Journal reference: Tibballs J, Watson T. Symptoms and signs differentiating croup and epiglottitis. Journal of Paediatrics and Child Health. (2010) 10.1111/j.1440-1754.2010.01892.x (published online, 21 November 2010) ^[1]

Link: http://dx.doi.org/10.1111/j.1440-1754.2010.01892.x

Published: 21 November 2010

Evidence cookie says...

Cough and drooling can be helpful in differentiating croup and epiglottitis in children with acute stridor. cough + absence of drooling → consider croup drooling + absence of cough → consider epiglottitis Note: the study results have some severe limitations and must be taken with care.

More details:

Article details

Study design:

Study aim:

to determine differentiating symptoms and signs of epiglottitis and laryngotracheobronchitis (croup)

Methods summary:

• From March 1990 to June 1993
• children presenting to the paediatric intensive care unit (PICU) at the Royal Children’s Hospital (RCH) with acute upper airway obstruction were examined carefully and signs recorded on a prepared form, when one of the two investigors was on duty
• In cases of suspected epiglottitis:
  • severe airway obstruction → intubation under general inhalational anaesthetic and throat swab taken
  • where intubation not immediately indicated → lateral x-ray of the neck
  • blood cultures and urine Hib antigen test performed
  • treatment with chloramphenicol or a third generation cephalosporin
• In cases of suspected croup:
  • the children were examined
  • treated with inhaled adrenaline and cephalosporin
• Children with acute airway obstruction due to other causes were recorded but not analysed further
• Children with airway obstruction due to anatomical or functional syndromic causes were excluded from the study
• The reliability of selected signs to diagnose croup and epiglottitis was determined by calculation of:
  • sensitivity
  • specificity
  • positive predictive value (PPV)
  • negative predictive value (NPV)

Results summary:

• 606 eligible patients
  • 374 diagnosed with croup
  • 189 diagnosed with epiglottitis
  • 43 with other conditions
• However, detailed symptoms and signs were recorded for only 203 children
  • 102 diagnosed with croup
  • 101 diagnosed with epiglottitis
• Symptoms:
  • children with croup had significantly greater incidence of:
    • cough, P < 0.001
    • breathing difficulty, P = 0.029
    • noisy breathing, P = 0.018
    • coryzal symptoms, P < 0.001
  • children with epiglottitis had significant greater incidence of:
    • drooling, P < 0.001
    • dysphonia, P < 0.001
    • fever, P = 0.012
    • preference to sit, P < 0.001
    • refusal of food, P < 0.001
    • dysphagia, P < 0.001
    • sore throat, P < 0.001
    • vomiting, P < 0.001
• Signs:
  • children with croup had greater incidence of:
    • cough, P < 0.001
    • retraction, P < 0.001
  • children with epiglottitis had greater incidence of:
    • drooling, P < 0.0001
    • preference to sit, P < 0.001
    • temperature above 38 °C
• Coughing for croup:
  • sensitivity = 1.00 (95% confidence interval [CI], 0.96-1.00)
  • specificity = 0.98 (95% CI, 0.93-0.99)
  • PPV = 0.98 (0.93-0.99)
  • NPV = 1.00 (0.96-1.00)
• Drooling for epiglottitis:
  • sensitivity = 0.79 (95% CI, 0.70-0.86)
  • specificity = 0.94 (0.88-0.97)
  • PPV = 0.93 (0.85-0.97)
  • NPV = 0.83 (0.74-0.89)

Study conclusion:

Epiglottitis and croup are often confused because they share symptoms and signs including stridor. However, differentiation in early illness is possible by additional observation of coughing and absence of drooling in croup and by the additional observation of drooling with absence of coughing in epiglottitis.

Participants:

• 606 children presenting to the PICU with acute upper airway obstruction in 1990-1993
• None of the children were vaccinated against Hib
• No other demographic details reported
• The children as a group were very unwell with almost half requiring endotracheal intubation
  • 374 diagnosed with croup
    • intubation required for 134 (36%)
  • 189  diagnosed with epiglottitis
    • intubation required for 167 (88%)
• However, only 203 children with detailed symptoms and signs recorded.
  • croup: 102 children
  • epiglottitis: 101 chidren
  • that is, all the study participants in the analysis were either diagnosed with croup or epiglottitis

Methodological weaknesses

• Selection bias presents a severe threat to the validity of the results; it is probable that the numerical values for the sensitivity, specificity, PPV and NPV for coughing and drooling are not externally valid.
  • firstly, the study population is already narrow; very unwell children presenting to PICU with symptoms of upper airway obstruction
  • secondly, attempt at capturing consecutive patients failed when it is considered that only 203 of 606 eligible children had detailed symptoms and signs recorded and thus were included in the analysis
  • thirdly, the statistical analysis insofar as sensitivity, specificity, PPV and NPV were conducted on a population of children where they all had one of two possible diagnoses, croup or epiglottitis.
• The methods section of the article seemingly describes a prospective case series. However, given that the data collection should have been completed 15 years prior to the publication of the article, and the fact that only a small proportion of the eligible cohort had the necessary data collected, it seems more likely that the study was in fact a retrospective review of medical records.

Methodological strengths

• Presents data on epiglottitis, a condition that is now rare.
• Provides clear data on differentiation of symptoms and signs between croup and epiglottitis (within the context of the narrow population)

Biases and conflicts of interests

Nil declared and nil seem obvious.

Clinical relevance to primary health care

Epiglottitis has been a rare condition since the introduction of routine vaccination against Hib. Nevertheless, sporadic cases remain and epiglottitis is potentially lethal especially if diagnosis is delayed.  Croup and epiglottis can be confused since both can present with stridor.

The results from this study cannot be validly generalised to the primary health care population and needs to be viewed in light of its study participants; very unwell children admitted to a paediatric intensive care unit (PICU) presenting with acute upper airway obstruction. Almost half of these children required intubation; all the children had either the diagnosis of croup or epiglottitis. For example, a positive predictive value (PPV) of 0.93 for drooling being diagnostic of epiglottitis (i.e., 93% of children with drooling have epiglottis) is obviously invalid in children presenting with stridor to a community general practice, and likely to be so in almost any modern Australian health setting.

Nevertheless, the negative predictive values (NPV) for cough and drooling remain useful and the authors’ cautious conclusion is likely to be true.

• croup is very unlikely in the absence of cough
• epiglottitis is unlikely in the absence of drooling

In the setting of acute airway obstruction in children, e.g., stridor:

• cough + absence of drooling → suspect diagnosis of croup
• drooling + absence of cough → suspect diagnosis of epiglottitis

References

1. Tibballs J, Watson T. Symptoms and signs differentiating croup and epiglottitis. Journal of Paediatrics and Child Health. (2010) 10.1111/j.1440-1754.2010.01892.x (published online, 21 November 2010)