• 
• 

Three Part Question

In [children with croup] does [the dose of dexamethasone administered] have [an affect on outcome]?

Clinical Scenario

A two year old boy presents to the emergency department with a 4 hour history of barking cough, rhinorrhoea, stridor on exertion. A diagnosis of croup is made.

What does of dexamethasone should be given?

Search Strategy

Croup OR laryngotracheobronchitis AND dexamethasone AND dose OR dosage AND paediatric OR child.

Athens access to the NHS Evidence Journals and Databases.
The following database were searched:
MEDLINE 1946 to 27/04/2016
EMBASE 1980 to 27/04/2016
CINAHL 1981 to 27/04/2016

Search Outcome

113 articles in total.
4 RCTs, 1 retrospective observational study and 2 meta-analysis.
1 previous Best BET (4) from 2009 looked at one article.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Geelhoed GC, Macdonald WB 20th December 1995 Australia	120 children. Aged 6 - 160 months. Observation ward of a tertiary paediatric hospital in Western Australia. Croup score >2	Two sequential double blind randomised controlled trial comparing efficacy of a single dose of oral dexamethasone of varying sizes. Trial A randomised 60 children to 0.3mg/kg (max 6mg) or 0.6mg/kg (max 12mg) dexamethasone syrup. Trial B randomised 60 children to 0.3mg/kg (max 6mg) or 0.15mg/kg (max 3mg). Principle outcome measures were: duration of hospitalisation, reduction in croup scores, proportion hospitalised at 24hrs and use of adrenaline nebuliser. Level of evidence 2b, Jadad 3	No outcome measure was different for the two groups in each study	p <0.05	1. The numbers were small as powered at 80% to demonstrate a difference of a doubling in duration of hospital stay. 2. No mention re funding. 3. Although stated double blind and randomised no information regarding method of randomisation or concealment. 4. Families that did not speak English or had no telephone were excluded, raising the possibility of selection bias.
			Mean duration of hospitalisation in hours	A: 0.3mg/kg 7 vs 0.6mg/kg 8. B: 0.15mg/kg 9 vs 0.3mg/kg 9
			Reduction in croup scores	Graphs available in paper
			Proportion hospitalised at 24hrs	A: 0.3mg/kg 6.9% vs 0.6mg/kg 6.5%. B: 0.15mg/kg 3.4% vs 0.3mg/kg 0%
			Use of adrenaline nebuliser first hour	A; 0.3mg/kg 21% vs 0.6mg/kg 16%. B: 0.15mg/kg 45% vs 0.3mg/kg 19%
			Use of adrenaline nebuliser after 1hour	A: 0.3mg/kg 0% vs 0.6mg/kg 3%. B: 0.15mg 0% vs 0.3mg/kg 0%
Alshehr M, Almegamsi T 2005 Saudi Arabia	84 children were enrolled, 72 completed the study. Aged 6 -160 months. Emergency department and outpatients from three medical institutes in Saudi Arabia.	Double blind randomised trial to assess the efficacy of 0.15mg/kg compared to 0.6mg/kg in children with croup from 1998 - 2002. Primary outcome was change in total croup scores per 12 hour interval and classification of patients having a favourable outcome at 12 and 24 hours. Level of evidence 2b, Jadad 4	No difference in rate of hospitalisation	p = 0.36	1. Small study. 2. No follow up of patients. 3. Concerns are applicability to UK as a different population, difference in management with longer hospital admissions and mist therapy. 4. 12 withdrew with no comment on why, at what stage or from which group. 5. No mention of funding.
			No difference in mean hospital stay	p = 0.64
			No difference in total croup score at 12hrs	p = 0.15
			Two patients developed bronchopneumonia and on bacterial tracheitis all in the 0.6mg/kg group
Fifoot AA, Ting JS 2007 Australia	99 children, aged 6-79 months. Brisbane tertiary children's hospital emergency department. Croup Score >1	A randomised, double blinded, parallel design clinical trial to compare the effectiveness of 3 corticosteroid regimes (1mg/kg prednisone, 0.15mg/kg dexamethasone and 0.6mg/kg dexamethasone) in children with mild to moderate croup. Secondary aim was to determine if lower dose dexamethasone is as effective as the currently recommended dose. Oxford level of evidence 2b. Jadad 5.	No difference in croup score at 4hrs	p = 0.478	1. Pharmaceutical funding. 2. Data collected one week post attendance subject to recall bias.
			No difference in rate of return for medical care	p = 0.573
			No difference in numbers needing further steroids	p = 1.000
			No difference in admission rates	p = 0.498
			No difference in need for salvage therapy	p = 0.616
			No difference in representation rate	not available
Chub-Ippakarn and Sngsupawanich 2007 Thailand	41 children aged 7 months - 5 years recruited from paediatric ward of Hatvai hospital in southern Thailand. Westley croup scores 4-7.	Double blinded randomised controlled trial using computer generated random permuted blocks. Primary outcome was croup score at 12 hours. Oxford level of evidence 2b, Jadad 5	No difference in croup score at 12 hours	p = 0.40	1. Questions over generalisability due to population differences. 2. Admission bias as ward patients only. 3. Dexamethasone given intravenously, not orally. 4. All patients received a single dose of nebuliser adrenaline 1ml 1:1000 at presentation.
Dobrovoljac M, Geelhoed GC. 2009 Australia	Princess Margaret Hospital for Children. Data collected from 1996 - 2006 from emergency department, paediatric wards and intensive care unit records.	Retrospective observational study. Aimed to ascertain whether a reduction from a dose of 0.6mg/kg to 0.15mg/kg of dexamethasone maintaine the improved outcomes since in their previous study which made giving dexamethasone mandatory. Oxford level of evidence 2b	Representation rate had risen	4.1% in 1998 to 5.9% in 2006	1. Retrospective so limited to data recorded. 2. Possibility of historical control bias.
			No change in ward admissions, ICU admissions, ICU days or intubations
			Admission rates continued to fall	30% in early 1990s to 15% most recently
Russell KF, Liang Y, O’Gorman K, Johnson DW, Klassan TP. 2011	Included 3 papers answering the question of dexamethasone dosage: Geelhoed Alshehar and Fifoot (all included in this Best BET).	Aimed to provide evidence to guide clinicians in their treatment of children with croup, to examine the effectiveness of glucocorticoids compared to placebo or active treatment in croup patient to identify areas of uncertainty for future research. Primary outcomes change in clinical croup score from baseline to 6 hours, 12 hours and 24hours and return visits and/or (re)admissions.	No difference in croup score at 6 hours (0.15mg vs 0.3mg/kg)	Standard mean difference 0.17 (95% CI -0.34 to 0.67) p = 0.52	RCTs only
			Reattendance rates (0.15mg/kg vs 0.3mg/kg)	Risk ratio 2.81 (CI 0.12 to 66.4)
			No difference in croup score at 6 hours (0.6 vs 0.3mg/kg)	Standard mean difference of 0.21 (CI -0.30 to 0.70)
			Reattendance rates (0.6 vs 0.3mg/kg)	Risk ratio 1.87 (CI 0.18 to 19.55)
			No difference in croup score at 6 hours (0.15 vs 0.6mg/kg)
			Overall effect p value = 0.90

Comment(s)

Although studies are small, they appear generally well conducted. Each of the 4 RCTs here are double blinded with Jadad scores ranging from 3 to 5. No evidence of superiority of one dose over another was found. The retrospective observational study (Oxford level of evidence 2b) showed a continued low westely score after treatment, although did show a small increase in representation rates.

Clinical Bottom Line

0.15mg/kg of dexamethasone orally appears to be as effective clinically as other doses in treating children with croup. The ToPDoG trial comparing prednisolone 1mg/kg, 0.15mg/kg dexamethasone and 0.6mg/kg dexamethasone is currently underway.

References

1. Geelhoed GC, Macdonald WB Oral dexamethasone in the treatment of croup: 0.15 mg/kg versus 0.3 mg/kg versus 0.6 mg/kg. Pediatric Pulmonology 1995 Dec;20(6):362-8.
2. Alshehr M, Almegamsi T Efficacy of a small dose or oral dexamethasone in croup. Biomedical Research 2005;16 (1):65-72
3. Fifoot AA, Ting JS Comparison between single-dose oral prednisolone and oral dexamethasone in the treatment of croup: a randomized, double-blinded clinical trial Emergency Medicine Australasia 2007;19(1):51-8
4. Chub-Uppakarn S, Sangsupawanich derate to severe croup A randomized comparison of dexamethasone 0.15mg/kg versus 0.6mg/kg for the treatment of moderate to severe croup Int J Pediatr Otorhinolaryngol. 2007 March;71(3):473–7
5. Dobrovoljac M, Geelhoed GC. 27 years of croup: An update highlighting the effectiveness of 0.15mg/kg of dexamethasone Emerg Med Australas Aug 2009; 21(4); 309-14
6. Russell KF, Liang Y, O’Gorman K, Johnson DW, Klassan TP. Glucocorticoids for croup. Cochrane Database Syst Rev Cochrane Database 2011 Jan 19 (1); CD 001955