Pediatric croup: an evidence-based clinical review of assessment, dexamethasone, and home vs ED triage

The barking cough at midnight is one of the most recognizable presentations in pediatrics, but the distance between mild self-limited croup and respiratory compromise turns on a small number of signs. Dexamethasone has changed almost everything about how this disease is managed. The decision tree, however, still rests in the family's living room.

1. What croup is — and what it isn't

Croup, formally laryngotracheobronchitis or laryngotracheitis, is a viral inflammation of the subglottic airway that produces the classic triad of barking ("seal-like") cough, inspiratory stridor, and hoarseness. Symptoms typically follow 1–3 days of mild upper-respiratory prodrome, peak overnight, and resolve within 3–7 days¹. Most children with croup are not particularly sick by global appearance; the airway sound is the prominent finding.

Etiology is dominantly viral. Parainfluenza virus types 1, 2, and 3 account for roughly three quarters of laboratory-confirmed cases. RSV, influenza, adenovirus, rhinovirus, and SARS-CoV-2 produce smaller fractions, with SARS-CoV-2-associated croup over-represented in Omicron-era pediatric ED data⁵. Bacterial superinfection (bacterial tracheitis) is rare but life-threatening and presents distinctly: high fever, toxic appearance, copious purulent secretions, and a poor response to standard croup therapy.

What croup is not: it is not epiglottitis, retropharyngeal abscess, anaphylaxis with angioedema, or foreign-body aspiration. Each of those has its own evaluation pathway. The clinical features that pull the diagnosis away from viral croup — drooling, refusal to lie down, the "tripod" posture, sudden onset without a viral prodrome, unilateral wheeze — are flagged in §6 below as red flags for immediate medical evaluation.

2. Epidemiology and natural history

Croup incidence in the developed world is estimated at 3–6% of children annually under age 6, with peak incidence between 6 months and 3 years and a male predominance of roughly 1.4:1¹. The condition is seasonal in temperate climates, with parainfluenza type 1 driving a fall/winter peak and parainfluenza type 3 producing scattered spring cases. The CDC's parainfluenza surveillance documents these patterns in the National Respiratory and Enteric Virus Surveillance System (NREVSS) data⁵.

Natural history is reassuring at the population level. Hospitalization occurs in roughly 1–8% of presenting cases, intubation in <1%, and mortality is exceedingly rare¹. Most children with croup never see an emergency department; of those who do, the great majority go home from the ED after a single dexamethasone dose and observation period.

Recurrence is common. Roughly 5% of children meet criteria for "recurrent croup" (more than three episodes in a year), and a subset of those have an underlying subglottic abnormality or atopic predisposition. The AAP HealthyChildren guidance recommends pulmonology or otolaryngology evaluation in children with frequent recurrences or atypically severe episodes⁴.

3. Assessment: the Westley score

The Westley croup score, originally described in the 1978 nebulized-epinephrine trial that established that drug's role, remains the standard severity instrument in pediatric emergency medicine³. It is a five-domain ordinal scale, computed without instrumentation, and validated across decades of subsequent trials.

Total scores stratify severity:

• Mild (≤2): Barking cough, no stridor at rest, no significant retractions. Most cases.
• Moderate (3–7): Stridor at rest, mild-moderate retractions, no agitation.
• Severe (8–11): Stridor at rest with significant retractions, agitation or distress.
• Impending respiratory failure (≥12): Decreased consciousness, cyanosis at rest, decreased air entry. Airway intervention pathway.

4. Therapy: dexamethasone is the foundation

Glucocorticoids transformed croup care. The 2018 Cochrane review pooled 43 trials with 4,565 children and found glucocorticoids reduce return-visit rates, length of hospital stay, ED length-of-stay, and total time to symptom resolution across the severity spectrum². The effect is robust across mild, moderate, and severe disease — the historical practice of reserving steroids for moderate or worse cases is no longer supported.

Dosing. The most widely studied regimen is oral dexamethasone 0.6 mg/kg as a single dose, maximum 16 mg⁷. Lower doses — 0.15 mg/kg and 0.3 mg/kg — have demonstrated non-inferiority in multiple randomized trials and are used preferentially in some centers to limit hyperglycemic and behavioral side effects. The 2004 Bjornson trial in the New England Journal of Medicine specifically established benefit for mild croup with 0.6 mg/kg, with a 30% reduction in return visits and a meaningfully lower rate of subsequent ED care⁷.

Route. Oral, IM, and nebulized routes are all evidence-supported. Oral is the default for cooperative children; IM is acceptable when vomiting or refusal precludes oral; nebulized budesonide is an alternative for children who cannot tolerate oral medication, with non-inferior outcomes to oral dexamethasone in head-to-head trials².

Why a single dose is enough. Dexamethasone's half-life (~36–54 hours) covers the natural-history window of clinical illness. Repeat dosing is not supported by trial evidence in uncomplicated croup. Children whose symptoms recur outside the 48-hour window should be re-evaluated rather than redosed at home.

5. Therapy: nebulized epinephrine for moderate-to-severe disease

For children with stridor at rest, significant retractions, or distress, nebulized epinephrine provides rapid airway-edema reduction. Onset is within 10–30 minutes; duration is approximately 2 hours; effects then taper, with potential for rebound symptoms³.

Dosing. Racemic epinephrine 2.25% solution, 0.05 mL/kg (max 0.5 mL) diluted in 3 mL saline; or L-epinephrine 1:1000, 0.5 mL/kg (max 5 mL). Either preparation is acceptable per the Cochrane evidence base².

Disposition rule. Any child who receives nebulized epinephrine requires observation for a minimum of 3–4 hours from the dose to confirm absence of rebound stridor before discharge home. This is the most commonly missed step in community emergency departments and is responsible for a small but real fraction of bounce-back visits.

6. Disposition: when home is enough, when ED is required

Home management is appropriate for the well-appearing child with a Westley score ≤2 who has received a single oral dexamethasone dose (when accessible via urgent care or telehealth) or is in a family with a clear escalation plan. Cool humid air (open freezer briefly, cool night air), upright positioning, fluids, and calm management of the child's distress are the supportive measures supported by the AAP parent-facing guidance and NIH MedlinePlus⁴⁶.

The CDC parainfluenza guidance and AAP HealthyChildren materials both emphasize one practical detail: croup peaks overnight, and the same child can be a Westley-1 at noon and a Westley-5 at midnight. Families discharged with mild croup should be counseled explicitly about the diurnal pattern and given a clear "call back if" list⁴⁵.

7. Common myths and evidence corrections

"Cool mist therapy resolves croup." Multiple randomized trials, including a 2006 trial in JAMA, have failed to demonstrate any benefit from humidified-air therapy in moderate croup¹. Cool air is a reasonable, low-harm supportive measure that helps some children, but it is not therapy. The clinical effect of "stepping outside with the baby" likely owes more to calming and brief positioning than to humidity.

"Steroids are too strong for mild croup." The 2004 Bjornson trial and the 2018 Cochrane review specifically demonstrate benefit in mild croup²⁷. The historical practice of reserving steroids for moderate-or-worse cases predates this evidence and is now actively discouraged in pediatric emergency literature.

"Antibiotics for croup." Croup is viral; antibiotics have no role in standard laryngotracheitis. Bacterial tracheitis is a distinct, rare entity treated with parenteral antibiotics and frequently airway intervention — its toxic appearance and lack of dexamethasone response distinguish it clinically.

"Albuterol works in croup." Croup is an upper-airway condition; albuterol targets lower-airway smooth muscle and has no demonstrated benefit. Concurrent wheeze suggests a different or overlapping diagnosis that warrants separate evaluation.

8. Recurrent and atypical presentations

Children with more than three croup episodes in a year, or any single episode requiring intubation, should be referred for pediatric otolaryngology or pulmonology evaluation to rule out an anatomic predisposition. Subglottic hemangioma, subglottic stenosis (acquired or congenital), and laryngomalacia are the principal differentials. A "spasmodic croup" presentation — sudden nighttime stridor without preceding viral symptoms and with rapid resolution — is sometimes associated with gastroesophageal reflux and atopic predisposition, and may benefit from reflux evaluation in select cases¹.

For the parent-facing complement to this clinical review, see Wermom's RSV and bronchiolitis primer, which covers the related but distinct lower-airway syndrome, and the pediatric fever management guide for the antipyretic guidance that applies to most croup presentations.

9. Care pathway summary

10. Where the evidence base is still soft

Three areas remain less settled. First, the optimal dexamethasone dose: 0.15 mg/kg may be sufficient for most cases, but the historical 0.6 mg/kg dose is what most trial outcomes are anchored to and remains the most defensible default. Second, the role of heliox for severe croup: small studies suggest benefit, but the evidence base does not yet support routine use. Third, post-SARS-CoV-2 croup management: Omicron-era pediatric ED data showed higher severity and ICU admission rates in SARS-CoV-2-associated croup than in parainfluenza croup, but treatment principles remain unchanged. The CDC's parainfluenza and respiratory virus surveillance pages continue to track these patterns⁵.