Share this article on:

State of the Evidence for Standard-of-Care Treatments for Croup: Are We Where We Need to Be?

Cherry, James D. MD, MSc

The Pediatric Infectious Disease Journal: November 2005 - Volume 24 - Issue 11 - p S198-S202
doi: 10.1097/01.inf.0000188156.23182.eb
Supplement Article

Background: Croup is a term that groups several different clinical syndromes with inspiratory stridor. The failure to delineate the specific syndromes has led to suboptimal treatment in many instances.

Methods: A literature review and personal experience have been analyzed.

Results: Specific croup syndromes have been identified. Most croup steroid-treatment studies have failed to adequately identify the specific illness being treated. With 2 exceptions, all studies done to date have been too small to sufficiently evaluate risks of steroids if the risk is 1% or less.

Conclusions: Three conclusions were reached: (1) no steroid treatment studies of moderate or severe laryngotracheitis have been of adequate size to determine the risk of progressive viral infection or secondary bacterial infection; (2) single dose steroid treatment or limited nebulized use is probably safe; and (3) we should encourage better clinical diagnosis of croup illnesses.

From the Department of Pediatrics, David Geffen School of Medicine at UCLA, Los Angeles, CA

Address for reprints: James D. Cherry, MD, Department of Pediatrics, MDCC 22-442, David Geffen School of Medicine at UCLA, 10833 Le Conte Avenue, Los Angeles, CA 90095-1752. E-mail

The word “croup” is derived from the Anglo-Saxon word kropan, meaning “to cry aloud.”1 Today croup is a broad, clinical, diagnostic term used for several different respiratory illnesses that have varying degrees of inspiratory stridor and cough due to obstruction in the supraglottic, glottic and subglottic regions (Table 1). 1,2 The classification presented in Table 1 notes specific illnesses by anatomic location, clinical characteristics or etiology. Unfortunately during the past ∼30 years, there has been much confusion relating to terminology. For example, papers entitled “Laryngotracheobronchitis” usually discuss laryngotracheitis and spasmodic croup. In addition, many papers about bacterial croup are entitled “bacterial tracheitis,” which is incorrect considering most bacterial croups also have bronchial and pulmonary involvement.



Proper therapy in croup illnesses should depend upon specific clinical diagnostic characteristics. Unfortunately, today this is often not the case. The following discussion will include the epidemiology, causative pathogens and pathophysiology of specific syndromes, and will address clinical diagnosis and treatment controversies. Also, a major emphasis will be placed on the evidence relating to the use of steroids in croup.

Back to Top | Article Outline


In ancient times, as well as the first third of the 20th century, the term “croup” referred to diphtheria.1 Nondiphtheria croup, which was called “false croup,” was first recognized in 1852. During the first 40 years of the 20th century, it was recognized that laryngotracheobronchitis was caused by Corynebacterium diphtheriae, as well as other bacteria. In the 1940s, Davison identified spasmodic croup and described its differences from more severe forms of croup.3 In 1948, Rabe indicated that there were 3 types of croup: epiglottitis due to Haemophilus influenzae; diphtheritic croup; and viral croup.4 Nondiphtheritic bacterial croup went unrecognized for nearly another 30 years, until it was rediscovered in 1976.5

In the 20th century, the history of croup was marked by 3 important events: immunization, which led to the rapid decline in diphtheria; the introduction and widespread use of antibiotics, which decreased bacterial croup; and the advent of tissue culture, which allowed the identification of the causative role of viruses in most croup syndromes.

Back to Top | Article Outline


Etiologic agents in laryngotracheitis and spasmodic croup are presented in Table 2. Parainfluenza virus type 1 is the most common cause of both. Parainfluenza virus types 2 and 3 and influenza A virus are also frequent causes. The other agents listed in Table 2 are usually generally associated with more mild illness.



Bacterial croup manifests as laryngotracheobronchitis or laryngotracheobronchopneumonitis. In most instances the bacterial infection is secondary to a viral infection with a parainfluenza or an influenza virus. Today the most common implicated bacteria is Staphylococcus aureus, however, other bacteria, such as Streptococcus pyogenes, Streptococcus pneumoniae, H. influenzae and Moraxella catarrhalis, can also be causative.

Parainfluenza viruses types 1 to 3 are all major causes of respiratory infections in children.6 Infection with parainfluenza virus type 3 occurs in the majority of children during the first year of life, whereas infections with parainfluenza virus types 1 and 2 are more common in older children; one third of children will have antibodies against these agents by 4 years of age.

Outbreaks of croup (laryngotracheitis and spasmodic croup) regularly occur in the fall. In general, large outbreaks occur because of parainfluenza virus type 1 during odd-numbered years. Outbreaks due to parainfluenza virus type 2 are less predictable but also tend to occur during the fall. Croup due to parainfluenza virus type 3 is usually not outbreak-related and occurs in the late spring. Croup due to influenza viruses A and B occurs when these viruses are epidemic in a population.

Back to Top | Article Outline



Onset of illness is similar to that of a cold with coryza, followed by cough, sore throat and fever. Upper airway obstructive signs usually develop during the second or third day of illness but can occasionally occur during the first day. The cough becomes “croupy” (sounding like a barking seal or sea lion) and this is followed by respiratory stridor (difficulty with inspiration). The child is febrile (37.8–40.5°C) and on physical examination is noticeably hoarse and has coryza and an increased respiratory rate with a prolonged inspiratory phase.

The illness progresses at different rates and the final degree of airway obstruction is variable. Some children have only hoarseness and barking cough without evidence of obstruction whereas other children experience progressive obstruction. This may lead to severe respiratory distress with supraclavicular, infraclavicular and sternal retractions, cyanosis of varying degrees and apprehension. In severe cases hypoxia occurs and the cardiac rate increases. Without treatment to relieve the obstruction, an asphyxial death will occur in some children. In other cases, the prolonged hypoxia and increased respiratory effort can lead to respiratory fatigue and, consequently, to the patient's death. The duration of illness in those most severely affected is rarely less than 7 days.

Back to Top | Article Outline
Spasmodic Croup.

The onset of spasmodic croup is always at night. It can occur in children who are thought to be well or to have a mild cold. It is characterized by the child awakening with sudden dyspnea, croupy cough and inspiratory stridor, but no fever. Its occurrence tends to run in families and repeated attacks occur in some children. Following an episode, the child is likely to have another attack the same evening and on 3 or 4 successive evenings. Employing mild sedation at bedtime and ensuring that the room air is adequately humidified can prevent these secondary attacks.

Back to Top | Article Outline
Laryngotracheobronchitis and Laryngotracheobronchopneumonitis.

These illnesses are far less common than laryngotracheitis and spasmodic croup. The severity of illness in most instances is due to secondary bacterial infection. Initially, signs and symptoms are similar to those seen in laryngotracheitis, as described above. The afflicted child initially has mild to moderately severe illness, which lasts for 2 to 7 days. The child suddenly becomes markedly worse. Signs and symptoms associated with extension of the infection to the bronchi, bronchioles and lung include rales, air trapping, wheezing and an increase in the respiratory rate. On occasion, upper and lower airway obstruction seems to occur concurrently. The extent of the obstruction in bacterial croup is usually to such an extreme degree that either intubation or tracheotomy is necessary. Fever is usually marked.

In several instances, toxic shock syndrome has occurred in children with laryngotracheobronchopneumonitis due to S. aureus.

Back to Top | Article Outline


In laryngotracheitis, laryngoscopic studies reveal redness and swelling of the lateral walls of the trachea. This inflammatory swelling encroaches on the patency of the airway. The tracheal lumen becomes further obstructed by fibrinous exudates and the surface is covered by pseudomembranes composed of exudative material. The vocal chords are frequently swollen and their mobility is impaired. Histologic examination of the larynx and trachea reveals significant edema and cellular infiltration in the lamina proprin, submucosa and adventitia. This infiltrate includes histiocytes, lymphocytes, plasma cells and polymorphonuclear leukocytes.

Spasmodic croup is an enigma. In spite of the fact that it occurs and is associated with the same respiratory viruses that cause laryngotracheitis, the subglottic tissues have only noninflammatory edema.

The bronchitis and pneumonia of laryngotracheobronchitis and laryngotracheobronchopneumonitis are the results of the extension of the infection from the trachea. The progressive obstructive disease with exudates and pseudomembrane formation at the bronchial and bronchiolar levels is most often the result of secondary bacterial involvement (bacterial croup). In bacterial croup, the tracheal wall is infiltrated with inflammatory cells. Ulcerating pseudomembranes and microabscesses occur. Thick pus is present within the lumen of the trachea and lower air passages.

Back to Top | Article Outline


In contrast with the plethora of studies relating to the pathogenesis of bronchiolitis due to respiratory syncytial virus infection, there have been few similar studies of croup and parainfluenza viruses.2,6,7–15 From the pathologic data presented above it seems that the findings in laryngotracheitis are due to the direct cytopathic effect of a virus (parainfluenza virus types 1 and 2 and influenza virus types 1 and 2 and influenza virus types A and B) and the concomitant host response. The pathogenesis of laryngotracheobronchitis and laryngotracheobronchopneumonitis is similar, but with extension of the infection to the lower respiratory tract and, in most instances, the occurrence of secondary bacterial infection.

Several studies suggest that allergic factors play a role in recurrent croup.7–15 In a definitive study by Welliver et al,7 it was found that children with croup caused by parainfluenza viral infections had 3.6-fold higher titers of parainfluenza virus IgE antibody in nasopharyngeal secretions than similarly infected children with only upper respiratory infections. In the same study, the children with croup had a 1.6-fold greater cell-mediated immune response (measured by lymphocyte transformation) to the parainfluenza virus antigen than the children with only upper respiratory infections.7

In a subsequent review, Welliver et al9 noted that an atopic disposition may be associated with the development of croup. Specifically children with croup caused by parainfluenza virus infection have specific IgE antibodies and release histamine into the airway more frequently than patients with upper respiratory infection caused by parainfluenza viral infections. He also noted that children with recurrent croup have several features in common with atopic children, such as positive skin tests to environmental allergies, and that they are more likely to develop asthma when they grow older. In addition, some children with a history of recurrent croup develop stridor with histamine challenge.

From the above it is reasonable to accept the role of atopy in recurrent croup. It is difficult to explain, however, that some apparent primary infections result in spasmodic croup. Perhaps initial sensitization is parainfluenza virus group specific and not type specific. This hypothesis suggests that early infection (primary infection) with parainfluenza virus type 3 would set the stage for spasmodic croup with parainfluenza viruses types 1 and 2.

Back to Top | Article Outline



During the last 60 years there have been a number of controversies relating to the treatment of croup, such as tracheostomy versus no tracheostomy, intubation versus tracheostomy, warm versus cold humidification, antibiotics versus no antibiotics, sedation versus no sedation, racemic epinephrine treatment versus mist alone and steroids versus no steroids.2

Back to Top | Article Outline

Since 1952, literally hundreds of research studies, reviews and testimonials have described the use of corticosteroids in croup. Additionally, a number of meta-analyses have presented evidence supporting the use of steroids in croup; however, since controlled croup studies are still being published, it seems that there is still doubt about steroid efficacy.16 During the last 26 years, personal experience has led me to be one of the critics of the use of steroids in croup. Nevertheless steroids are recommended and routinely used to treat croup. For example, in the parainfluenza virus section of the 2003 edition of the Report of the Committee on Infectious Diseases of the American Academy of Pediatrics it is stated: “… parenteral dexamethasone in high doses (>0.3 mg/kg), oral dexamethasone (0.15–0.6 mg/kg), and nebulized corticosteroids have been demonstrated to lessen the severity and duration of symptoms and hospitalization in patients with moderate to severe laryngotracheobronchitis. Oral dexamethasone (0.15 mg/kg) also is effective for outpatients with less severe croup.”17

In the latest and most complete metaanalysis to date, Russell et al18 concluded glucocorticoids bring about clinical improvement in children with croup within 6 hours, nebulized budesonide or dexamethasone given either orally or intramuscularly are effective in treating croup, and the use of glucocorticoids is associated with a lower rate of epinephrine use, fewer return visits and/or (re)admission and shorter time spent in the hospital.

In contrast, my opinion as of 2004 was as follows: “In summary, a single dose of a systemic steroid, such as dexamethasone, administered intramuscularly or orally (0.6 mg/kg) or the limited use of nebulized budesonide probably is safe and may be useful in the child with more severe spasmodic croup. Steroids should not be used in laryngotracheobronchitis, laryngotracheobronchopneumonitis or epiglottitis.”2

In the same chapter, however, I wrote the following: “Of most concern to me today is the fact that risks of steroid use have not been evaluated and, because of small sample sizes in all available studies, they cannot be evaluated by meta-analysis. I have seen 3 children develop lower respiratory tract complications during steroid treatment of croup. In one case, an adenoviral pneumonia worsened and, in the other cases, bacterial tracheitis with pneumonia occurred. [It should be noted that in the 3 cases referred to, all received steroid treatment over several days and not single dose treatment.19] In a study by Super et al,20 there were 2 steroid-treated patients who developed pneumonia during therapy; none of the controls developed pneumonia. In a more recent trial in which 28 children received nebulized dexamethasone, 2 children with neutropenia developed bacterial tracheitis.21 Burton et al22 reported the occurrence of Candida laryngotracheitis as a complication of steroid and antibiotic treatment in a child with croup.

My opinions are supported by the findings in the extensive review by Russell et al.18 The review examined the comparative clinical responses in 24 different placebo-controlled studies. A listing of these studies is presented in Table 3. Despite the magnitude of studies evaluated, only 2 had enough subjects to assess the risks of steroids in regard to progressive viral infection or secondary bacterial complications. The largest study only included children with very mild croup (a Westley croup score of ≤2),23 suggestive of spasmodic croup and not laryngotracheitis.16 Some might argue that the use of dexamethasone in children with mild croup is unnecessary, since, objectively, symptoms are mild and self-limited.16



The second study, by Mühlendahl et al,24 is unusual because the children were typically older than expected (66% were ≥2 years of age). It was therefore more likely that their illnesses had an allergic component, rather than croup seen during first 2 years of life.

The remaining studies were not of adequate size to evaluate the risk of either progressive viral disease or secondary bacterial infection. Since the number of subjects was not sufficient enough to receive statistically and clinically significant results, a metaanalysis would be of little use.

Back to Top | Article Outline


Literature review has found that no steroid-treatment studies of moderate or severe laryngotracheitis have been of adequate size to determine the risk of progressive viral infection or secondary bacterial infection. However, current information suggests that the use of single dose steroid treatment or limited nebulized treatment is probably safe. To determine appropriate treatment, further understanding of croup is essential, and better clinical diagnosis of the illnesses should be encouraged.

Back to Top | Article Outline


Question: Are glucocorticoids considered at this time to be first line therapy in a child who presents with a barking cough, stridor and tachypnea, which is assumed to be croup?

James Cherry, MD: When referring to the evidence presented in the most recent review by Russell et al, treatment with corticosteroids compared with placebo results in reduced symptoms, less need for racemic epinephrine, fewer admissions to the hospital from emergency departments and shorter hospital stays. Steroids administered intramuscularly, orally or by aerosol have all demonstrated to be effective in relieving some symptoms associated with croup. However, as discussed in this presentation, the effects of repeated doses of corticosteroids in patients with severe croup have not been evaluated. Before it can be considered whether corticosteroids should be considered first-line therapy for all croup, it is imperative to study the risks in investigations which have ample size to determine serious but low-rate problems.

Question: Can nebulized epinephrine be used as a safe and effective alternative to steroids or racemic epinephrine to reduce edema in the treatment of croup?

James Cherry, MD: There has been limited research comparing the 2 treatments, even though both have been shown to be efficacious in the treatment of croup. In a multicenter, randomized, parallel study, Fitzgerald et al looked at 66 hospitalized children with viral or spasmodic croup. The subjects received either budesonide or l-adrenaline via nebulization. The change in total croup symptom score, duration of croup attack and side effects were measured. There was no significant difference observed in baseline scores. All patients had significant improvement from baseline; however, there was not a significant difference between the 2 treatment groups, as measured by change in croup score, change in oxygen saturation, duration of hospitalization, number of additional treatments and adverse events. Thus, the results demonstrate no difference in efficacy and safety between nebulized epinephrine and nebulized steroids in the treatment of croup. These results support nebulized epinephrine as an alternative to nebulized steroids. There is no reason to believe that there is any significant difference between treatment with nebulized racemic epinephrine versus nebulized l-adrenaline.

Back to Top | Article Outline


1. Cherry JD. Croup. In: Kiple KF, ed. The Cambridge World History of Human Disease. Cambridge, United Kingdom: Cambridge University Press; 1993:654–657.
2. Cherry JD. Croup (laryngitis, laryngotracheitis, spasmodic croup, laryngotracheobronchitis, bacterial tracheitis, and laryngotracheobronchopneumonitis). In: Feigin RD, Cherry JD, Demmler GJ, Kaplan S, eds. Textbook of Pediatric Infectious Diseases. 5th ed. Philadelphia, PA: W. B. Saunders Co.; 2004:252–266.
3. Davison FW. Acute obstructive laryngitis in children. Penn Med J. 1950;53:250–254.
4. Rabe EF. Infectious croup, I: etiology. Pediatrics. 1948;2:255–265.
5. Jones R, Santos JI, Overall JC Jr. Bacterial tracheitis. JAMA. 1979;242:721–726.
6. Hall CB. Parainfluenza viruses. In: Feigin RD, Cherry JD, Demmler GJ, Kaplan S, eds. Textbook of Pediatric Infectious Diseases. 5th ed. Philadelphia, PA: W. B. Saunders Co.; 2004:2270–2283.
7. Welliver RC, Sun M, Rinaldo D. Defective regulation of immune response in croup due to parainfluenza virus. Pediatr Res. 1985;19:716–720.
8. Welliver RC, Wong DT, Middleton E Jr, et al. Role of parainfluenza virus-specific IgE in pathogenesis of croup and wheezing subsequent to infection. J Pediatr. 1982;101:889–896.
9. Welliver RC. Croup: continuing controversy. Semin Pediatr Infect Dis. 1995;6:90–95.
10. Zach MS. Airway reactivity in recurrent croup. Eur J Respir Dis Suppl. 1983;128(pt 1):81–88.
11. Hide DW, Guyer BM. Recurrent croup. Arch Dis Child. 1985;60:585–586.
12. Castro-Rodriguez JA, Holberg CJ, Morgan WJ, et al. Relation of two different subtypes of croup before age three to wheezing, atopy, and pulmonary function during childhood: a prospective study. Pediatrics. 2001;170:512–518.
13. Wolf IJ. Allergic factors in the etiology of spasmodic croup and laryngo-tracheitis. Ann Allergy. 1966;24:79–82.
14. Nicolai T, Mutius E. Risk of asthma in children with a history of croup. Acta Pædiatr. 1996;85:1295–1299.
15. Van Bever HP, Wieringa MH, Weyler JJ, et al. Croup and recurrent croup: their association with asthma and allergy. Eur J Pediatr. 1999;158:253–257.
16. Bjornson CL, Klassen TP, Williamson J, et al. A randomized trial of a single dose of oral dexamethasone for mild croup. N Engl J Med. 2004;351:1306–1313.
17. American Academy of Pediatrics. Parainfluenza viral infections. In: Pickering LK, ed. Red Book: 2003 Report of the Committee on Infectious Diseases. 26th ed. Elk Grove Village, IL: American Academy of Pediatrics; 2003:454–455.
18. Russell K, Weibe N, Saenz A, et al. Glucocorticoids for croup [review]. The Cochrane Database of Syst Rev. 2005;Issue 1:1–67.
19. Cherry JD. Acute epiglottitis, laryngitis and croup. In: Remington JS, Swartz MN, eds. Current Clinical Topics in Infectious Diseases. New York, NY: McGraw-Hill; 1981:1–30.
20. Super DM, Cartelli NA, Brooks LJ, et al. A prospective randomized double-blind study to evaluate the effect of dexamethasone in acute laryngotracheitis. J Pediatr. 1989;115:323–329.
21. Johnson DW, Schuh S, Koren G, et al. Outpatient treatment of croup with nebulized dexamethasone. Arch Pediatr Adolesc Med. 1996;150:349–355.
22. Burton DM, Seid AB, Kearns DB, et al. Candida laryngotracheitis: A complication of combined steroid and antibiotic usage in croup. Int J Pediatr Otorhinolaryngol. 1992;23:171–175.
23. Westley CR, Cotton EK, Brooks JG. Nebulized racemic epinephrine by IPPB for the treatment of croup: a double-blind study. Am J Dis Child. 1978;132:484–487.
24. Mühlendahl KE, Kahn D, Spohr HL, Dressler F. Steroid treatment of pseudo-croup. Helv Paediart Acta. 1982;37:431–436.

croup; spasmodic croup; laryngotracheitis; laryngotracheobronchitis; laryngotracheobronchopneumonitis

© 2005 Lippincott Williams & Wilkins, Inc.