A critical analysis and appraisal of the management of croup in the UK out-of-hospital environment

Background and epidemiology

Croup is a common respiratory infection with an incidence of around 3% of the paediatric population per year, most commonly affecting children between 6–36 months and more commonly males than females (Johnson, 2009; Brashers and Huether, 2014). The prevalence within this age group is commonly thought to be due to socialisation of the child in nurseries and schools, with additional predisposing factors such as pollution, passive smoking and immunological immaturity being noted (Mansi et al, 2009). In addition, a general consensus in the literature identifies a seasonal variance in the prevalence of croup, with a peak in patient presentations to health services in late spring/autumn to winter months, with the opposite being true for summer months (Segal et al, 2005; D'Souza et al, 2007; Bjornson and Johnson, 2008; Rihkanen et al, 2008). This is also supported by Public Health England (2014), who show that activity of one of the dominant causative agents for croup, the human parainfluenza virus (HPIV) is greatest in autumn and winter months.

The predominance of this illness is highlighted in the literature such as an epidemiological study by Rosychuk et al (2010), who report 20 019 emergency department (ED) attendances by young children under 24 months of age for 27 355 episodes of croup across a 6-year period. Consequently, the noted commonplace of croup provides many socioeconomic repercussions, whereby visits to general practitioners and EDs incurred costs of treatment, lost school days for children and lost work days for their parents (Mansi et al, 2009). In addition, while there is a paucity in epidemiological literature specific to croup in the paramedic pre-hospital environment, feverish respiratory illnesses in children such as croup are prevalent within the community and are a common reason for accessing out-of-hours health care (National Institute for Health and Care Excellence, 2013). It may therefore be argued that robust, evidence-based practice surrounding the treatment and management of croup should be encouraged within the pre-hospital environment in order to address these findings and promote patient-centred care.

Aetiology and pathophysiology

It is well noted in literature that croup is primarily a viral infection, with the HPIV acting as the dominant infective agent, with up to 75% of incidences reported HPIV positive (Johnson, 2009; Mansi et al, 2009). This is supported in a prospective study by Rihkanen et al (2008) who explored the virus types contained within the nasopharyngeal mucus of 144 children diagnosed with croup. Here, the top five percentages for virus types included HPIV 1 (30.6%), respiratory syncytial virus (14.6%), human bocavirus (12.5%), rhinovirus (11.8%), enterovirus (9.0%) and influenza A (9.0%) (Rihkanen et al, 2008). Additionally, a total of 41% of patients were found positive for HPIV-1–HPIV-3 (Rihkanen et al, 2008).

Invasion of the host with such pathogens usually occurs within the epithelium of the nasopharynx, commonly causing coryzal upper respiratory tract symptoms such as rhinorrhoea, hoarseness and low-grade fever for 12–48 hours (Bjornson and Johnson, 2008; Balachandran, 2013). Progression of the infection leads to inflammatory oedema and spasm of the pharyngeal, laryngeal, tracheal and bronchial mucosa with necrosis and shedding of the epithelial cells (Fitzgerald, 2006; Brashers and Huether, 2014).

At the larynx, oedema leads to a decreased mobility of the vocal cords, manifesting as a characteristic hoarseness (Tebruegge and Curtis, 2012). Laryngeal oedema is, however, limited due to the tight adherence of its mucosal membrane to underlying cartilage (Brashers and Huether, 2014). The more distal subglottic mucosal membrane has a looser adherence, allowing for greater mucosal oedema (Brashers and Huether, 2014). This region is well demarcated as the narrowest portion of the child's upper-airway, where oedematous accumulation substantially reduces the cross-sectional area (Eid, 2009).

Furthermore, infection with HPIV causes inhibition of sodium absorption and activates chloride secretion across the epithelial membranes of the trachea, further promoting oedematous accumulation (Kunzelmann et al, 2004). Obstruction of the upper-airway, and its associated increase in resistance to airflow, produces inspiratory and/or expiratory stridor and generates an increased negative intra-thoracic pressure (Brashers and Huether, 2014). This manifests as an increased work of breathing, nasal flaring and chest wall recession with a non-productive ‘barking’ cough (Balachandran, 2013).

In severe/life-threatening cases, the patient may become fatigued with ensuing asynchronous abdominal and chest wall motion, leading to hypoxia, acidosis and acute respiratory failure (Pfleger and Eber, 2013). This form of severe croup is, however, estimated to occur in only 1% of children (Pfleger and Eber, 2013). Having established the epidemiology, aetiology and pathophysiology surrounding croup, this article will now explore evidence surrounding current pharmacotherapies for croup.

Corticosteroids in croup

Evidence to support the use of corticosteroids as the mainstay of treatment for croup has recently gained a consensus within literature. Here, meta-analyses of several randomised controlled trials (RCT) exploring the use of a corticosteroid versus a placebo, provide significant findings of patient improvement with corticosteroids and reductions in the requirements of secondary treatment (Kairys et al, 1989; Griffin et al, 2000). More recently, a Cochrane Review by Russel et al (2011) of 38 studies reported an improvement in the Westley score (see Table 1) of participants after both 6 and 12 hours following treatment with corticosteroids. Furthermore, a decrease in re-admissions or return visits was seen in these patients, while amount of time spent in medical facilities also significantly decreased. From these results the authors conclude that both dexamethasone and budesonide are effective in the treatment of croup, reducing length of hospital stays and secondary access to health care (Russel et al, 2011). This recent review arguably forms a definitive support for the use of corticosteroids in croup and further promotes the already historic uptake of this pharmacotherapy into hospital practice guidelines (Bjornson and Johnson, 2008). Despite this, dissemination of the aforementioned findings and uptake of the evidence into paramedic practice has been relatively slow, with guidelines for paramedic use of corticosteroids for croup having only recently been introduced (Association of Ambulance Chief Executives (AACE), 2013).

Current dosages for pre-hospital paramedics

Within UK paramedic practice, oral dexamethasone is indicated for moderate to severe croup with two set dosages dependant on age: 2 mg for ages 1–12 months and 4 mg for 18 months–6 years (AACE, 2013). These doses are singular, with no repeat dose required (AACE, 2013). This differs from alternate guidance for practitioners, such as in the British National Formulary (Joint Formulary Committee, 2014), which advocates 150 µg per kg, therefore allowing for a more accurately calculated dose. While risk of adverse effects of corticosteroid therapy in children are relatively low, it may be contended that accurate doses of systemic steroids should be sought, especially in children who suffer with recurrent croup requiring more frequent pharmacotherapy (Bjornson and Johnson, 2008). This is supported by Brown (2009), who argues that nebulised budesonide should be used for recurrent spasmodic croup, in order to negate the need for multiple systemic courses of corticosteroid.

It may therefore be argued that lower and correctly calculated doses of dexamethasone administered by paramedics may be preferable over two set doses for a wide age range. This is disputed by Eastwood et al (2009), who conducted a literature review surrounding paramedics' ability to perform drug calculations. Through the analysed literature it was found that a significant lack of mathematical proficiency exists within the paramedic profession, through which drug dosages may frequently be miscalculated. An argument may therefore be made for continuing a fix dose regimen for drugs administered by paramedics in order to avoid inadvertent under or overdosing of children. This argument is supported in a later cross-sectional study undertaken by Eastwood et al (2013), who used a paper-based questionnaire on general maths and drug calculations to test the accuracy of undergraduate paramedic students in performing drug calculations. Results showed a mean score of 39.5% for its participants, highlighting a significant inability of the students involved in performing accurate calculations. This further suggests that the current fixed dosage guidance for UK paramedics is safer and reduces the chance of miscalculation.

It is, however, noted that this study was conducted on a small sample of undergraduate paramedics from outside the UK and thus its generalisability to UK paramedics is limited (Hulley et al, 2013). Additionally, it may be contended that the use of paper-based questionnaires cannot accurately reflect real-life practice, thus limiting the external validity of the study (Babbie, 2009). Furthermore the study only achieved a 52.3% response rate, arguably reducing the significance of its results due to non-response bias (Babbie, 2009). In summary, although a paucity in robust and high-level evidence regarding UK paramedics is noted, currently evidence surrounding paramedic drug calculation ability seems to substantiate the benefit of a fixed dosage guideline for UK paramedics.

The superiority of higher doses of corticosteroid against lower doses has also been explored in past literature. Here, the meta-analysis by Kairys et al (1989) reports that higher doses of corticosteroids equates to increased numbers of children who improved 12 hours after treatment, compared to those treated with lower doses or the placebo. While this forms an argument towards higher doses of corticosteroids for the pharmacotherapy of croup, it is noted that due to the differing designs in all of the studies reviewed, the prospect of bias ensues (Bjornson and Johnson, 2008).

Additional RCTs have been published comparing alternate doses of dexamethasone (Geelhoed and Macdonald, 1995b; Alshehri et al, 2005; Fifoot and Ting, 2007). Findings from these trials form a consensus that a 0.15 mg/kg dose is sufficient, which contrasts to findings from Kairys et al (1989) and supports the administration of lower doses of corticosteroids. It may, however, be argued that the clinical impact and significance brought about from these RCTs is restricted due to small sample sizes in each study (Hulley et al, 2013). Here it is argued that a larger sample size may have increased the statistical power and external validity of the results (Hulley et al, 2013). Moreover, the absence of further robust RCTs in this area highlights a need for further research.

From analysis of the above literature it may further be maintained that although lower doses have been shown to be adequate in the treatment of croup, pragmatic use of larger doses may be of benefit when treating more severe or refractory cases, as highlighted by Kairys et al (1989) and supported by Bjornson and Johnson (2008). Further research in this area is therefore required, particularly relating to dose-specific, pre-hospital administration of corticosteroids. From this discussion, exploration surrounding the routes of administration and applicability in pre- hospital paramedic practice will now be made.

Routes of administration

In UK paramedic guidelines (AACE, 2013), dexamethasone administration is currently advocated via the oral route, using an intravenous preparation of 4mg/ml. While oral administration of corticosteroids is historical within healthcare practice, other routes of administration such as the intramuscular (IM) and nebulised routes (See Figure 1) are available (Bjornson and Johnson, 2008). Furthermore, as with many of the drugs currently within the UK paramedic's armamentarium, several routes of administration are available for administration using clinical judgement. For dexamethasone, UK paramedics are only able to administer via the oral route, calling into question the efficacy and applicability of other routes such as the IM or nebulised route.

Nebulisation and IM injection of corticosteroids have certain practical advantages over the oral route. Firstly, the nebulisation/IM route may arguably be of benefit in patients suffering with vomiting, or for whom oral administration may be difficult due to respiratory distress or an inability/refusal to swallow (Bjornson and Johnson, 2008). Similarly, the palatability of oral drugs is an equally important concern, where poor taste may reduce tolerance (Çetinkaya et al, 2004). Both issues are commonly seen in children with croup and may cause inadequate or unknown doses to be administered as a result, or induce vomiting with the same issues ensuing (Çetinkaya et al, 2004). In light of this, comparative RCTs regarding the efficacy of alternative methods of oral administration have been found; for example, Qureshi et al (2001) show dexamethasone tablets crushed into softs foods to be better tolerated by children than prednisolone syrup.

Alternatively, though the IM route is arguably a quick and efficient way of administering corticosteroids, it is at a disadvantage in this case due to providing risk of infection, while also causing pain and anxiety in an already distressed child (Çetinkaya et al, 2004). Conversely, nebuliser therapy is relatively non-invasive and also allows for simultaneous oxygen administration, which may be of benefit to hypoxic patients suffering with more severe croup (Advanced Life Support Group, 2011). In such cases, decreased gastrointestinal tract perfusion may lead to poor absorption of oral corticosteroids, while poor local tissue perfusion may impair absorption of drugs via the IM route (Richards, 2009; Singh et al, 2011). An argument for the routine use of nebulised corticosteroids in such circumstances by paramedics may therefore be presented, while also considering the advantages of the IM route (Singh et al, 2011).

While considering the practical advantages of alternate routes, evidence into their efficacy in treating croup must also be analysed. In their recent Cochrane Review, Russel et al (2012) reviewed three studies where a direct comparison was made between nebulised/inhaled budesonide and oral or IM dexamethasone (Klassen et al, 1998; Pedersen et al, 1998; Duman et al, 2005). Combining these results showed no statistically significant difference in croup scores at 6 hours post treatment, indicating no clinically significant difference between the routes. When viewed in isolation, however, findings from Pedersen et al (1998) produced a significant difference (p=0.001, p=0.0004) at 12 hours post-treatment in the participant group administered with IM dexamethasone, thus outlining this route as potentially more efficacious. This significantly different finding is, however, only reported from one of the three reviewed studies. This fact, alongside the small number of participants recruited for each study and the presence of confounding variables, meant that conclusive recommendations regarding the efficacy of either route over the other were unable to be made by Russel et al (2012).

In support of this, results from an RCT by Geelhoed and Macdonald (1995a) report no significant difference in croup scores or duration of hospitalisation in patients treated with oral dexamethasone compared to nebulised budesonide. Importantly, the authors contend that oral dexamethasone was more easily administered to distressed patients than nebulised budesonide, supporting the use of oral dexamethasone by paramedics over nebulised budesonide. Moreover, a study by Johnson et al (1998) comparing the effectiveness of nebulised budesonide and IM dexamethasone over a placebo found greater improvements in patients treated with IM dexamethasone compared to nebulised budesonide (p=0.003). Here it was contended that poor acceptance of the nebuliser mask may have led to low concentrations of the drug in the airways (Johnson et al, 1998; Çetinkaya et al, 2004). These findings suggest a poor acceptance of the nebulised route and thus support the use of IM and oral corticosteroids over nebulised corticosteroids by UK paramedics.

A final argument for the use of the nebulised route for corticosteroids is that nebulised adrenaline may be administered concurrently within the same mask, in cases of moderate to severe croup. In literature, the use of nebulised adrenaline is advocated in a Cochrane Review conducted by Bjornson et al (2011), while concurrent therapy with corticosteroids is supported in a randomised control trial by Duman et al (2005). In this study, nebulised L-adrenaline and concurrent IM or nebulised corticosteroid was compared against administration of cool mist and IM corticosteroids when treating moderate to severe croup. Results showed that a higher number of participants had Westley croup scores below 2 at 30 and 60 minutes when administered with either nebulised L-adrenaline and IM corticosteroid or nebulised L-adrenaline and nebulised corticosteroid (p=0.004 and p=0.032, respectively), with no significant difference between the two regimes.

From these results the authors advocate early combined therapy of L-adrenaline with either IM or nebulised corticosteroid for effective management of severe to moderate croup, in order to reduce hospitalisation. These findings support the use of both the IM and nebulised route of administration and importantly highlights a further practical advantage of nebulisation as an easy and efficient way to provide concurrent adrenaline therapy, should the mask be well tolerated.

While the previously discussed studies often compare two of the three commonly available routes against each other, only one study was identified comparing all three routes. Here, an RCT by Çetinkaya et al (2004) explored the efficacy of nebulised budesonide, oral and IM dexamethasone versus a placebo. Results were measured against Westley croup scored at 24, 48 and 72 hours where all patients in steroid treatment groups scored significantly lower than the placebo. In addition, the authors reported no significant difference between the steroid type and route, supporting the previous findings by Russel et al (2012).

Cost-effectiveness

The practical advantages and efficacy of the doses and routes of corticosteroids for croup have been explored within literature; however, a further aspect to importantly consider is the cost-effectiveness of such treatment regimens. A literature search was therefore conducted in order to extract appropriate evidence in this area. The search was conducted using Boolean operators and the terms ‘dexamethasone’ AND ‘budesonide’ AND ‘cost-effectiveness’; with a second search performed using ‘dexamethasone’ AND ‘budesonide’ AND ‘cost-analysis’ (Booth et al, 2012). From both searches no literature was found to comprehensively compare the cost-effectiveness of the mentioned drugs or either route of administration.

It is, however, noted that after finding no clinically significant difference between patients treated with oral dexamethasone, nebulised budesonide or both therapies, Klassen et al (1998) recommend oral dexamethasone as the drug of choice when treating croup. This is due to oral dexamethasone costing C$0.50 Canadian dollars per treatment compared to C$6.00 Canadian dollars per treatment of nebulised budesonide. The authors do, however, also advocate nebulised budesonide where oral dexamethasone is not well tolerated.

Alternatively, a cost analysis by Sumner et al (2010) was found comparing the cost-effectiveness of nebulised adrenaline only, oral dexamethasone only or both therapies in the treatment of paediatric bronchiolitis. Results from this study showed that the combination of nebulised adrenaline and oral dexamethasone was the most cost-effective treatment option, where average societal costs totalled $1 360 for dexamethasone alone, $1 322 for adrenaline alone and $1 115 for combination therapy.

Similar concurrent therapy has been previously discussed and has been shown to be efficacious in the treatment of croup in the study by Duman et al (2005). It may be therefore contended that a cost-effective management of croup, particularly in the more moderate-severe cases would incorporate treatment with both oral dexamethasone and nebulised adrenaline. This recommendation must, however, be supported by robust, UK-based cost analyses into such treatment as it is recognised that a Canadian, hospital-based cost analysis studying a different disease is not generalisable to UK paramedic practice and has limited external validity (Hulley et al, 2013).

Finally, a cost comparison of oral and IM dexamethasone against nebulised budesonide was also performed using prices sourced from the British National Formulary. Here it is found that dexamethasone as an oral solution of 2 mg/5 ml equates to £2.82 per 4 mg dose (British National Formulary, 2015b). As an intravenous preparation, currently administered orally by UK paramedics, 3.8 mg/1 ml forms a singular dose costing £1.99, which may also be given via the IM route (British National Formulary, 2015b). Nebulised budesonide, however, appears to be the most expensive, with a standard 2 mg dose equating to £4.00 per dose (British National Formulary, 2015a). An intravenous preparation administered orally or via injection has therefore proven to be the most cost-efficient method of administering a standard dose of corticosteroid, though it is maintained that research into this area is required to sufficiently support this.

Conclusions

From overall analysis of the literature, definitive support for the routine use of corticosteroids by paramedics when treating croup has been highlighted. At present, the fixed dose regimen remains appropriate, given that current evidence has highlighted a lack of proficiency by paramedics to perform drug calculations. This is supported by the opinion that adverse effects of corticosteroids are rarely seen, and also by evidence supporting the use of larger doses of corticosteroids in moderate/severe cases of croup. Evidence surrounding the efficacy of oral dexamethasone, IM dexamethasone and nebulised budesonide shows oral and IM dexamethasone to be comparable or occasionally superior to nebulised budesonide and also potentially more cost-effective.

Conversely, in cases where nebulised adrenaline is indicated, concurrent treatment with nebulised budesonide may be more beneficial to patients. Practitioners must therefore consider the practical advantages of each route, such as the pragmatic use of nebulised/IM corticosteroids for a child who is vomiting. This therefore provides an argument for the introduction of budesonide for nebulisation and an IM route of dexamethasone administration, alongside the oral route, in UK paramedic practice in order to promote efficient patient-centred care.

Key Points