, HFNPO therapy in bronchiolitis has been more readily available on general medical wards and in the emergency department over the last decade.²¹  Widespread use was driven by the hypothesis that earlier use of noninvasive respiratory support might avoid intensive care admission. However, despite a substantial rise in use, there is no evidence that HFNPO has reduced length of stay or intensive care admissions. In fact, recent studies have shown the reverse trend, ie, higher rates of intensive care admission alongside greater use of noninvasive ventilation.²²  The role of HFNPO is as a rescue therapy for infants who have persistent hypoxaemia despite low flow oxygen therapy or have severe respiratory distress to the point of requiring transfer to a resuscitation area or high dependency unit.²¹ 

In this child, oxygen saturations were maintained on low flow oxygen, and the use of HFNPO to improve work of breathing was unwarranted. This led to the infant being unable to feed, and he subsequently became hungry and irritable, which was managed with unnecessary sedation.

Comfort measures such as swaddling, cuddling, and small frequent breast feeds should not be under-valued. Sometimes the simple things in life are the most extraordinary.

“It’s the simple things in life that are the most extraordinary.”

— Paul Coelho