crup 3.pdf

8/9/2019 crup 3.pdf

1/109

Glucocorticoids for croup (Review)

Russell KF, Liang Y, OGorman K, Johnson DW, Klassen TP

This is a reprint of a Cochrane review, prepared and maintained by The Cochrane Collaboration and published inThe Cochrane Library2011, Issue 1

http://www.thecochranelibrary.com

Glucocorticoids for croup (Review)

Copyright 2012 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
http://www.thecochranelibrary.com/http://www.thecochranelibrary.com/

8/9/2019 crup 3.pdf

2/109

T A B L E O F C O N T E N T S

1HEADER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2PLAIN LANGUAGE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3OBJECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Figure 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

11DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Figure 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

Figure 5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

15AUTHORS CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

15ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

15REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22CHARACTERISTICS OF STUDIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

51DATA AND ANALYSES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Analysis 1.1. Comparison 1 Any glucocorticoid versus placebo, Outcome 1 Croup score (change baseline - 6 hours) by

score. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56


score. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58


score. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

Analysis 1.4. Comparison 1 Any glucocorticoid versus placebo, Outcome 4 Westley score (change baseline - 6 hours) by

inpatient/outpatient. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60



Analysis 1.6. Comparison 1 Any glucocorticoid versus placebo, Outcome 6 Westley score (change baseline - 24 hours) byinpatient/outpatient. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62


glucocorticoid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63


glucocorticoid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64


glucocorticoid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

Analysis 1.10. Comparison 1 Any glucocorticoid versus placebo, Outcome 10 Return visits and/or (re)admissions by


Analysis 1.11. Comparison 1 Any glucocorticoid versus placebo, Outcome 11 Return visits and/or (re)admissions by

glucocorticoid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

Analysis 1.12. Comparison 1 Any glucocorticoid versus placebo, Outcome 12 Return visits and/or (re)admissions by croup

severity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68Analysis 1.13. Comparison 1 Any glucocorticoid versus placebo, Outcome 13 Length of stay by inpatient/outpatient. 69

Analysis 1.14. Comparison 1 Any glucocorticoid versus placebo, Outcome 14 Improvement (at 6 hours) by






Analysis 1.17. Comparison 1 Any glucocorticoid versus placebo, Outcome 17 Additional treatments. . . . . . . 73

iGlucocorticoids for croup (Review)


8/9/2019 crup 3.pdf

3/109

Analysis 2.1. Comparison 2 Any glucocorticoid versus epinephrine, Outcome 1 Croup score (change baseline - 6 hours) by


Analysis 2.2. Comparison 2 Any glucocorticoid versus epinephrine, Outcome 2 Croup score (change baseline - 12 hours)

by inpatient/outpatient. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76


by inpatient/outpatient. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77Analysis 2.4. Comparison 2 Any glucocorticoid versus epinephrine, Outcome 4 Croup score (change baseline - 6 hours) by

glucocorticoid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78


by glucocorticoid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79


by glucocorticoid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

Analysis 2.7. Comparison 2 Any glucocorticoid versus epinephrine, Outcome 7 Return visits and/or (re)admissions by


Analysis 3.1. Comparison 3 Dexamethasone versus budesonide, Outcome 1 Croup score (change baseline - 6 hours) by


Analysis 3.2. Comparison 3 Dexamethasone versus budesonide, Outcome 2 Croup score (change baseline - 12 hours) by


Analysis 3.3. Comparison 3 Dexamethasone versus budesonide, Outcome 3 Return visits and/or (re)admissions byinpatient/outpatient. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84

Analysis 4.1. Comparison 4 Budesonide and dexamethasone versus budesonide, Outcome 1 Croup score (change baseline -

6 hours) by inpatient/outpatient. . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

Analysis 4.2. Comparison 4 Budesonide and dexamethasone versus budesonide, Outcome 2 Return visits and/or

(re)admissions by inpatient/outpatient. . . . . . . . . . . . . . . . . . . . . . . . . . 86

Analysis 5.1. Comparison 5 Budesonide and dexamethasone versus dexamethasone, Outcome 1 Croup score (change

baseline - 6 hours) by inpatient/outpatient. . . . . . . . . . . . . . . . . . . . . . . . . 87

Analysis 5.2. Comparison 5 Budesonide and dexamethasone versus dexamethasone, Outcome 2 Return visits and/or

(re)admissions by inpatient/outpatient. . . . . . . . . . . . . . . . . . . . . . . . . . 88

Analysis 6.1. Comparison 6 Oral versus intramuscular dexamethasone, Outcome 1 Croup score (change baseline-discharge)


Analysis 6.2. Comparison 6 Oral versus intramuscular dexamethasone, Outcome 2 Return visits and/or (re)admissions by

inpatient/outpatient. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90Analysis 7.1. Comparison 7 Dexamethasone 0.30 mg/kg versus 0.15 mg/kg, Outcome 1 Croup score (change baseline - 6

hours) by inpatient/outpatient. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

Analysis 7.2. Comparison 7 Dexamethasone 0.30 mg/kg versus 0.15 mg/kg, Outcome 2 Return visits and/or (re)admissions


Analysis 8.1. Comparison 8 Dexamethasone 0.60 mg/kg versus 0.30 mg/kg, Outcome 1 Croup score (change baseline - 6




Analysis 9.1. Comparison 9 Dexamethasone 0.60 mg/kg versus 0.15 mg/kg, Outcome 1 Croup score (change baseline - 6




Analysis 10.1. Comparison 10 Dexamethasone versus betamethasone, Outcome 1 Croup score (change baseline - 6 hours)by inpatient/outpatient. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

Analysis 10.2. Comparison 10 Dexamethasone versus betamethasone, Outcome 2 Return visits and/or (re)admissions by


Analysis 11.1. Comparison 11 Dexamethasone versus prednisolone, Outcome 1 Croup (change baseline - 6 hours) by


Analysis 11.2. Comparison 11 Dexamethasone versus prednisolone, Outcome 2 Return visits and/or (re)admissions by


99ADDITIONAL TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

iiGlucocorticoids for croup (Review)


8/9/2019 crup 3.pdf

4/109

100APPENDICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

102FEEDBACK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

104WHATS NEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

104HISTORY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

104CONTRIBUTIONS OF AUTHORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

105DECLARATIONS OF INTEREST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105SOURCES OF SUPPORT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

105INDEX TERMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

iiiGlucocorticoids for croup (Review)


8/9/2019 crup 3.pdf

5/109

[Intervention Review]

Glucocorticoids for croup

Kelly F Russell2, Yuanyuan Liang3, Kathleen OGorman2, David W Johnson4, Terry P Klassen1

1Manitoba Institute of Child Health, Winnipeg, Canada. 2Department of Pediatrics, University of Alberta, Edmonton, Canada.3Department of Epidemiology and Biostatistics, University of Texas Health Science Center at San Antonio, San Antonio, Texas, USA.4Department of Pediatrics, Faculty of Medicine, University of Calgary, Alberta Childrens Hospital, Calgary, Canada

Contact address: Terry P Klassen, Manitoba Institute of Child Health, 513-715 McDermot Avenue, Winnipeg, Manitoba, MB R3E

3P4, [email protected].

Editorial group:Cochrane Acute Respiratory Infections Group.

Publication status and date:Edited (no change to conclusions), published in Issue 1, 2012.

Review content assessed as up-to-date: 23 July 2010.

Citation: Russell KF, Liang Y, OGorman K, Johnson DW, Klassen TP. Glucocorticoids for croup.Cochrane Database of SystematicReviews2011, Issue 1. Art. No.: CD001955. DOI: 10.1002/14651858.CD001955.pub3.


A B S T R A C T

Background

Since the initial publication of this systematic review in 1997, several randomized trials examining the benefit of glucocorticoids have

been published. The objective of this review is to provide evidence to guide clinicians in their treatment of patients with croup by

determining the effectiveness of glucocorticoids and to identify areas requiring future research.

Objectives

To determine the effect of glucocorticoids for children with croup.

Search methods

We searched CENTRAL (2010, Issue 3), which contains the Cochrane Acute Respiratory Infections Groups Specialized Register,

MEDLINE (1966 to July week 2, 2010) and EMBASE.com (1974 to July 2010). We also contacted authors of identified croup trials

published in the last 10 years to inquire about additional published or unpublished trials.

Selection criteria

Randomized controlled trials (RCTs) that examine children with croup and objectively measure the effectiveness of glucocorticoids.

Data collection and analysis

Two review authors identified studies for potential relevance based on the review of the title and abstract (when available). Two review

authors independently reviewed studies for relevance using a priori inclusion criteria and assessed trial quality. Differences were resolved

by consensus. One review author extracted data using a structured form and another review author checked the results for accuracy.

We performed standard statistical analyses.

Main results

Thirty-eight studies were included (n = 4299). Glucocorticoids were associated with an improved Westley score (maximum 17 points)

at six hours with a mean difference of -1.2 (95% confidence interval (CI) -1.6 to -0.8) and at 12 hours -1.9 (95% CI -2.4 to -1.3); at

24 hours this improvement was no longer significant (-1.3, 95% CI -2.7 to 0.2). Fewer return visits and/or (re)admissions occurred in

participants treated with glucocorticoids (risk ratio (RR) 0.5; 95% CI 0.3 to 0.7). Length of time spent in accident and emergency or

hospital (mean difference 12 hours, five to 19 hours) was significantly decreased for participants treated with glucocorticoids. Use of

epinephrine decreased for children treated with a glucocorticoid (risk difference 10%; 95% CI 1 to 20).

1Glucocorticoids for croup (Review)

mailto:[email protected]:[email protected]

8/9/2019 crup 3.pdf

6/109

Authors conclusions

Dexamethasone and budesonide are effective in relieving the symptoms of croup as early as six hours after treatment. Fewer return

visits and/or (re)admissions are required and the length of time spent in hospital is decreased. Research is required to examine the most

beneficial method for disseminating croup practice guidelines and to increase the uptake of evidence.

P L A I N L A N G U A G E S U M M A R Y

Glucocorticoids for croup

Croup is common in children and is thought to be triggered after a viral infection. Croup causes swelling in the throat and windpipe

(trachea) and causes hoarseness, a barking cough and noisy breathing. Croup usually gets better by itself but sometimes drugs are used

to try and improve this condition. The review looked at trials of one type of steroid drug, glucocorticoids. Glucocorticoids can reduce

the swelling and make it easier for the child to breathe. We found that glucocorticoids can start improving croup in children within

six hours (14 studies, 1031 children). The effect lasts about 12 hours (eight studies, 532 children), lessens the need for other drugs,

and shortens hospital stays by 12 hours (eight studies, 795 children). There were no adverse events associated with glucocorticoids.

Additional studies are needed to determine the best dose of glucocorticoids.

B A C K G R O U N D

Description of the condition

Croup (laryngotracheobronchitis) is a common cause of upper

airway obstruction in children and is characterized by hoarse-

ness, a barking cough and inspiratory stridor. These symptoms are

thought to occur as a result of oedema of the larynx and trachea,

which has been triggered by a recent viral infection. Para influenza

virus type 1 is the agent most commonly identified in cases of

croup (Henrickson 1994).

Although croup is a self-limiting illness, it is a large burden on

health care systems because of frequent visits to doctors and ac-

cident and emergency departments and, when necessary, hospi-

talizations. The annual incidence of croup in children younger

than six years ranges from 1.5% to 6% (Denny 1983). During a

six-year period in Alberta, Canada, 20,019 infants less than two

years of age visited the emergency department for 27,355 episodes

of croup and 8.0% of infants were admitted (Rosychuk 2010).Admission rates for croup in children seen in outpatient settings

range from 1.5% to 31% of cases seen; these figures vary widely,

depending on hospital admission practices and the severity of the

disease in the population being assessed (Marx 1997;To 1994).

Description of the intervention

Since the late 1980s it has been recognized that glucocorticoids

provide some clinical benefit for children with croup. In 1989

Kairys 1989 published a meta-analysis of clinical trials examin-

ing the benefit of glucocorticoids. Previous research has shown

that a greater proportion of children are receiving glucocorticoids

in the emergency department. An Australian hospital introduced

a mandatory glucocorticoid policy for children with croup and

this resulted in decreased admissions and intubation (Geelhoed1996b). A second Australian study found that by 1992, 87% of

children received corticosteroids upon admission and ICU admis-

sionsand intubation ratesdecreased(McDonogh 1994). Racaemic

adrenaline (epinephrine), or L-adrenaline, has been shown to pro-

vide temporary relief to patients with croup but is not thought to

have long-term benefits (Johnson 2009;Waisman 1992). In the

past, the standard management of croup included mist treatment

(that is, treatment with humidified air), although there is little ev-

idence of its effectiveness (Lavine 2001;Moore 2007;Neto 2002;

Scolnick 2006).

How the intervention might work

Localised inflammation of the upper airway caused by a virus

is believed to cause croup symptoms. The narrowest part of the

airway is the subglottic region. Even small amounts of oedema

in this region can significantly increase the work of breathing in

young children because airway resistance increases dramatically.

Poiseuilles law states that airway resistance is related to the radius

of the airway to the power of four. Glucocorticoids are believed to



8/9/2019 crup 3.pdf

7/109

reducethe degreeof inflammationand swelling through their anti-

inflammatory properties leading to decreased effort in breathing

for the child.

Why it is important to do this reviewIn 1999 this review was first published in The Cochrane Libraryand included 24 RCTs that examined the effectiveness of gluco-

corticoids for children with croup (Ausejo 2000). However, since

then a number of RCTs have been published and there has been

continued interest in the use of glucocorticoids to treat patients

with croup. This is an update of the original review. There are ar-

eas of uncertainty regarding the use of glucocorticoids in children

with croup, such as direct comparisons within a study examining

the various routes of administration of glucocorticoids and the

role of glucocorticoids in the treatment of mild croup. Also, in

the previous version of this review, publication bias was a threat

to the validity of results and incorporation of additional trials may

influence the presence of publication bias.

O B J E C T I V E S

The objective of this systematic review was to provide evidence

to guide clinicians in their treatment of children with croup, to

examine the effectiveness of glucocorticoids (oral, spray, nebulized

or injected) compared to placebo or active treatment(s) in croup

patients and to identify areas of uncertainty for future research.

M E T H O D S

Criteria for considering studies for this review

Types of studies

Participants had to have been randomly assigned to treatment

groups. We included studies regardless of language of publication

or publication status.

Types of participants

We included children with croup, pseudo croup or laryngotra-

cheitis. We excluded bronchiolitis. We defined croup as a syn-

drome consisting of hoarseness, barking cough and stridor, in

which alternative diagnosis of acute stridor was excluded. We

included both inpatients and outpatients; we classified patients

treated in an emergency department as outpatients.

Types of interventions

We included studies where the effectiveness of glucocorticoids (ad-

ministered orally, parenterally or by nebulization) was compared

to a placebo or any active treatment.

Types of outcome measures

Primary outcomes

1. Change in clinical croup score from baseline to six hours,

12 hours or 24 hours.

2. Return visits and/or (re)admissions.

Secondary outcomes

1. Length of stay in hospital or accident and emergency

department.

2. Patient improvement (yes or no).3. The use of additional treatments such as epinephrine,

supplemental glucocorticoids, mist treatment, intubation, or

antibiotics.

Studies that failed to meet all four of the inclusion criteria were

excluded.

Search methods for identification of studies

Electronic searches

For the original review we searched the Cochrane Central Register

of Controlled Trials (CENTRAL) (The Cochrane Library2006,Issue 1), which contains the Cochrane Acute Respiratory Infec-

tions Groups Specialized Register. In addition, we searched MED-

LINE (January 1966 to January 2006), exploding glucocorticoid

treatment (and each of the terms for corticosteroids) and croup;

we restricted the search to randomized controlled trials using a

previously validated strategy. We also searched Excerpta Medica

and EMBASE (January 1974 to April 2006) (seeAppendix 1for

details of the search strategies).

For this update, we searched Cochrane Central Register of Con-

trolled Trials (CENTRAL) (The Cochrane Library2010, Issue3), which contains the Cochrane Acute Respiratory Infections

Groups Specialized Register, MEDLINE (1966 to July week 2,

2010) and EMBASE.com (1974 to July 2010). We also contacted

authors of identified croup trials published in the last 10 years to

inquire about additional published or unpublished trials.

We used the following search strategy to search MEDLINE and

CENTRAL. We combined the MEDLINE search strategy with

the Cochrane HighlySensitive SearchStrategy for identifying ran-

domized trials in MEDLINE: sensitivity- and precision-maximiz-

ing version (2008 revision) Ovid format (Lefebvre 2008). The



8/9/2019 crup 3.pdf

8/109

search was adapted for EMBASE (see Appendix 2) and MED-

LINE in-process (seeAppendix 3).

MEDLINE

1 exp Laryngitis/

2 laryngit*.tw.

3 croup*.tw.4 laryngotracheobronchit*.tw.

5 (pseudocroup* or pseudo-croup*).tw.

6 laryngotracheit*.tw.

7 or/1-6

8 exp Glucocorticoids/

9 glucocorticoid*.tw,nm.

10 Adrenal Cortex Hormones/

11 corticosteroid*.tw,nm.

12 corticoid*.tw,nm.

13 Hydrocortisone/

14 hydrocortisone.tw,nm.

15 hydroxypregnenolone.tw,nm.

16 Pregnenolone/17 pregnenolone.tw,nm.

18 Tetrahydrocortisol/

19 tetrahydrocortisol.tw,nm.

20 Hydrocortisone/

21 hydrocortisone.tw,nm.

22 Cortodoxone/

23 cortodoxone.tw,nm.

24 Cortisone/

25 cortisone.tw,nm.

26 Corticosterone/

27 corticosterone.tw,nm.

28 Hydroxycorticosteroids/

29 hydroxycorticosteroid*.tw,nm.30 triamcinolone.tw,nm.

31 prednisone.tw,nm.

32 prednisolone.tw,nm.

33 paramethasone.tw,nm.

34 methylprednisolone.tw,nm.

35 dexamethasone.tw,nm.

36 clobetasol.tw,nm.

37 beclomethasone.tw,nm.

38 betamethasone.tw,nm.

39 Pregnenediones/

40 pregnenedione*.tw,nm.

41 budesonide*.tw,nm.

42 or/8-4143 7 and 42

Searching other resources

In addition, we sent letters to the trial authors of studies published

inthe past 10 years to enquire whetherthey knew of anyotherpub-

lished or unpublished trials. We retrieved more than 700 unique

references. There were no language or publication restrictions in

our searches.

Data collection and analysis

Selection of studies

Two review authors (original review AS, TPK; 2004 update KR,

LH, NW; 2010 update KR, KO) independently selected studies

from this list of references as being potentially relevant based on

a review of titles and, when available, the abstracts. We retrieved

the potentially relevant studies as full manuscripts. Two review au-

thors (original review AS, TPK; 2004 update KR, LH, NW; 2010

update KR, KO) independently reviewed all potentially relevant

studies. Discrepancies were resolved through discussion, or when

needed, by an arbitrator.

Data extraction and management

We extracted data using a structured form that captured patients

status (inpatient or outpatient) and intervention and control char-

acteristics, such as type of drug, route of administration and

dosage. Additionally, we collected data on the primary outcome

measures: change from baseline clinical croup scores; return visits

and/or (re)admissions; length of stay in hospital or accident and

emergency (hours); patient improvement (yes or no); and the use

of additional interventions such as epinephrine, supplemental glu-

cocorticoids, mist treatment, intubation or antibiotic treatment.

One review author (original review MA; 2004 update KR; 2010update KO) extracted data and a second review author (original

review TPK; 2004 update KC, NW; 2010 update KR) checked

the results for accuracy.

We extracted data from graphs (Fitzgerald 1996; Geelhoed 1995a;

Geelhoed1995 C&D; Husby 1993; Kuusela 1988; Roberts 1999)

and obtained additional data from two trial authors (Bjornson

2004;Luria 2001).

Assessment of risk of bias in included studies

We assessed methodological quality of the five studies that were

added tothis reviewusing theRiskof biastool.Thistool examines

the adequacy of the following domains: allocation sequence gen-

eration, allocation concealment, blinding, management of incom-

plete outcome data, selective outcome reporting,and other sources

of bias (Higgins 2009). Sponsorship of studies was recorded as

pharmaceutical company, other sources, or not mentioned (Cho

1996). Two review authors (original review MA, JK; 2004 update

KR, MA, AS, NW, LH; 2010 update KR, KO) independently

assessed quality. Differences were resolved by consensus.



8/9/2019 crup 3.pdf

9/109

Measures of treatment effect

The croup score was reported as either the Westley score, the

telephone outpatient score (TOP) score or various author-created

scales, hence standardized mean differences (SMDs) were used

to combine the outcome for any croup score. A treatment effect

(difference between treatment means) divided by its measurementvariation (for example, a pooled standard deviation) gives a SMD.

Mean differences in Westley scores were calculated for the primary

outcome and consequently used in the subgroup and sensitivity

analyses because estimates between Westleyand other croup scores

were found to be significantly different (seeResults).

We expressed length of stay as mean differences and calculated an

overall mean difference. We calculated risk ratios for binary data

(i.e. return visits and/or (re)admissions and improvement in signs

andsymptoms). Riskdifferences werederived whereoutcomes had

zero events in both groups (namely, the use of various additional

treatments). We calculated the number needed to treat (NNT) or

harm (NNH) for significant results. We reported the pooled base-

line rates (i.e. the control group rates) using the inverse variancemethod. We calculated NNTs or NNHs for this complete range

when the baseline rates varied extensively.

Unit of analysis issues

We calculated the change from baseline measures in cases where

change from baseline measures were not reported directly (

Geelhoed 1995a; Geelhoed 1995 C&D; Husby 1993;Johnson

1996;Klassen 1994;Klassen 1998;Koren 1983;Kuusela 1988;

Leipzig 1979; Martinez 1993; Roorda 1998; Super 1989; von

Muhlendahl 1982). We pooled counts, means and variances when

studies contained more than one treatment group (Geelhoed

1995a; Geelhoed 1995 C&D; Johnson 1998 A&B;Luria 2001)in the glucocorticoid versus placebo comparison.

Many variance imputations were performed according to the work

of Follman (Follmann 1992). The variance of a change from

baseline measure was derived from the common variance of a

single croup score assuming a correlation of 0.5 between pre-

and post-treatment scores (Cruz 1995; Johnson 1996; Klassen

1998; Martinez 1993; Roorda 1998; Super 1989; von Muhlendahl

1982). Variances were also derived or imputed from the measure-

ment of confidence intervals assuming a 0.5 correlation between

pre- and post-treatment scores (Fitzgerald 1996;Klassen 1998),

as well as range and inter-quartile range calculations (Cruz 1995;

Johnson 1996; Roorda 1998; Super1989). Average variances from

similar studies (same comparison, outcome and inpatient/outpa-

tient status) were substituted where no imputable information

was present (Geelhoed 1995a; Geelhoed 1995 C&D; Godden

1997;Kuusela 1988).

Medians were substituted for means in three studies (Cruz 1995;

Johnson 1996;Super 1989). Time points close to six, 12 or 24

hours were used when our exact time points were not provided

(for six hours:Cruz 1995;Fitzgerald 1996;Godden 1997;Husby

1993; Johnson 1996; Klassen 1996c; Klassen 1998; Massicotte

1973);for12hours: Massicotte 1973;andfor24hours,Massicotte

1973;von Muhlendahl 1982).

Assessment of heterogeneity

We assessed heterogeneity quantitatively with the Chi2 test for

heterogeneity (significance cutoff at P = 0.10) and the I 2 statis-

tic (Higgins 2002). The I2 statistic indicates the percent variabil-

ity due to between study (or inter-study) variability as opposed

to within study (or intra-study) variability. An I2 statistic value

greater than 50% shouldbe considered large.For the analyses of all

outcomes, we used random-effects models to combine treatment

effects regardless of quantified heterogeneity.

Subgroup analysis and investigation of heterogeneity

We explored heterogeneity between studies using subgroup and

sensitivity analyses performed on the primary outcomes of changein Westley scores from baseline to six hours and return visits and/

or (re)admissions using meta-regression. We considered the fol-

lowing subgroups: type and route of glucocorticoid administered,

quality (i.e. allocation concealment, funding), inpatient or outpa-

tient status.

In addition to funnel plots, we used the rank correlation test (Begg

1994) and weighted regression (Egger 1997) for the detection of

publication bias (Light 1984). Adjustment for publication bias in

the pooled estimates was performedusing the trim and fillmethod

(Taylor 1998). We used more than one method since the relative

merits of the methods are not well established.

Sensitivity analysis

We considered fixed-effect models in sensitivity analyses.

R E S U L T S

Description of studies

See: Characteristicsof included studies; Characteristicsof excluded

studies.

Results of the search

In this update we identified 35 potentially relevant records from

the searches.

We identified 124 studies as being potentially relevant. In this

update, we added seven unique studies and there are 38 stud-

ies in this review (seeCharacteristics of included studies; also at



8/9/2019 crup 3.pdf

10/109

www.bmj.comfor the original review). However, three of the in-

cluded studies had multiple treatment arms resulting in multi-

ple comparisons; these comparisons are differentially referenced

(Geelhoed 1995 C&D; Johnson 1998 A&B; Skowron 1966 A&

B). Thus, we analyzed 41 comparisons.

Included studies

All of the included trials were of parallel design. Thirty-eight of

the included studies were published in English, one in French

(Massicotte 1973),one inSpanish (Martinez 1993) andanotherin

Danish (Vad Pedersen 1998). Twenty-nine studies were placebo-

controlled and 18 trials examined more than one active treatment

(seeCharacteristics of included studies). Eighteen trials were con-

ducted on inpatients and the remaining 23 were conducted on

outpatients. Dexamethasone was evaluated in 31 trials and against

an active treatment in 17 of these. Dexamethasone was adminis-

tered intramuscularly in 13 studies, orally in 13, sub-cutaneously

in two, and two studies delivered dexamethasone via a nebulizer.There were two factorial studies examining placebo, epinephrine,

dexamethasone and epinephrine/dexamethasone (Kuusela 1988;

Martinez 1993). Five trials examined budesonide versus placebo

only and budesonide was compared to an active treatment in 13

studies; two studies had dexamethasone as a co-intervention in the

budesonide and placebo arm (Geelhoed 2005; Klassen 1996c).

Two studies had three treatment arms: budesonide versus dexam-

ethasone versus dexamethasone/budesonide (Klassen 1998) and

mist and dexamethasone versus epinephrine and dexamethasone

versus epinephrine and budesonide (Duman 2005). One study

had four treatment arms: budesonide versus oral dexamethasone

versus intramuscular dexamethasone versus placebo (Cetinkaya

2004). Prednisolone wascomparedto placebo in three trials(Eden

1964; Massicotte 1973; Tibballs 1992) and compared to dex-

amethasone in two other studies (Fifoot 2007;Sparrow 2006).

One study examined fluticasone versus placebo (Roorda 1998)

while another study compared dexamethasone and betamethasone

(Amir 2006). One study examined three different doses of dex-

amethasone (Geelhoed 1995a) and four studies compared differ-

ent routes of administrationfor dexamethasone (Donaldson 2003;Luria 2001;Rittichier 2000;Vad Pedersen 1998). Three studies

included only childrenwith mild croup(Bjornson 2004; Geelhoed

1996a;Luria 2001). Studies tended to be small with a median of

73 (inter-quartile range 53 to 99) participants.

Two previouslyvalidatedcroup scores were used.The first, Westley

score (Westley 1978) is a 17-point scale to assess air entry (two

points), stridor (two points), intercostal retractions (indrawing

of the chest wall between the ribs on inspiration) (three points),

cyanosis (five points) and level of consciousness (five points). The

Westley score was used in 12 trials to measure baseline and post-

treatment croup scores. The second validated croup score was the

Telephone Out Patient (TOP) that was validated to be used over

the telephone and it was used in one trial (Bjornson 2004). TheTOP score is a three-point telephone outpatient score that assesses

the presence of stridor and barky cough by asking parents about

their childs symptoms in the last 24 hours.

Excluded studies

We excluded a total of 82 studies. The precise reasons for exclusion

are described in theCharacteristics of excluded studiestable.

Risk of bias in included studies

Figure 1andFigure 2describe the quality parameters of the in-

cludedstudies.All studies weredescribedas randomized controlled

trials.


http://www.bmj.com/http://www.bmj.com/

8/9/2019 crup 3.pdf

11/109

Figure 1. Risk of bias graph: review authors judgements about each risk of bias item presented as

percentages across all included studies.



8/9/2019 crup 3.pdf

12/109

Figure 2. Risk of bias summary: review authors judgements about each risk of bias item for each included

study.



8/9/2019 crup 3.pdf

13/109

Allocation

Allocation concealment was adequate in 19 (44%) of the studies,

inadequate in one (2%), and unclear in 21 (46%).

Blinding

The majority of the studies were described as double-blind (n =

35).

Other potential sources of bias

Pharmaceutical sponsorship was identified in five (12%) studies,

support was from other sources in six (15%), funding source was

not mentioned in 27 (66%) and two studies reported that the

study was conducted without funding (5%).

Effects of interventions

Comparison: any glucocorticoid versus placebo

Twenty-four studies involving 2878 participants compared a glu-

cocorticoid to a placebo. These glucocorticoids were dexametha-

sone (n = 15), budesonide (n = six), methylprednisolone(n = three)

and fluticasone (n = one).

Croup score

Overall differences including any croup scores gave similar results,

smaller in magnitude but the same direction. At six hours, thestandardized mean difference (SMD) was -0.6, 95% confidence

interval (CI)-0.8to -0.4; I2 statistic 67% (Analysis 1.1).At12and

24 hours, the SMDs were -0.7, 95% CI -1.0 to -0.3; I 2 statistic

74% (Analysis 1.2) and -0.6, 95% CI -1.1 to 0; I2 statistic 83%

(Analysis 1.3), respectively.

Themost frequent outcome utilized in nine studies wasthe clinical

croup score based on a 17-point ordinal scale developedby Westley

(Westley 1978). Other scoring systems, only one other of which

have been validated, were utilized in seven studies; in eight studies

no clinical score was reported.

Westley scores at six hours showed significantly greater differences

then the other croup scores combined (meta-regression P = 0.02).

All croup scores combined at six hours attributed 67% of thetotal variability to between study variability; combined Westley

scores attributed 23% of the total variability to between study

variability. The Westley and TOP score are the only methods that

have undergone validation and reliability testing and been shown

to be sensitive to important changes in a patients clinical status.

The smaller treatment effect noted with non-Westley scores could

be the result of insensitivity to change or perhaps a greater degree

of variability caused by low reliability.

The mean difference in improvement in the Westleyscorebetween

any glucocorticoid treatment and placebo at six hours was -1.2,

95% CI -1.6 to -0.8; I2 statistic 41% (Analysis 1.4). At 12 hours

this difference was MD -1.9, 95% CI -2.4 to -1.3; I2 statistic 37%

(Analysis 1.5). This change was not significant at 24 hours (MD -

1.3, 95% CI -2.7 to 0.2; I2 statistic 82%,Analysis 1.6). However,

fewer participants were evaluated at 24 hours and hence the lack

of significance may be a reflection of reduced statistical power or a

reflection of large between studyvariability (82%). The magnitude

of change of one is similar to that seen at earlier evaluation points

but the 95% CI crosses zero. A decrease in Westley score of one

from baseline is thought to be a clinically important change. The

fixed-effect estimate at 24 hours was significant.

Meta-regression found no differences in the Westley score between

different glucocorticoid types (P = 0.107). The improvement inthe Westley croup score at six hours was a mean difference of -

1.3, 95% CI -1.7 to -0.9; I2 statistic 0% (Analysis 1.7(1.7.2))

for dexamethasone versus placebo; this was similar to budesonide

versus placebo (MD -1.4; 95% CI -2.1 to -0.7; I2 statistic 55%,

Analysis 1.7(1.7.1)). One study that compared fluticasone versus

placebo had non-significant findings (MD 0.1; 95% CI -1.4 to

1.6,Analysis 1.7(1.7.3)).

Improvement

Improvement was defined in a variety of ways, such as relief of

symptoms, a dichotomization of a croup score or by a minimal in-

cremental improvement in croup score. Heterogeneity, assessed bythe I2 statistic was not large; all I2 statistics were less than 60% for

each time point. At six hours, the RR for improvement was 1.4,

95% CI 1.1 to 1.8; I2 statistic 26% (Analysis 1.14) with a NNT

of five (three to 11). The baseline rate of clinical improvement was

45% (39% to 52%). At 12 hours the RR was 1.3, 95% CI 1.1 to

1.6; 52% (Analysis 1.15) with a NNT of five (three to 10). The

baseline rate of clinical improvement was 62% (55% to 69%). At

24 hours, the RR was 1.2, 95% CI 1.1 to 1.5; I2 statistic 31%

(Analysis 1.16) and the NNT was five (three to 13). The baseline

rate of clinical improvement was 86% (80% to 92%). The degree

of benefit of a NNT of five patients would be sufficient to support

the use of glucocorticoids over placebo.

Return visits and/or (re)admissions

There was a significant decrease in the rate of return visits and/or

(re)admissions for patients treated with glucocorticoid versus pa-

tients treated with placebo (RR 0.5, 95% CI 0.3 to 0.7, I2 statistic

49%,Analysis 1.10). The range of baseline rates was 2% to 72%.



8/9/2019 crup 3.pdf

14/109

Table 1 gives the NNT for this range of rates. The average (inverse

variance method) baseline rate was 12% (10% to 14%). This cor-

responds to a NNT of 17 (13 to 28). Total variability attributable

to between-study variation was not too large (I2 statistic 49%).

Nevertheless,the definitions for return visits and/or (re)admissions

varied between studies, for example, re-admittance for inpatientsversus unscheduled visits to a primary care giver or emergency

room. In any case, the effect remains consistent in direction.

Length of stay

Overall, a significantly shorter time was spent in accident and

emergency (Leipzig 1979) and hospital (nine studies) when chil-

dren were treated with a glucocorticoid as compared with placebo

(825 patients); the mean difference was -12 (-19 to -five) hours

(Analysis 1.13). Total variability attributable to between-study

variation was large (I2 statistic 64%). Two inpatient studies

(Roorda 1998;Super 1989) did not favour glucocorticoids, but

both were very small and had insignificant results; the remaining

studies were consistent in direction.

Use of additional interventions

All but one random-effects model showed no significant differ-

ences for use of any particular additional intervention for any

glucocorticoid versus placebo. Children treated with a glucocor-

ticoid were 10% (1% to 20%; I2 statistic 55%, Analysis 1.17.4)

less often treated additionally with epinephrine. The NNT was

10 (five to 100). In the sensitivity analyses, use of mist treatment

for the fixed-effect model was significant and favoured any gluco-

corticoid. Additional use of mist treatment was measured in two

studies (Johnson 1996;Super 1989); their respective RDs were -

12% (95% CI -19% to -6%) and -15% (95% CI -25% to -4%).

Their baseline rates were 16% (12% to 21%) and 11% (6% to

16%). Use of mist treatment was quite heterogeneous (I2 statis-

tic 95%) between the two included studies and this result should

be interpreted with caution. There was no difference in the need

for intubation or tracheostomy between the two groups (RD 0%,

95% CI -1% to 1%; I2 statistic 0%).

Subgroup and sensitivity analyses

In the subgroup analyses (between study comparisons), all quality

items (funding, methodological quality, allocation concealment)

were insignificant for both Westley score at six hours and return

visits and/or (re)admissions. Clinical setting (inpatient versus out-

patient), route of administration (oral, spray, intramuscular, sub-

cutaneous), and type of glucocorticoid were also not significantly

different for both primary outcomes. However, fluticasone was

close to being significantly less beneficial than dexamethasone and

budesonide (P = 0.11) forthe Westley score at sixhours. This study

(Roorda 1998) was small (n = 17); the meta-regression was likely

underpowered to detect a difference, if one existed. This study did

not record return visits or (re)admissions. Three studies included

only mild croup patients (Bjornson 2004;Geelhoed 1996a;Luria

2001) and compared dexamethasone to placebo. One study used

0.6 mg/kg of oral dexamethasone (Bjornson 2004), one study ad-

ministered 0.15 mg/kg of oral dexamethasone (Geelhoed 1996a)

andthe final study examined 0.60 mg/kg of oral dexamethasone ornebulized dexamethasone (Luria 2001).Bjornson 2004included

children with a Westley croup score of 3) who were

treated in the emergency department

Interventions Oral 0.6 mg/kg dexamethasone (n = 36) or oral 0.15 mg/kg dexamethasone (n = 36)

Outcomes Westley Croup score

Re-visit and/or (re)-admissions

Notes

Risk of bias

Bias Authors judgement Support for judgement

Random sequence generation (selection

bias)

Low risk

Allocation concealment (selection bias) Low risk

Blinding (performance bias and detection

bias)

All outcomes

Low risk

Incomplete outcome data (attrition bias)

All outcomes

Low risk

Selective reporting (reporting bias) Low risk

Other bias Low risk

Pharmaceutical funding? Unclear risk Not mentioned

Amir 2006

Methods Randomized controlled trial. No drop-outs

Participants 52 children aged 6 months to 6 years with mild to moderate croup who were treated in

the emergency department (Westley 1-11)

Interventions Intramuscular 0.6 mg/kg dexamethasone (n = 26) or oral 0.4 mg/kg betamethasone (n

= 26)



8/9/2019 crup 3.pdf

28/109

Amir 2006 (Continued)

Outcomes Westley Croup score

Additional treatment in the ED

ED hospitalizations rateFollow-up visits

Additional treatment

Notes

Risk of bias



bias)

Low risk

Allocation concealment (selection bias) Unclear risk Unclear


Other bias Low risk

Pharmaceutical funding? High risk No funding

Bjornson 2004

Methods Randomized by the pharmacy, placebo-control led, double-blind trial. 13were lostto fol low

up

Participants 720childrenaged3 monthsto9 yearswithmildcrouptreatedintheemergencydepartment

(Westley

8/9/2019 crup 3.pdf

29/109

Bjornson 2004 (Continued)

Pharmaceutical funding? High risk Other funding

Cetinkaya 2004

Methods Randomized, placebo-controlled trial

Participants 60 children aged 6 to 36 months who presented to the emergency department with croup

Interventions Nebulised budesonide 500 ug and placebo (n = 15) or intramuscular dexamethasone 0.

6 mg/kg and placebo (n = 15) or oral dexamethasone 0.6 mg/kg and placebo (n = 15) or

nebulized salbutamol and placebo (n = 15)

Outcomes Westley croup score

Notes

Risk of bias




Cruz 1995

Methods Randomized by the pharmacy, placebo-controlled, double-blind trial. 3 children required

hospitalizations, 2 withdrew and 2 were lost to follow up

Participants 45 children aged 6 months to 5 years with moderate croup in the emergency department.

Exclusion: patients who required more than one racemic epinephrine treatment

Interventions Placebo (n = 19) or intramuscular dexamethasone 0.6 mg/kg (n = 19), unique dose. Dex-

amethasone maximum dose of 10 mg

Outcomes Re-presented and re-admitted to the hospital

Notes

Risk of bias






8/9/2019 crup 3.pdf

30/109

Donaldson 2003

Methods Randomized by the pharmacy, controlled, double-blind trial. 1 child withdrew (treatment

group and reason not stated)

Participants 96 children aged 3 to 84 months with croup (Westley >= 2) in the emergency departmentwere enrolled (collected complete follow-up data for 95)

Exclusion: croup score < 2, epiglottitis, bacterial tracheitis, foreign body, chronic lung dis-

ease, severe co-morbidities, inability of parents to give informed consent, or glucocorticoids

therapy within 4 weeks of presenting

Interventions Intramuscular dexamethasone 0.6 mg/kg (n = 49) or oral dexamethasone 0.6 mg/kg (n =

46)

Outcomes Parental assessment of symptom relief at 24 hours

Presence of stridor, barky cough, expiratory sounds and normal sleeping patterns at 24

hours

Unscheduled visits

All symptoms absent at 24 hours and 4 days

Co-interventions: mist tent

Notes

Risk of bias


Allocation concealment (selection bias) Low risk Adequate


Duman 2005

Methods Randomized controlled trial

Participants 76 children aged >= 6 months with croup (Westley >= 2) in the emergency department

Interventions Cool mist therapy and intramuscular 0.6 mg/kg dexamethasone (n = 26) or 2.5 mL

for 0 to 20 kg or 5.0 mL for 20 to 40 kg nebulized epinephrine and intramuscular

dexamethasone 0.6 mg/kg (n = 31) or 2.5 mL for 0 to 20 kg or 5.0 mL for 20 to 40 kg

nebulized epinephrine and nebulized budesonide 2 mg (n = 19)

Outcomes Croup score at 0.5, 1, 1.5 and 2 hoursHospitalization

Readmitted

Co-intervention: epinephrine

Notes Quality = 1

Risk of bias



8/9/2019 crup 3.pdf

31/109

Duman 2005 (Continued)


Random sequence generation (selectionbias)

Low risk



bias)

All outcomes

Low risk


All outcomes

Low risk ITT analysis


Other bias Low risk

Pharmaceutical funding? High risk No funding

Eden 1964

Methods Randomized, placebo-controlled, double-blind trial. 3 drop-outs

Participants 50 children hospitalized with croup

Interventions Placebo (n = 25) or intramuscular methyl-prednisolone 4 mg/kg/day (n = 25)

Outcomes Percentage of improvement

Notes

Risk of bias






8/9/2019 crup 3.pdf

32/109

Eden 1967

Methods Randomized, placebo-controlled, double-blind study


Interventions Placebo (n = 25) or intramuscular dexamethasone 0.4 mg/kg/day (n = 25)

Outcomes Clinical score based on a multi-item scale (stridor, dyspnoea, retraction, cyanosis) at 0, 6,

12,18 and 24 hours

Co-intervention: intubation

Notes

Risk of bias




Fifoot 2007

Methods Randomized, double-blind study

Participants 99 children aged 6 months to 6 years who present to the emergency department with

croup (2 or more on Westley croup score). 86 parents/guardians were available for 1

week telephone follow up

Interventions Oral prednisolone 1 mg/kg (n = 34), oral dexamethasone 0.15 mg/kg (n = 34) or oral

dexamethasone 0.60 mg/kg (n = 31). Single doses

Outcomes Westley croup score at 0, 1, 2, 3 and 4 hours

Admitted

Requiring salvage therapy (nebulized epinephrine)

Follow up: return to medical care with croup or the need for further treatment with

steroids in the week following discharge

Notes

Risk of bias



bias)

Low risk




8/9/2019 crup 3.pdf

33/109

8/9/2019 crup 3.pdf

34/109

Geelhoed 1995 C&D (Continued)

Outcomes

Notes

Risk of bias


Allocation concealment (selection bias) Unclear risk SeeGeelhoed 1995candGeelhoed 1995d

Pharmaceutical funding? Unclear risk SeeGeelhoed 1995candGeelhoed 1995d

Geelhoed 1995a

Methods Randomized, controlled, double-blind trial. 1 withdrew

Participants 120 patients aged 6 months to 13 yearspresenting to the emergency department with croup

Interventions Oral dexamethasone 0.6 mg/kg (n = 31) or oral dexamethasone 0.3 mg/kg (n = 29) (trial

A). Single doses. Oral dexamethasone 0.15 mg/kg (n = 29) or oral dexamethasone 0.3 mg/

kg (n = 31) (trial B)

Outcomes Clinical score based on multi-item scale (stridor and retraction) at 0, 1, 2, 3, 4 and 8

Length of stay

Return visits/re-admissions

Notes

Risk of bias




Geelhoed 1995c

Methods Randomized, placebo-control led, double-bl ind. 11 withdrawals (7 children from the

placebo group)

Participants 80 childrenaged 3 months to12 yearspresenting to the emergency department with croup

and then admitted to the hospital

Interventions Placebo (n = 30) or nebulized budesonide 2 mg single dose (n = 27) or oral dexamethasone

0.6 mg/kg maximum 12 mg single dose (n = 23)

Budesonide branch



8/9/2019 crup 3.pdf

35/109

Geelhoed 1995c (Continued)

Outcomes Clinical score based on multi-item scale (stridor, retractions modified-modified Westley)

at 0, 1, 3, 6, 9 and 12 hours

Length of stayCo-interventions: racemic epinephrine


Notes

Risk of bias




Geelhoed 1995d

Methods

Participants

Interventions Dexamethasone branch

Outcomes

Notes

Risk of bias




Geelhoed 1996a

Methods Randomized, placebo-controlled, double-blind trial. 4 patients were lost to follow up

Participants 100 children aged 0.4 to 11 years presenting with mild croup to emergency department

Interventions Placebo (n = 50) or oral dexamethasone 0.15 mg/kg single dose (n = 50)

Outcomes Return visits/re-admissions

Notes



8/9/2019 crup 3.pdf

36/109

Geelhoed 1996a (Continued)

Risk of bias




Geelhoed 2005

Methods Randomized trial. Children and all staff except nurses administering the drug were blinded

to treatments. 1 lost to follow up

Participants 71 children older than 3 months who presented to the emergency department with croup

requiring admission

Interventions Placebo and dexamethasone 0.15 mg/kg (n = 36) or nebulized budesonide 2 mg and

dexamethasone 0.15 mg/kg (n = 36)

Outcomes Croup score

Length of stay


Duration of symptoms

Notes

Risk of bias




Godden 1997

Methods Randomized, placebo-controlled, double-blind trial with 7 withdrawals

Participants 87 children aged 0.6 to 10 years hospitalized with croup

Interventions Placebo (n = 42) or nebulized budesonide 2 mg initially and 1 mg every 12 hours (n = 47)

Outcomes Clinical score of a multi-item scale (Westley modified) at 0, 0.5, 1, 2, 4, 12 and 24 hours.

Co-interventions: racemic epinephrine, intubation

Length of stay



8/9/2019 crup 3.pdf

37/109

Godden 1997 (Continued)

Notes

Risk of bias




Husby 1993

Methods Randomized, placebo-controlled, double-blind trial

Participants 37 children aged 0.4 to 4.9 years hospitalized with croup. 1 drop-out before placebo wasadministered

Interventions Placebo (n = 16) or nebulized budesonide 1 mg twice (n = 20)

Outcomes Clinical score based on multi-item scale (Westley modified). Average differences between

0 and 6 hours

Co-intervention: antibiotics

Notes

Risk of bias




James 1969


Participants 88 children (median age of 14 months) hospital ized with croup

Interventions Placebo (n = 43) or intramuscular dexamethasone 4 to 12 mg single dose (n = 45)

Outcomes Co-intervention: antibiotics, tracheostomy

Notes

Risk of bias



8/9/2019 crup 3.pdf

38/109

James 1969 (Continued)



Pharmaceutical funding? Low risk Pharmaceutical support

Johnson 1996

Methods Block randomization by the hospital pharmacy, placebo-controlled, double-blind trial

Participants 55 children aged 9 months to 2.5 years presenting to the emergency department with

moderate croup

Interventions Placebo (n = 27) or nebulized dexamethasone 10 to 20 mg (n = 28)

Outcomes Clinical score of a multi-item scale (Westley modified) at 0, 2, 4 and last score


Co-intervention: mist-tent, supplemental steroids

Notes

Risk of bias




Johnson 1998 A&B

Methods SeeJohnson 1998aandJohnson 1998b

Participants

Interventions

Outcomes

Notes

Risk of bias


Allocation concealment (selection bias) Unclear risk SeeJohnson 1998aandJohnson 1998b



8/9/2019 crup 3.pdf

39/109

Johnson 1998 A&B (Continued)

Pharmaceutical funding? Unclear risk SeeJohnson 1998aandJohnson 1998b

Johnson 1998a

Methods Block randomization (by the pharmaceutical company), placebo-controlled, double-blind

trial

Participants 144 children aged from 3 months to 9 years with moderate croup presenting to the emer-

gency department

Interventions Placebo (n = 49) or nebulized budesonide 4 mg single dose (n = 48) or intramuscular

dexamethasone 0.6 mg/kg single dose (n = 47)

Dexamethasone branch

Outcomes Clinical score based on a multi-item scale (Westley modified) at 0 and 2 hoursReturn visits/re-admissions

Co-interventions: racemic epinephrine, intubation

Notes

Risk of bias




Johnson 1998b

Methods

Participants

Interventions Budesonide branch

Outcomes

Notes

Risk of bias





8/9/2019 crup 3.pdf

40/109

Johnson 1998b (Continued)


Klassen 1994

Methods Block randomization by the pharmacy department, placebo-controlled, double-blind trial.

4 patients were lost to follow up

Participants 54 children aged from 3 months to 5 years presenting to the emergency department with

mild to moderate croup

Interventions Placebo (n = 27 ) or nebulized budesonide 2 mg (n = 27). Single dose

Outcomes Clinical score based on a multi-item scale (Westley modified) at 0 and 4 hours. Co-inter-

vention: supplemental steroids


Notes

Risk of bias




Klassen 1996

Methods Blockrandomizationbypharmacydepartment,controlled,double-blindtrialwith amatch-

ing placebo

Participants 50 children aged from 3 months to 5 years presenting to the emergency department with

mild to moderate croup

Interventions Placebo plus oral dexamethasone 0.6 mg/kg (n = 25) or 2 mg nebulized budesonide plus

oral dexamethasone 0.6 mg/kg (n = 25). Unique doses

Outcomes Clinical score based on multi-item scale (Westley modified) at 0 and 4 hours

Co-intervention: racemic epinephrine


Notes

Risk of bias




8/9/2019 crup 3.pdf

41/109

Klassen 1996 (Continued)



Klassen 1998

Methods Block randomization by the hospital pharmacy, controlled, double-bl ind trial

Participants 198 children with croup presenting to the emergency department

Interventions Nebulised budesonide 2 mg single dose (n = 65) or oral dexamethasone 0.6 mg/kg single

dose (n = 69) or budesonide + dexamethasone (n = 64)

Outcomes Clinical score based on multi-item scale (Westley modified)

Return visits/re-admissionsLength of stay

Co-interventions: racemic epinephrine, supplemental steroids

Notes

Risk of bias




Koren 1983

Methods Randomized, placebo-controlled, double-blind trial. No description of drop-outs

Participants 78 children aged from 8 months to 8 years hospital ized with croup

Interventions Placebo (n = 38) or intramuscular dexamethasone 0.6 mg/kg single dose (n = 40)

Outcomes Respirations per minute at 0, 1, 2, 3, 4, 5 and 6 hours

Notes

Risk of bias





8/9/2019 crup 3.pdf

42/109

Koren 1983 (Continued)


Kuusela 1988

Methods Randomized, placebo-controlled, double-blind trial; pre-numbered unlabelled ampoules

in random order

Participants 72 children from 6 months to 11 years of age hospitalized with croup

Interventions Placeboplusplacebo(n= 21)ornebulizedL-epinephrine2.25%0.25ml/kgplusplacebo (n

= 16) or intramuscular dexamethasone 0.6 mg/kg plus placebo (n = 16) or dexamethasone

plus L-epinephrine (n = 19). L-epinephrine every 2 hours if necessary. Dexamethasone

maximum dose of 10 mg

Outcomes Clinical score based on dyspnoea and cough scale at 0, 6, 12 and 24 hoursLength of stay

Notes

Risk of bias




Leipzig 1979


Participants 30 childrenaged from8 months to 6 yearspresenting with croup to the emergency depart-

ment

Interventions Placebo (n = 14) or intramuscular dexamethasone 0.3 mg/kg twice (n = 16)

Outcomes Clinical score based on a multi-item scale (stridor, cyanosis, retractions, respiration rate,

pulse rate) at 0 and 24 hours

Length of stay

Co-interventions: intubation

Notes

Risk of bias




8/9/2019 crup 3.pdf

43/109

Leipzig 1979 (Continued)



Luria 2001


Participants 264 children aged 6 months to 6 years presenting to the emergency department with mild

croup of less than 48 hours

Interventions Placebo (n = 88) or 0.6 mg/kg of oral dexamethasone (n = 85), or 160 ug of nebulized

dexamethasone sodium phosphate (n = 91)

Outcomes Treatment failure (requiring corticosteroids or racemic epinephrine up to 7 days aftertreatment)


Parentals assessments of childs condition (improvement)

Notes

Risk of bias




Martinez 1993



Interventions Placebo (n= 17) or nebulized L-epinephrine 0.14% quid if necessary (n= 15) or intramus-

cular dexamethasone 0.5 mg/kg single dose (n = 16) or L-epinephrine+dexamethasone (n

= 18)

Outcomes Clinical score based on multi-item scale (stridor, dyspnoea, cough, cyanosis) at 0, 6,12,18

and 24 hours

Notes Language = Spanish

Risk of bias



8/9/2019 crup 3.pdf

44/109

Martinez 1993 (Continued)




Massicotte 1973

Methods Randomized by table of random numbers, matching placebo-controlled, double-blind trial

Participants 42 children less than 4 years hospitalized with croup

Interventions Placebo (n = 17) or intravenous methyl-prednisolone 4 mg (n = 25) at 0, 4 and 8 hours.

Total dose day = 12 mg

Outcomes Percentage of improved patients at 4, 8, 14 and 20 hours

Notes Language = French

Risk of bias




Rittichier 2000

Methods Randomized (computer-generated) controlled trial. 27 children withdrew (unable to be

contacted by phone, treatment arm not specified)

Participants 277 children a ged 3 m onths to 1 2 y ears w ith moderate croup o f < 48 hours onset presenting

to the emergency department

Interventions Intramuscular dexamethasone 0.6 mg/kg (n = 139) or oral dexamethasone 0.6 mg/kg (n =

138). Maximum dosage of both routes of administration was 8 mg

Outcomes Unscheduled return visitCaretakers report of symptoms (improvement)

Additional treatments given at unscheduled visits

Co-interventions: all children received cool mist therapy

Notes

Risk of bias



8/9/2019 crup 3.pdf

45/109

Rittichier 2000 (Continued)




Roberts 1999

Methods Randomized by a computer-generated random numbers table, double-blind, placebo con-

trolled trial. Randomization code kept at Astra Draco (gave funding). 1 child withdrew

Participants 82 children hospitalized with croup (0.5 to 8 years)

Interventions Nebulised placebo (n = 40) or nebulized budesonide (n = 42). Dose of 2 mg/4 mL every

12 hours for a maximum of four doses

Outcomes Croup score similar to Leipzig 1979with alterations to stridor assessment, oxygen saturation

and temperature

Co-interventions: intubation, use or epinephrine


Notes

Risk of bias




Roorda 1998

Methods Randomized, double-blind, placebo-controlled trial (prel iminary study)

Participants 17 children hospitalized with croup (4 to 52 months)

Interventions Metered dose inhaler placebo (n = 8) or metered dose inhaler fustians with a dose of 2000

ug/day (n = 9)

Outcomes Westley croup score up to 24 hours

Co-interventions: supplemental corticosteroids, intubation

Length of stay

Notes



8/9/2019 crup 3.pdf

46/109

Roorda 1998 (Continued)

Risk of bias




Skowron 1966 A&B

Methods SeeSkowron 1966aandSkowron 1966b

Participants

Interventions

Outcomes

Notes

Risk of bias


Allocation concealment (selection bias) Unclear risk SeeSkowron 1966aandSkowron 1966b

Pharmaceutical funding? Unclear risk SeeSkowron 1966aandSkowron 1966b

Skowron 1966a


Participants 200 children (average age of 2.3 years) hospitalized with croup

Interventions Placebo (n = 44) or subcutaneous dexamethasone 6 mg single dose (n = 56) (Trial a)

Placebo (n = 53) or subcutaneous dexamethasone 4 mg single dose (n = 41) (Trial b)

Branch 6 mg

Outcomes Co-intervention: tracheostomy


Notes

Risk of bias




8/9/2019 crup 3.pdf

47/109

Skowron 1966a (Continued)



Skowron 1966b

Methods

Participants

Interventions Branch 4 mg

Outcomes

Notes

Risk of bias




Sparrow 2006

Methods Randomized by the pharmacy, double-blind controlled trial. No drop-outs

Participants 133 children a ged 3 to 1 42 m onths w ith m ild t o moderate croup t reated a t the e mergency

department

Interventions Oral 0.15 mg/kg dexamethasone (n = 68) or oral 1 mg/kg prednisolone (n = 65)

Outcomes Admissions

Duration of croup symptoms

Duration of viral symptoms

Return visits

Additional interventions

Time in ED

Adverse events

Notes

Risk of bias




8/9/2019 crup 3.pdf

48/109

Sparrow 2006 (Continued)


bias)

Low risk



bias)

All outcomes

Low risk


All outcomes

Low risk


Other bias Low risk


Super 1989

Methods Randomized, placebo-controlled, double-blind

Participants 29 children aged from 6 months to 3 years hospital ized with croup

Interventions Placebo (n = 13) or intramuscular dexamethasone 0.6 mg/kg single dose (n = 16)

Outcomes Clinical score based on multi-item scale (Westley modified) at 0, 12 and 24 hours

Percentage of improvement

Co-interventions: supplemental steroids, mist tent

Length of stay

Notes

Risk of bias






8/9/2019 crup 3.pdf

49/109

Tibballs 1992

Methods Randomized, placebo-controlled, double-blind trial. Interventions were supplied by the

pharmacy in an order determined by a table of random numbers. 3 withdrawals from the

placebo group and were excluded from the analysis

Participants 73 hospitalized children aged 6 months to 8 years, were eligible if endotracheal intubation

had been required for upper airway obstruction caused by croup

Interventions Placebo (n = 35) or nasogastric prednisolone 2 mg/kg/day (n = 38), until 24 hours after

extubation

Outcomes Intubation duration

Re-intubation rates

Co-intervention: racemic epinephrine

Notes

Risk of bias




Vad Pedersen 1998

Methods Variable block randomization (2 to 6), non-blinded trial

Participants 59 were i ncluded, but 2 w ere l ater e xcluded, children b etween 3 months to 6 y ears a dmitted

to the hospital with croup

Interventions Inhaled budesonide (n = 28) dose of 1000 ug twice or intramuscular dexamethasone (n =

29) at a dose of 0.6 mg/kg

Outcomes 7-point croup score a 0, 3, 6 and 12 hours

Length of stay

Need for further dexamethasone injections

Notes Language = Danish

Risk of bias






8/9/2019 crup 3.pdf

50/109

von Muhlendahl 1982

Methods Randomized, placebo-controlled, double-blind trial with 57 withdrawals

Participants 406 children aged less than 10 years hospitalized with croup

Interventions Placebo (n = 173) or oral dexamethasone 6 mg (n = 176), unique dose

Outcomes Clinical score based on multi-item scale at 0, 6, 12 and 18 hours

Improvement

Co-interventions: tracheostomy

Length of stay

Notes

Risk of bias




ED = emergency department

ITT = intention-to-treat

Characteristics of excluded studies [ordered by study ID]

Study Reason for exclusion

Adelman 1996 Review

Anene 1996 No diagnosis of croup

Anonymous 1989 Letter to Kairys meta-analysis (1989)

Anonymous 1992 Position statement

Anonymous 1994 Review

Anonymous 1995 Summary of Klassens study (1994)

Anonymous 1996 Review

Asher 1981 Letter to the editor



8/9/2019 crup 3.pdf

51/109

(Continued)

Bass 1972 Letter to Edens study (1971)

Bass 1980 Letter to Leipzigs study (1979)

Baugh 1986 Review

Bauman 1996 Comment about croup management

Berni 1999 Diagnostic and treatment guidelines

Bollobas 1965 No diagnosis of croup

Chub-Uppakarn 2007 All children received epinephrine prior to corticosteroids

Cichy 1983 No diagnosis of croup

Cluster 2005 Review of Bjornsons trial

Connolly 1969 No diagnosis of croup

Couser 1992 No diagnosis of croup

De Boeck 1995 Review

Denoyelle 1992 Comments/review

Doull 1995 Editorial

Eden 1971 Letter to the editor

Feingold 1999 Commentary

Flisberg 1973 No clinical outcomes reported

Folland 1997 Commentary about croup management

Freezer 1990 Retrospective study

Friedman 1990 Review

Geelhoed 1994 Review

Geelhoed 1996b Review

Geelhoed 1997 Review

Goddard 1967 No diagnosis of croup



8/9/2019 crup 3.pdf

52/109

(Continued)

Goldhagen 1983 Review

Haque 1981 Not randomized

Harel 1995 Letter to the editor

Havaldar 1997 No diagnosis of croup

Hawkins 1980 Review

Helowicz 1996 Letter to the editor

Herxheimer 1996 Letter to the editor

Kairys 1989 Meta-analysis

Karoliny 1990 Case series

Kelley 1992 Retrospective study

Klassen 1997a Review

Klassen 1997b Review

Kunkel 1996 Not randomized

Ledwith 1995 Not randomized

Lukina 1991 Review

Lynk 1993 Letter to the editor

Marchi 1993 Review

Martensson 1960 Not randomized

McDonogh 1994 Retrospective study

Mortensen 1994 Duplication of Husbys study (1993)

Novik 1960 Not randomized

Oski 1961 No diagnosis of croup

Osvath 1994 Not randomized

Prendergast 1994 Uncontrolled trial



8/9/2019 crup 3.pdf

53/109

(Continued)

Quan 1992 Review

Richter 1983 Uncontrolled

Rizos

crup 3.pdf

Documents