Nebulized epinephrine for croup in children
Abstract
Background
Croup is a common childhood illness characterized by barky cough, stridor, hoarseness and respiratory distress. Children with severe croup are at risk for intubation. Nebulized epinephrine may prevent intubation.
Objectives
To assess the efficacy (measured by croup scores, rate of intubation and health care utilization such as rate of hospitalization) and safety (frequency and severity of side effects) of nebulized epinephrine versus placebo in children with croup, evaluated in an emergency department (ED) or hospital setting.
Search methods
We searched CENTRAL 2013, Issue 6, MEDLINE (1966 to June week 3, 2013), EMBASE (1980 to July 2013), Web of Science (1974 to July 2013), CINAHL (1982 to July 2013) and Scopus (1996 to July 2013).
Selection criteria
Randomized controlled trials (RCTs) or quasi‐RCTs of children with croup evaluated in an ED or admitted to hospital. Comparisons were: nebulized epinephrine versus placebo, racemic nebulized epinephrine versus L‐epinephrine (an isomer) and nebulized epinephrine delivered by intermittent positive pressure breathing (IPPB) versus nebulized epinephrine without IPPB. Primary outcome was change in croup score post‐treatment. Secondary outcomes were rate and duration of intubation and hospitalization, croup return visit, parental anxiety and side effects.
Data collection and analysis
Two authors independently identified potentially relevant studies by title and abstract (when available) and examined relevant studies using a priori inclusion criteria, followed by methodological quality assessment. One author extracted data while the second checked accuracy. We use the standard methodological procedures expected by the Cochrane Collaboration.
Main results
Eight studies (225 participants) were included. In general, children included in the studies were young (average age less than two years in the majority of included studies). Severity of croup was described as moderate to severe in all included studies. Six studies took place in the inpatient setting, one in the ED and one setting was not specified. Six of the eight studies were deemed to have a low risk of bias and the risk of bias was unclear in the remaining two studies.
Nebulized epinephrine was associated with croup score improvement 30 minutes post‐treatment (three RCTs, standardized mean difference (SMD) ‐0.94; 95% confidence interval (CI) ‐1.37 to ‐0.51; I2 statistic = 0%). This effect was not significant two and six hours post‐treatment. Nebulized epinephrine was associated with significantly shorter hospital stay than placebo (one RCT, MD ‐32.0 hours; 95% CI ‐59.1 to ‐4.9). Comparing racemic and L‐epinephrine, no difference in croup score was found after 30 minutes (SMD 0.33; 95% CI ‐0.42 to 1.08). After two hours, L‐epinephrine showed significant reduction compared with racemic epinephrine (one RCT, SMD 0.87; 95% CI 0.09 to 1.65). There was no significant difference in croup score between administration of nebulized epinephrine via IPPB versus nebulization alone at 30 minutes (one RCT, SMD ‐0.14; 95% CI ‐1.24 to 0.95) or two hours (SMD ‐0.72; 95% CI ‐1.86 to 0.42). None of the studies sought or reported data on adverse effects.
Authors' conclusions
Nebulized epinephrine is associated with clinically and statistically significant transient reduction of symptoms of croup 30 minutes post‐treatment. Evidence does not favor racemic epinephrine or L‐epinephrine, or IPPB over simple nebulization. The authors note that data and analyses were limited by the small number of relevant studies and total number of participants and thus most outcomes contained data from very few or even single studies.
PICOs
Plain language summary
Nebulized epinephrine for croup in children
Croup is a common childhood illness which primarily affects those between the ages of six months and three years, with a peak annual incidence in the second year of life of nearly five per cent. Males and females are affected equally. Croup is most often caused by a viral infection. Symptoms of croup include a hoarse voice, a 'barking' cough and noisy breathing. These symptoms are the result of swelling that occurs in the area of the windpipe (trachea) just below the voice box (larynx). Although most cases of croup are mild and resolve on their own, occasionally the swelling can be severe enough to cause difficulty in breathing. In these children, epinephrine (also called adrenaline) is a medication that is inhaled as a mist to temporarily shrink the swollen area in the trachea.
This review looked at trials of inhaled epinephrine for the treatment of children with croup and is comprised of only eight studies with 225 participants. Of the eight included studies, six were assessed as having low risk of bias and two as unclear risk of bias (based upon assessment of adequate random sequence generation, allocations concealment, blinding of participants and personnel, blinding of outcome assessment, completeness of outcome data, and selective reporting). Studies assessed a variety of outcome measures and few studies examined the same outcomes; therefore, most outcomes contained data from a maximum of three studies, and in some cases only single studies.
Compared to no medication, inhaled epinephrine improved croup symptoms in children at 30 minutes following treatment (three studies, 94 children). This treatment effect disappeared two hours after treatment (one study, 20 children). However, children's symptoms did not become worse than prior to treatment. No study measured adverse events.
The evidence is current to July 2013.
Authors' conclusions
Background
Description of the condition
Croup (laryngotracheobronchitis) is a common respiratory illness of childhood, estimated to affect approximately three per cent of children under six years of age annually (Denny 1983; Johnson 2003). The clinical picture is characterized by the abrupt onset of a distinctive barky cough, which may be accompanied by stridor, hoarse voice and respiratory distress. Croup symptoms are often preceded by non‐specific symptoms such as cough, rhinorrhea and fever. The most common etiology is a viral infection, predominantly parainfluenza virus (Marx 1997).
Description of the intervention
While most children with croup have mild symptoms (defined as presence of a 'barky cough', no audible stridor at rest and no to mild chest wall indrawing) (Johnson 2001; Johnson 2003), a small minority have severe symptoms characterized by marked chest wall indrawing, agitation and lethargy (Johnson 2001; Johnson 2003). These children are at significant risk for intubation. Corticosteroids decrease both the frequency and duration of hospitalization and intubation (Kairys 1989; Russell 2011; Tibballs 1992). A systematic review of corticosteroids found that compared to placebo, corticosteroids significantly reduced six and 12‐hour croup scores, reduced length of hospital stay and reduced return visits (Russell 2011). However, corticosteroids take approximately 30 minutes after treatment to lessen respiratory distress (Dobrovoljac 2012). Consequently in children with the most severe symptoms, a therapy with a rapid rate of onset may lessen the need for intubation. The therapy most commonly used for this purpose is nebulized epinephrine. Children with moderate to severe croup are usually administered both epinephrine and corticosteroids concurrently to reduce respiratory distress while awaiting the effects of the corticosteroid treatment. Alternative therapies are mist and heliox (helium‐oxygen mixture) (Johnson 2005; Vorwerk 2010). However, recent evidence suggests that mist provides no significant clinical benefit (Moore 2007; Neto 2002; Scolnik 2006). Heliox has greater promise, but it is more cumbersome to use, and has little evidence of effectiveness (Johnson 2005).
How the intervention might work
Nebulized epinephrine has been widely used for more than 30 years, after the first report of its effectiveness in 1971 (Adair 1971). It is thought that epinephrine‐induced vasoconstriction decreases upper airway edema (Brown 2002; Johnson 2005; Kaditis 1998; Klassen 1999). A small number of randomized controlled trials (RCTs) have been published (Johnson 2005), along with one systematic review (Bjornson 2011).
Initial studies published in the 1970s and early 1980s suggested that epinephrine might be more effective when nebulized via IPPB (intermittent positive pressure breathing) than by nebulization alone. The use of IPPB has now fallen out of favor and it is no longer routinely used.
Clinical severity in children with croup can be determined by both direct measures (clinical scores, transcutaneous carbon dioxide concentrations, pulsus paradoxus, impedance plesmography, radiographic measurement of tracheal diameter) and indirect ones (rate and duration of intubation, rate and duration of hospitalization, rate of return to seek medical care for ongoing croup symptoms, sleep lost by parents, parental stress). Several objective techniques have been reported (Corkey 1981; Davis 1993; Fanconi 1990; Steele 1998), but none are easy to use, nor measure change across the full range of clinical severity. Consequently they have not been routinely used.
The most widely used direct measures of respiratory severity are different types of croup scores (Bourchier 1984; Corkey 1981; Downes 1975; Geelhoed 1995; Godden 1997; Husby 1993; Leipzig 1979; Massicotte 1973; Taussig 1975; von Muhlendahl 1982; Westley 1978). The Westley croup score has been shown to be reliable (Johnson 1998; Klassen 1994) and, to some extent, valid (Klassen 1994; Klassen 1995). None of the other croup scores, to our knowledge, have been assessed for either validity or reliability. On one hand, clinical scores are inevitably subjective at least to some degree, and may not represent truly significant clinical change. On the other hand, clinical scores are arguably more sensitive to change, and therefore more likely to detect smaller degrees of improvement.
Indirect measures of severity, such as health care utilization, are thought by many to capture significant clinical change better than clinical scores. Some would argue that if a treatment cannot be shown to reduce health care utilization, it is of no benefit.
We have chosen to assess and report both direct and indirect measures of clinical change, with the intent of allowing the reader to make their own judgement as to the relative merits of the different measures.
Why it is important to do this review
Croup is a common childhood illness and may result in presentation to the emergency department (ED) due to difficulty in breathing. Epinephrine has been used as a treatment option to reduce tracheal swelling for over 30 years. However, currently there is no systematic review examining the efficacy of this intervention.
Objectives
Primary objective
To assess the efficacy (measured by croup scores, rate of intubation and health care utilization such as rate of hospitalization) and safety (frequency and severity of side effects) of nebulized epinephrine versus placebo in children with croup, evaluated in an ED or hospital setting.
Secondary objectives
To assess the efficacy and severity of side effects of nebulized racemic epinephrine versus nebulized L‐epinephrine in children with croup evaluated in an ED or hospital setting and to assess the efficacy and severity of side effects of nebulized epinephrine delivered with intermittent positive pressure breathing (IPPB) versus nebulized epinephrine alone.
Methods
Criteria for considering studies for this review
Types of studies
Randomized controlled trials (RCTs) or quasi‐RCTs (Q‐RCTs) which compared the use of nebulized epinephrine to placebo, nebulized racemic epinephrine versus nebulized L‐epinephrine, and nebulized epinephrine delivered with intermittent positive pressure breathing (IPPB) versus nebulized epinephrine alone in children with croup.
Types of participants
Studies included children with a clinical diagnosis of croup (defined as acute onset of a 'barky cough' and stridor) whether evaluated in an ED or admitted to hospital.
Types of interventions
-
Nebulized epinephrine (either racemic or L‐epinephrine, or delivered with or without IPPB) versus placebo.
-
Nebulized racemic epinephrine versus L‐epinephrine.
-
Nebulized epinephrine delivered by IPPB versus nebulized epinephrine delivered without IPPB.
Types of outcome measures
Primary outcomes
-
Change in clinical croup scores following treatment.
Secondary outcomes
-
Rate and duration of intubation.
-
Rate and duration of hospitalization.
-
Rate of return to medical care for ongoing croup symptoms.
-
Improvement.
-
Parental anxiety.
-
Side effects such as hypertension.
-
Evidence of myocardial injury or cardiac arrhythmias.
We evaluated all studies that met the above criteria; we used no exclusion criteria.
Search methods for identification of studies
Electronic searches
We searched the Cochrane Central Register of Controlled Trials (CENTRAL) 2013, Issue 6, part of The Cochrane Library, www.thecochranelibrary.com (accessed 3 July 2013), which contains the Cochrane Acute Respiratory Infections Group's Specialized Register, MEDLINE (1966 to June week 3, 2013), EMBASE (1980 to July 2013), Web of Science (1974 to July 2013), CINAHL (1982 to July 2013) and Scopus (1996 to July 2013).
We used the following search terms to search MEDLINE and CENTRAL. We combined the MEDLINE search with the Cochrane Highly Sensitive Search Strategy for identifying randomized trials in MEDLINE: sensitivity‐ and precision‐maximizing version (2008 revision) Ovid format (Lefebvre 2011). We adapted the search strategy to search EMBASE (see Appendix 2), CINAHL (see Appendix 3), Web of Science (see Appendix 4) and Scopus (see Appendix 5).
MEDLINE (Ovid)
1 exp Laryngitis/
2 (croup or laryngit*).tw.
3 laryngotracheit*.tw.
4 (laryngotracheobronchit* or laryngo‐tracheo‐bronchit*).tw.
5 (pseudocroup or pseudo‐croup).tw.
6 or/1‐5
7 Epinephrine/
8 exp Adrenergic Agonists/
9 exp Adrenal Cortex Hormones/
10 epinephrin*.tw,nm.
11 (adrenalin* or l‐adrenalin*).tw,nm.
12 (adrenergic agonist* or adrenergic alpha‐agonist* or adrenergic beta‐agonist*).tw,nm.
13 or/7‐12
14 exp "Nebulizers and Vaporizers"/
15 Aerosols/
16 respiratory therapy/ or oxygen inhalation therapy/ or respiration, artificial/ or exp positive‐pressure respiration/
17 (inhal* or vapor* or vapour* or atomiz* or atomis* or nebuliz* or nebulis* or spray* or mist* or aerosol*).tw.
18 (positive‐pressure adj3 (breathing or respiration)).tw.
19 ippb.tw.
20 or/14‐19
21 6 and 13 and 20
Searching other resources
We contacted experts in the field of acute respiratory infections to locate additional studies. There were no language or publication restrictions.
Data collection and analysis
Selection of studies
Two review authors (CB, KR) independently scanned abstracts from the initial search results to identify trials that broadly met the inclusion criteria. We then reviewed the full‐text articles of the selected trials and the same two review authors independently applied the inclusion criteria. We resolved differences by consensus.
Data extraction and management
We extracted data using a structured form that captures patient status (ED or inpatient, the author's description of their clinical severity), intervention and control characteristics (such as type of drug: racemic versus L‐epinephrine), dosage and method of administration (nebulization alone versus nebulized with intermittent positive pressure breathing (IPPB)). In addition, we collected data on the primary and secondary outcome measures if available: change from baseline clinical croup scores; rate and duration of intubation; rate and duration of hospitalization; and occurrence of hypertension, myocardial injury or cardiac arrhythmias. One review author (KR) extracted data and a second review author (CB) checked this for accuracy. If necessary, we extracted data from graphs or directly from the trial authors.
Assessment of risk of bias in included studies
Two review authors (CB, KR) independently assessed the quality of studies in three ways. We used more than one method since the relative merits of each remain controversial. First, we used the method outlined in the Cochrane Handbook for Systematic Reviews of Interventions that focuses on selection, performance, attrition and detection bias (Higgins 2011). Second, we classified concealment of allocation as being a high, low or unclear risk of bias (previously referred to as adequate, inadequate or unclear (Schulz 1995)). Lastly, we classified studies by who sponsored them: pharmaceutical company, other sources or not mentioned (Cho 1996). We resolved differences by consensus. We compared and reported the results from the three different classification methods.
Measures of treatment effect
With regard to continuous variables, we calculated the change from baseline measures if change from baseline measures were not reported directly. As needed, we performed variance imputations according to the work of Follmann (Follman 1992). We anticipated that different clinical croup scores would be reported. If this was the case, we used trial standardized mean differences (SMDs) for pooled estimates. (A treatment effect (difference between treatment means) divided by its measurement variation (for example, a pooled standard deviation) gives a SMD). We also anticipated that croup scores would be assessed at a variety of time periods. Because the treatment effect of epinephrine is rapid, we assessed change in croup score at 30 minutes, two hours and six hours. If a study did not assess change in croup scores at our time points of interest, we used the closest time point (for example, a 15‐minute change in croup score as used for the change in croup score at 30 minutes outcome). We expressed duration of intubation and hospitalization as MDs and calculated an overall MD.
For binary data (that is, intubation and hospitalization rates), we calculated risk ratios (RR). If zero events occurred in both groups, we derived risk differences (RD). If we found significant results, we calculated the number needed to treat to benefit (NNTB) or harm (NNTH). We reported the pooled baseline rates using the inverse variance method.
Unit of analysis issues
We calculated the change from baseline measures in cases where change from baseline measures were not reported directly (Corkey 1981; Fogel 1982; Waisman 1992; Westley 1978).
Dealing with missing data
Many variance imputations were performed according to the work of Follman 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 (Corkey 1981; Fogel 1982; Westley 1978; Waisman 1992). We imputed variances from upper bound P values (Kristjansson 1994).
Assessment of heterogeneity
We assessed heterogeneity quantitatively with the I2 statistic (Higgins 2002). The I2 statistic indicates the per cent variability due to between‐study (or inter‐study) variability as opposed to within‐study (or intra‐study) variability. We considered an I2 statistic greater than 50% to be large.
Assessment of reporting biases
If at least eight studies were found for our primary outcome we tested for publication bias. In addition to funnel plots, we used the rank correlation test (Begg 1994) and weighted regression (Egger 1997) for the detection of publication bias. We performed adjustment for publication bias in the pooled estimates using the trim and fill method. We used more than one method since the relative merits of the methods are not well established.
Data synthesis
For the analyses of all outcomes, we used a random‐effects model to combine treatment effects regardless of quantified heterogeneity. We considered a fixed‐effect model in sensitivity analyses.
Subgroup analysis and investigation of heterogeneity
We explored heterogeneity between studies using subgroup and sensitivity analyses performed on the primary outcome of the change in clinical croup scores from baseline to 30 minutes and 60 minutes. We considered the following subgroups: quality (that is, the risk of bias, allocation concealment, funding and simple classification of bias (low, moderate or high)) and inpatient versus ED status.
Sensitivity analysis
There were not enough studies in any one meta‐analysis to perform any meaningful sensitivity analyses.
Results
Description of studies
Results of the search
For the 2011 Cochrane Review, the electronic literature search identified 316 unique studies. After assessing the titles and, when available, the abstracts, we identified 50 studies as being potentially relevant. Reviewing the reference lists of the included studies did not identify any additional studies.
The electronic literature search performed in July 2013 identified 23 studies, one of which was a duplicate, leaving 22 unique studies. After assessing the titles and the abstracts for each of these 22 studies, we identified no new relevant studies.
Included studies
After applying the a priori inclusion criteria, eight of the 50 studies met the inclusion criteria. Three comparisons were examined: nebulized epinephrine versus placebo (Corkey 1981; Fernandez 1993; Gardner 1973; Kristjansson 1994; Kuusela 1988; Westley 1978), nebulized racemic epinephrine versus L‐epinephrine (Waisman 1992) and nebulized epinephrine with IPPB versus nebulized epinephrine without IPPB (Fogel 1982). All studies were randomized, placebo‐controlled trials with the exception of the study comparing nebulized epinephrine with and without IPPB, which was a cross‐over RCT. Seven studies were published in English and one in Spanish (Fernandez 1993). Studies tended to be small, ranging from 14 to 78 total participants per comparison.
Nebulized epinephrine versus placebo
Four studies took place in an inpatient setting, one in an outpatient setting and one did not specify setting. In general, children within the studies were young (average age two years or less in two studies, two to three years in two studies and not specified in two studies). Severity of croup in the majority of enrolled children was judged to be moderate or moderate to severe in all six studies. There was minor variance between studies in the type of epinephrine used (racemic epinephrine in five studies, L‐epinephrine in one study) and dose of racemic epinephrine used (0.5 ml of 2.25% racemic epinephrine in three studies, 0.5 mg/kg of 2.25% racemic epinephrine in two studies, and 0.25 ml of 2.25% racemic epinephrine per five kilograms of body weight in one study). Epinephrine was administered via IPPB in two of the six studies. One study (Westley 1978) used a croup score which has been validated. The Westley 17‐point croup score incorporates assessment of level of consciousness (five points), cyanosis (five points), stridor (two points), air entry (two points) and chest wall retractions (three points). The remaining five studies used a variety of non‐validated croup scores as outcome measures, with total possible points ranging from six to 12 points.
Nebulized racemic epinephrine versus L‐epinephrine
We identified a single study suitable for inclusion, of 66 children in an inpatient setting (Waisman 1992). Average age was young (11 months) and baseline severity of croup on study admission was moderate. The doses of racemic epinephrine and L‐epinephrine were 0.5 ml of 2.25% and 5 ml of 1:1000 dilution, respectively. A non‐validated 10‐point croup score (inspiratory breath sounds, two points; stridor, two points; cough, two points, retractions/nasal flaring, two points; cyanosis, two points) was utilized as the primary outcome measure.
Nebulized epinephrine with IPPB versus nebulized epinephrine without IPPB
We identified a single study suitable for inclusion, of 14 children in an inpatient setting (Fogel 1982). Average age was young (1.9 years) and baseline severity of croup on study admission was moderate. The dose of racemic epinephrine used was 0.25 ml of 2.25%. The IPPB pressure used was 15 cm to 17 cm of water and two hours after the first treatment, cross‐over to the other group occurred. A modified 16‐point croup score based upon the validated Westley croup score (level of consciousness, four points; color, four points; stridor, three points; air entry, two points; and chest wall retractions, three points) was utilized as the primary outcome measure.
Excluded studies
We excluded a total of 50 studies. The exact reason for exclusion is provided in the Characteristics of excluded studies table.
Risk of bias in included studies
In the original review, we assessed methodological quality using the Jadad scale and concealment of allocation. No studies were added to this update. However, the conclusions concerning methodological quality were revised to reflect the 'Risk of bias' tool. We deemed six of the eight studies to have a low risk of bias and the risk of bias was unclear in the remaining two studies (Fernandez 1993; Kuusela 1988). Half of the studies were reported to be of low risk of bias due to adequate concealment of allocation (Fernandez 1993; Kuusela 1988; Waisman 1992; Westley 1978). Only one study reported a funding source (Kristjansson 1994). The results from the various quality indicators are pictorially displayed in Figure 1 and Figure 2.
'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study.
Allocation
Allocation concealment was adequate in three of the studies (Fernandez 1993; Kuusela 1988; Waisman 1992), inadequate in two (Corkey 1981; Westley 1978) and unclear in three (Fogel 1982; Gardner 1973; Kristjansson 1994).
Blinding
All trials were double‐blind. Two studies did not report who was blinded (Gardner 1973; Westley 1978).
Incomplete outcome data
Four studies reported losses to follow‐up (Kuusela 1988; Kristjansson 1994; Waisman 1992; Westley 1978). However, only one study explained the losses to follow‐up (Waisman 1992).
Selective reporting
All but two studies (Fogel 1982; Waisman 1992) reported a primary outcome (6/8; 75%).
Other potential sources of bias
Two studies conducted an intention‐to‐treat (ITT) analysis (Corkey 1981; Fogel 1982).
Effects of interventions
Nebulized epinephrine versus placebo
There were six studies (183 participants) that reported data comparing nebulized epinephrine versus placebo. Four (109 participants) of these six studies took place in an inpatient setting, while one (54 participants) took place in an outpatient setting and one (20 participants) did not specify the setting. While six studies were included in this section, no more than three were included in any particular analysis.
Primary outcome
1. Croup score
Since all studies used different systems in their croup scoring, we used standardized mean differences (SMDs) to synthesize the data. Three studies showed that epinephrine had a statistically significant smaller croup score than placebo after 30 minutes (SMD ‐0.94; 95% confidence interval (CI) ‐1.37 to ‐0.51; I2 statistic = 0%) (Analysis 1.1). The change in croup score was almost identical for the two studies that looked at inpatients (SMD ‐0.82; 95% CI ‐1.47 to ‐0.17; I2 statistic = 0%) (Analysis 1.1) and the one study that looked at outpatients (SMD ‐1.03; 95% CI ‐1.60 to ‐0.46) (Analysis 1.1). Heterogeneity was negligible in these comparisons.
There were also data on croup score after two hours (SMD ‐0.15; 95% CI ‐1.03 to 0.73; one study) (Analysis 1.2) and six hours (SMD ‐0.60; 95% CI ‐1.50 to 0.30; two studies; I2 statistic = 70%) (Analysis 1.3). Neither of these estimates was statistically significant.
Secondary outcomes
1. Rate and duration of intubation
Not reported.
2. Rate and duration of hospitalization
There were two studies that reported length of hospital stay (in hours), one each in the inpatient and outpatient settings. The inpatient study showed children on epinephrine spent a statistically significant smaller amount of time in the hospital than placebo participants in the hospital (MD ‐32.00; 95% CI ‐59.14 to ‐4.86) (Analysis 1.4), while the outpatient study had a much smaller and non‐significant difference (MD ‐1.80; 95% CI ‐4.07 to 0.47) (Analysis 1.4). We did not combine these two studies as we would not expect similar hospital stays between inpatients and outpatients.
3. Rate of return to medical care for ongoing croup symptoms
Return visits were reported by one study. However, that study reported no re‐admissions in either the epinephrine or the placebo group, thus we computed no statistics.
4. Improvement
Two studies counted the number of participants that had significant improvement. Improvement was defined in one study as a decrease in the croup score of greater that two or more points (2/10 total points) (Gardner 1973) and in the other study as a decrease in the croup score of greater than two or more points (2/15) (Kristjansson 1994). The combined risk ratio (RR) did favor epinephrine but it was not statistically significant (RR 1.46; 95% CI 0.82 to 2.60; I2 statistic = 17%) (Analysis 1.6). Heterogeneity was moderate.
5. Parental anxiety
Not reported.
6. Side effects such as hypertension
Not reported.
7. Evidence of myocardial injury or cardiac arrhythmias
Not reported.
Nebulized racemic epinephrine versus L‐epinephrine
Primary outcome
1. Croup score
Waisman 1992 (28 participants) compared nebulized racemic epinephrine versus L‐epinephrine in an outpatient setting. There was no significant difference in croup score between the two types of epinephrine after 30 minutes (SMD 0.33; 95% CI ‐0.42 to 1.08) (Analysis 2.1) but after two hours, L‐epinephrine showed significant reduction over racemic epinephrine (SMD 0.87; 95% CI 0.09 to 1.65) (Analysis 2.2).
Secondary outcomes
Rate and duration of intubation. The racemic epinephrine group had an intubation rate of 3/16 and the L‐epinephrine group had a rate of 0/14, which was a non‐significant difference (RD 0.19; 95% CI ‐0.03 to 0.40) (Analysis 2.3).
1. Rate and duration of intubation
Not reported.
2. Rate of return to medical care for ongoing croup symptoms
Not reported.
3. Improvement
Not reported.
4. Parental anxiety
Not reported.
5. Side effects such as hypertension
Not reported.
6. Evidence of myocardial injury or cardiac arrhythmias
Not reported.
Nebulized epinephrine with intermittent positive pressure breathing (IPPB) versus nebulized epinephrine without IPPB
Primary outcome
1. Croup score
Fogel 1982 (14 participants) compared nebulized epinephrine with IPPB versus nebulized epinephrine without IPPB in an inpatient setting. After 30 minutes, there was no significant difference in croup score between the two groups (SMD ‐0.72; 95% CI ‐1.86 to 0.42) (Analysis 3.1), while at two hours there was a larger, still non‐significant difference (SMD ‐0.14; 95% CI ‐1.24 to 0.95) (Analysis 3.2).
Secondary outcomes
1. Rate and duration of intubation
Intubation was reported as an outcome. However, both groups had zero intubations, thus we computed no statistics.
2. Rate of return to medical care for ongoing croup symptoms
Not reported.
3. Improvement
Not reported.
4. Parental anxiety
Not reported.
5. Side effects such as hypertension
Not reported.
6. Evidence of myocardial injury or cardiac arrhythmias
Not reported.
Subgroup, sensitivity and publication bias analysis
Since none of our analyses contained more than three studies, no further analysis exploring heterogeneity or publication bias could be conducted.
Discussion
Summary of main results
Nebulized epinephrine has become standard management, especially in North America, for the treatment of moderate and severe croup. Nebulized epinephrine is typically administered to a child with moderate or severe upper airway obstruction in either an emergency department (ED) or inpatient setting. Though widely used in clinical practice, our search identified only six relevant clinical trials comparing nebulized epinephrine to placebo that included a total of 183 participants (Corkey 1981; Fernandez 1993; Gardner 1973; Kristjansson 1994; Kuusela 1988; Westley 1978).
Nebulized epinephrine versus placebo
Primary outcome: croup score
Data from this review have shown that, compared to placebo, nebulized racemic epinephrine effectively improves croup symptoms as measured by clinical croup scores in children 30 minutes following treatment. Three trials assessed croup score at 30 minutes (Corkey 1981; Kristjansson 1994; Westley 1978). Although the number of children was small (94 participants), the pooled data indicated a moderate to large reduction in croup score as compared with placebo (Cohen 1988). All three trials favored racemic epinephrine over placebo and the magnitude of benefit was similar and consistent between studies and in both inpatient and outpatient settings.
At 120 minutes following treatment, the mean difference in croup score was similar in both racemic epinephrine and placebo groups in the single, small (20 participants) study (Westley 1978). Although the observed clinical benefit of epinephrine had essentially dissipated at 120 minutes, there was no evidence, on average, to suggest that there was an increase or worsening of croup score, as compared to pre‐treatment or baseline in the group of children treated with epinephrine.
Two studies (69 participants) assessed croup score at six hours following treatment in an inpatient setting (Fernandez 1993; Kuusela 1988) and found no difference in croup score between the nebulized racemic epinephrine and placebo groups. This is not surprising given the relatively brief duration of action of nebulized epinephrine and is also consistent with the finding of no difference in croup score at the 120‐minute assessment.
Secondary outcomes
In a single inpatient study, hospital stay was shorter among those treated with nebulized racemic epinephrine group as compared with placebo (Kuusela 1988). The mean difference of 32 hours was both statistically and clinically significant. Confidence intervals were wide, however, reflecting the small study size. Also, corticosteroid was administered to eight of 37 participants and the breakdown by treatment group was not provided. This could account for the shorter hospital stay in the epinephrine group. Length of stay in the outpatient setting was assessed in one study and it was similar between epinephrine and placebo groups (Kristjansson 1994). Equal proportions of children (52% and 58% in the racemic epinephrine and placebo groups, respectively) received concomitant corticosteroid treatment. No difference in length of stay is consistent with resolution of clinical effect by two hours after treatment.
We chose to report the outcome proportion of patients improved even though we had not pre‐specified it as a secondary outcome because one study (Gardner) did not report the actual croup score values, only the proportion of children whose croup score had improved by two or more points versus those which had not.
Racemic epinephrine versus L‐epinephrine
A small, methodologically sound study comparing nebulized racemic epinephrine with L‐epinephrine in the outpatient setting showed no difference in croup score 30 minutes following treatment (Waisman 1992). By 120 minutes, there was a statistically significant result showing L‐epinephrine to have a longer duration of benefit than racemic epinephrine.
Racemic epinephrine is composed of equal proportions of D‐ and L‐isomers of epinephrine and was originally used to treat croup because it was hypothesized that racemic epinephrine would cause fewer cardiovascular side effects than L‐epinephrine. L‐epinephrine is the type of epinephrine routinely used for other indications in medicine in either 1:1000 or 1:10,000 concentrations. Waisman 1992 administered an identical 5 mg of L‐epinephrine (0.5 ml of 2.25% racemic epinephrine versus 5.0 ml of 1:1000 epinephrine) and found no statistically significant differences in cardiovascular side effects between the two drugs. In addition, it is now known that only L‐epinephrine is pharmacologically active; the R‐isomer has no activity (Westfall 2009).
Nebulized epinephrine with IPPB versus nebulized epinephrine without IPPB
One study of good methodological quality found that nebulized epinephrine with IPPB was similar to nebulized epinephrine without IPPB in the inpatient setting (Fogel 1982). However, this study was small and not designed to show equivalence. Nonetheless, given the technical challenges of IPPB and no evidence of superiority, routine use of nebulized epinephrine without IPPB seems justified.
Safety of nebulized epinephrine
Though none of the clinical trials reported significant adverse events associated with nebulized epinephrine, the small total number of participants and rarity of adverse events provides insufficient data to conclude that the use of nebulized epinephrine is universally safe. Further, none of these trials assessed the effectiveness or safety of epinephrine when administered repeatedly in a row. A single case report of ventricular tachycardia and myocardial infarction in a previously healthy child who received three doses of nebulized epinephrine in one hour raises concerns about potential cardiac toxicity (Butte 1999). While this case is of concern, given the widespread use of epinephrine over several decades, it seems unlikely, however, that one or even two nebulized doses of epinephrine poses significant risk to a child.
Overall completeness and applicability of evidence
The number of relevant studies was small, as were total numbers of participants. Studies assessed a wide range of outcomes and few studies examined the same outcomes, thus most outcomes contained data from very few or even single studies. Timing of outcome measures also varied between studies. Finally, co‐interventions (for example, corticosteroid administration) were not explicitly described for some studies. There were too few studies included to formally assess publication bias.
We did not search for articles which included co‐treatment with corticosteroids, as the question of whether adjunct epinephrine therapy provides additional benefit to corticosteroid treatment alone will be addressed in a separate Cochrane Review.
Finally, none of the studies included children with mild croup. However, mild croup is defined as minimal to no symptoms of airway obstruction on presentation; epinephrine is used to provide temporary relief of airway obstructive symptoms and, thus, there is not a strong rationale for the use of epinephrine in children with mild croup.
Quality of the evidence
The majority of the studies were at low risk of bias and half of the studies reported adequately concealed patient allocation methods. While all studies were double‐blind and the majority of the studies reported their primary outcome, several studies reported a loss to follow‐up and few studies conducted an intention‐to‐treat analysis.
Potential biases in the review process
We undertook a broad and extensive search strategy of multiple databases, contacted experts in the field and applied no language or publication restrictions to minimize the chance of bias due to missing published studies or non‐English language studies. This review did contain one Spanish language study. As none of our analyses contained more than three studies, no further analysis exploring heterogeneity or publication bias could be performed. Thus, the possibility of publication bias cannot be ruled out.
Agreements and disagreements with other studies or reviews
We did not identify any other relevant studies or systematic reviews in publication.
'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study.
Comparison 1 Nebulized epinephrine versus placebo, Outcome 1 Croup score (change baseline ‐ 30 minutes).
Comparison 1 Nebulized epinephrine versus placebo, Outcome 2 Croup score (change baseline ‐ 2 hours).
Comparison 1 Nebulized epinephrine versus placebo, Outcome 3 Croup score (change baseline ‐ 6 hours).
Comparison 1 Nebulized epinephrine versus placebo, Outcome 4 Length of hospitalization in hours.
Comparison 1 Nebulized epinephrine versus placebo, Outcome 5 Return visits and/or (re)admissions.
Comparison 1 Nebulized epinephrine versus placebo, Outcome 6 Improvement.
Comparison 2 Nebulized racemic epinephrine versus L‐epinephrine, Outcome 1 Croup score (change baseline ‐ 30 minutes).
Comparison 2 Nebulized racemic epinephrine versus L‐epinephrine, Outcome 2 Croup score (change baseline ‐ 2 hours).
Comparison 2 Nebulized racemic epinephrine versus L‐epinephrine, Outcome 3 Intubation.
Comparison 3 Nebulized epinephrine with IPPB versus nebulized epinephrine without IPPB, Outcome 1 Croup score (change baseline ‐ 30 minutes).
Comparison 3 Nebulized epinephrine with IPPB versus nebulized epinephrine without IPPB, Outcome 2 Croup score (change baseline ‐ 2 hours).
Comparison 3 Nebulized epinephrine with IPPB versus nebulized epinephrine without IPPB, Outcome 3 Intubation.
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Croup score (change baseline ‐ 30 minutes) Show forest plot | 3 | 94 | Std. Mean Difference (IV, Random, 95% CI) | ‐0.94 [‐1.37, ‐0.51] |
| 1.1 Inpatient | 2 | 40 | Std. Mean Difference (IV, Random, 95% CI) | ‐0.82 [‐1.47, ‐0.17] |
| 1.2 Outpatient | 1 | 54 | Std. Mean Difference (IV, Random, 95% CI) | ‐1.03 [‐1.60, ‐0.46] |
| 2 Croup score (change baseline ‐ 2 hours) Show forest plot | 1 | Std. Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 2.1 Inpatient | 1 | Std. Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 2.2 Outpatient | 0 | Std. Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 3 Croup score (change baseline ‐ 6 hours) Show forest plot | 2 | 138 | Std. Mean Difference (IV, Random, 95% CI) | ‐0.60 [‐1.12, ‐0.08] |
| 3.1 Inpatient | 2 | 69 | Std. Mean Difference (IV, Random, 95% CI) | ‐0.60 [‐1.50, 0.30] |
| 3.2 Outpatient | 0 | 0 | Std. Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] |
| 3.3 IPPB administration | 1 | 37 | Std. Mean Difference (IV, Random, 95% CI) | ‐1.06 [‐1.76, ‐0.36] |
| 3.4 No IPPB administration | 1 | 32 | Std. Mean Difference (IV, Random, 95% CI) | ‐0.14 [‐0.84, 0.55] |
| 4 Length of hospitalization in hours Show forest plot | 2 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 4.1 Inpatient | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 4.2 Outpatient | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 5 Return visits and/or (re)admissions Show forest plot | 1 | Risk Difference (M‐H, Random, 95% CI) | Totals not selected | |
| 5.1 Inpatient | 0 | Risk Difference (M‐H, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 5.2 Outpatient | 1 | Risk Difference (M‐H, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 6 Improvement Show forest plot | 2 | 74 | Risk Ratio (M‐H, Random, 95% CI) | 1.46 [0.82, 2.60] |
| 6.1 Inpatient | 0 | 0 | Risk Ratio (M‐H, Random, 95% CI) | 0.0 [0.0, 0.0] |
| 6.2 Outpatient | 1 | 54 | Risk Ratio (M‐H, Random, 95% CI) | 1.83 [0.96, 3.50] |
| 6.3 Setting not reported | 1 | 20 | Risk Ratio (M‐H, Random, 95% CI) | 1.0 [0.42, 2.40] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Croup score (change baseline ‐ 30 minutes) Show forest plot | 1 | Std. Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.1 Inpatient | 0 | Std. Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 1.2 Outpatient | 1 | Std. Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 2 Croup score (change baseline ‐ 2 hours) Show forest plot | 1 | Std. Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 2.1 Inpatient | 0 | Std. Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 2.2 Outpatient | 1 | Std. Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 3 Intubation Show forest plot | 1 | Risk Difference (M‐H, Random, 95% CI) | Totals not selected | |
| 3.1 Inpatient | 0 | Risk Difference (M‐H, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 3.2 Outpatient | 1 | Risk Difference (M‐H, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Croup score (change baseline ‐ 30 minutes) Show forest plot | 1 | Std. Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.1 Inpatient | 1 | Std. Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 1.2 Outpatient | 0 | Std. Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 2 Croup score (change baseline ‐ 2 hours) Show forest plot | 1 | Std. Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 2.1 Inpatient | 1 | Std. Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 2.2 Outpatient | 0 | Std. Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 3 Intubation Show forest plot | 1 | Risk Difference (M‐H, Random, 95% CI) | Totals not selected | |
| 3.1 Inpatient | 1 | Risk Difference (M‐H, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 3.2 Outpatient | 0 | Risk Difference (M‐H, Random, 95% CI) | 0.0 [0.0, 0.0] | |
