Inhaled short-acting bronchodilators for managing emergency childhood asthma: an overview of reviews

Transcription

1 Allergy REVIEW ARTICLE Inhaled short-acting bronchodilators for managing emergency childhood asthma: an overview of reviews M. Pollock 1, I. P. Sinha 2, L. Hartling 1, B. H. Rowe 3, S. Schreiber 1 & R. M. Fernandes 4,5 1 Alberta Research Centre for Health Evidence, Department of Pediatrics, University of Alberta, Edmonton, AB, Canada; 2 Institute of Child Health, Alder Hey Children s Hospital, University of Liverpool, Liverpool, UK; 3 Department of Emergency Medicine and School of Public Health, University of Alberta, Edmonton, AB, Canada; 4 Clinical Pharmacology Unit, Instituto de Medicina Molecular, University of Lisbon Lisboa, Portugal; 5 Department of Pediatrics, Lisbon Academic Medical Centre, Santa Maria Hospital, Lisboa, Portugal To cite this article: Pollock M, Sinha I, Hartling L, Rowe BH, Schreiber S, Fernandes RM. Inhaled short-acting bronchodilators for managing emergency childhood asthma: an overview of reviews. Allergy 2017; 72: Keywords asthma; bronchodilator; children; emergency department; overview of reviews. Correspondence Lisa Hartling, Edmonton Clinic Health Academy, Avenue NW, Edmonton, AB T6G-1C9, Canada. Tel.: Fax: hartling@ualberta.ca Accepted for publication 30 August 2016 DOI: /all Edited by: Marek Sanak Abstract International guidelines provide conflicting recommendations on how to use bronchodilators to manage childhood acute wheezing conditions in the emergency department (ED), and there is variation within and among countries in how these conditions are managed. This may be reflective of uncertainty about the evidence. This overview of systematic reviews (SRs) aimed to synthesize, appraise, and present all SR evidence on the efficacy and safety of inhaled short-acting bronchodilators to treat asthma and wheeze exacerbations in children 0 18 years presenting to the ED. Searching, review selection, data extraction and analysis, and quality assessments were conducted using methods recommended by The Cochrane Collaboration. Thirteen SRs containing 56 relevant trials and 5526 patients were included. Results demonstrate the efficacy of short-acting beta-agonist (SABA) delivered by metered-dose inhaler as first-line therapy for younger and older children (hospital admission decreased by 44% in younger children, and ED length of stay decreased by 33 min in older children). Short-acting anticholinergic (SAAC) should be added to SABA for older children in severe cases (hospital admission decreased by 27% and 74% when compared to SABA and SAAC alone, respectively). Continuous nebulization, addition of magnesium sulfate to SABA, and levosalbutamol compared to salbutamol cannot be recommended in routine practice. Acute wheeze in children is one of the most common reasons for presentation to the emergency departments (ED) (1). In school-aged children, acute wheeze usually signifies underlying asthma, while in preschool children this sign may be the result of either asthma or viral-induced wheeze with distinct phenotypes and underlying mechanisms. In these age groups, regardless of the underlying pathology, exacerbations of recurrent wheeze and dyspnoea are largely caused by bronchoconstriction of the small airways. Although most children improve with bronchodilator therapy, many require hospital admission, and others may even require admission to critical or intensive care units. It is important, therefore, that children with wheeze and dyspnoea are treated promptly and appropriately in the ED to reduce symptoms and respiratory distress, mitigate hospital admission, and prevent respiratory failure. Across the UK and many other countries, there is a 10-fold difference in rates of hospital admission for acute asthma in children (2). Although the reasons for this are complex, one factor may be variations among EDs in the management of pediatric asthma, which have recently been described for key interventions such as inhaled bronchodilators (3). The three most common groups of inhaled bronchodilators used to treat acute asthma and wheeze in EDs are short-acting beta-agonists (SABA) such as salbutamol (albuterol), levosalbutamol (levalbuterol), and adrenaline (epinephrine); short-acting anticholinergics (SAACs) such as ipratropium bromide; and, more recently, a third category of muscle relaxants comprising magnesium sulfate (MgSO 4 ) (4). These drugs act on different cellular pathways (5), and can be administered using different routes (e.g. via metered-dose inhaler (MDI) with or without a spacer device, or as a nebulized formulation, either individually or in a combined preparation) and at various intervals (e.g. intermittently, back-toback, and continuously). Allergy 72 (2017) John Wiley & Sons A/S. Published by John Wiley & Sons Ltd 183

2 Bronchodilators for emergency childhood asthma Pollock et al. To inform how acute asthma and wheezing in children should be managed in the ED, all available evidence regarding therapeutic management must be collated and appraised. Systematic reviews (SRs) that do this tend to focus on one particular clinical question; however, given the several options around inhaled bronchodilators, a detailed structured overview may help inform the development of local, national, and international guidelines. The objective of this overview is to present a comprehensive synthesis of evidence from SRs on the efficacy and safety of inhaled short-acting bronchodilators to treat asthma and wheezing exacerbations in children aged 0 18 years presenting to the ED. The focus of this overview is the use of bronchodilators rather than other interventions, such as corticosteroids, and we excluded parenteral administration. It is also restricted to the ED setting as the severity of asthma and goals of therapy typically vary among children presenting to the ED compared to those admitted with refractory or severe asthma. Moreover, the overview is restricted to pediatric studies as the pharmacology and pathophysiology of acute asthma and wheeze differs in adults. Methods This overview was conducted using methods presented in The Cochrane Handbook for Systematic Reviews of Interventions (6). This methodology encourages the presentation of a robust, succinct summary of evidence of clinically important outcomes from SRs addressing different questions. Through all stages of this project, a protocol was followed in which all outcomes and analyses were specified a priori. Reporting adhered to PRISMA guidelines (7). Inclusion criteria for reviews Studies We used a standard definition of SRs (8) and included all SRs containing randomized controlled trials (RCTs) and/or controlled clinical trials (CCTs). Participants Systematic reviews evaluating therapies for children aged 0 18 years treated in the ED or equivalent care setting for acute exacerbation of asthma or recurrent wheeze were included; studies of single first episodes of wheeze or bronchiolitis were excluded. From the eligible SRs, we included results from trials that contained children only, and trials that contained both children and adults only if the pediatric data could be extracted separately. Interventions and comparators Permitted interventions included any inhaled short-acting bronchodilator, including SABAs, SAACs, and other shortacting bronchodilators such as MgSO 4, administered via MDI with or without a spacer or valved holding chamber (VHC), or nebulizer. All comparators were permitted, including other short-acting bronchodilators, placebos, or comparators described as standard care or similar by SR authors. All doses and frequencies of administration (i.e. single, repeated, intermittent, and continuous administration) were included, and co-interventions were permitted and reported. Outcome measures The primary and secondary outcomes were selected a priori based on clinical relevance, using a consensus process among the study s clinicians (IS, BHR, RMF). Primary outcomes were as follows: hospital admission; length of stay (LOS) in the ED; and admission to the intensive care unit (ICU), with or without ventilator support. Secondary outcomes were as follows: clinical severity scores, assessed using any appropriate tool (as determined by SR authors); vital signs (i.e. respiratory rate, heart rate, and oxygen saturation); and adverse effects (e.g. nausea, vomiting, tremor, and other general or specific outcomes considered undesirable by SR authors). Data on pulmonary function were also collected when available, including peak expiratory flow (PEF) and forced expiratory volume in 1-s (FEV 1 ). Search methods for identification of reviews A comprehensive search of the following electronic databases was conducted by an information specialist (RF) between March and May 2014: Cochrane Database of Systematic Reviews (CDSR) via Cochrane Library (from 2005), MED- LINE via Ovid (from 1946), EMBASE via Ovid (from 1980), Database of Abstracts of Reviews of Effects (DARE) (2nd quarter 2014), the Health Technology Assessment database (2nd quarter 2014) via Cochrane Library, and the Cumulative Index to Nursing and Allied Health Literature (CINAHL) Database via EBSCOhost (from 1937). Searches used validated child (9) and SR (10) filters, as appropriate. Searches were restricted to published English-language documents from 1990 to the present. We reviewed reference lists of relevant SRs to identify additional SRs. Update searches were conducted in the CDSR and Ovid MEDLINE in November 2015 (Appendix S1). Selection of reviews Titles and abstracts were screened by one reviewer (CJ), and a second reviewer (MP) assessed all excluded documents to ensure accuracy. Full texts of all potentially relevant documents were retrieved and assessed independently for inclusion by two reviewers (MP, AW), with discrepancies resolved by third-party adjudication (IS, RMF). Assessment of methodological quality of reviews Methodological quality of each included SR was assessed independently by two reviewers (MP, IS, SS) using A MeaSurement Tool to Assess systematic Reviews (AMSTAR) (11). Discrepancies were resolved through consensus. Data collection and analysis Data were extracted and analyzed by one reviewer and verified for accuracy by a second reviewer, with discrepancies 184 Allergy 72 (2017) John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

3 Pollock et al. Bronchodilators for emergency childhood asthma resolved through consensus (MP, SS). For each SR, evaluations were made to see whether trials matched the eligibility criteria of the overview. The following data were extracted from SRs: descriptive characteristics of SRs and their included trials; risk of bias (RoB) assessments for the trials contained within the included SRs as conducted by the SR authors, if available (for the domains of sequence generation, allocation concealment, blinding, incomplete outcome data, and selective outcome reporting) (12); and outcome data. In cases where data on trial characteristics and RoB of trials were not available in the included SRs, the missing data were extracted and/or the missing RoB assessments were completed independently by two reviewers with discrepancies resolved through consensus (MP, SS), using the primary research paper. When a trial was contained within more than one included SR, the outcome data were extracted from the most comprehensive SR to ensure accuracy of effect estimates. For dichotomous outcomes, data were extracted at the latest reported time point. For continuous outcomes, data were extracted at the latest time point up to and including 1 h after treatment, and at the latest time point after 1 h of treatment. Outcome data were extracted from SRs and meta-analyzed using REVIEW MANAGER 5.3. (The Nordic Cochrane Centre, The Cochrane Collaboration, 2014.) Dichotomous outcomes were expressed using risk ratios (RRs) with 95% confidence intervals (CIs); for primary outcomes with statistically significant results, the number needed to treat (NNT) with 95% CI was also calculated, with baseline risk estimates coming from pooled estimates of control groups (13). Continuous outcomes were expressed using mean differences with 95% CIs, except for results from clinical scores, which were expressed using standardized mean differences (SMDs) with 95% CIs. Standardized mean differences were used because we anticipated heterogeneity in measurement tools used across SRs, and results were described as small (<0.40), moderate ( ), or large (>0.70) (14). All analyses used random-effects models. Presentation of outcome data was planned separately for three age groups: 0 2, 2 6, and 6 18 years. This was decided a priori in an effort to examine whether age influenced any treatment or adverse effects. Due to the nature of the age subgroups contained within the included SRs, outcome data were extracted and presented for younger children aged 0 3 years, and older children aged 3 18 years. For primary outcomes, post hoc subgroup analyses were conducted based on severity of asthma, using the definitions provided by SR authors and/or trialists. This resulted in six subgroups: mild asthma, mild and moderate asthma, moderate asthma, moderate and severe asthma, severe asthma, and all asthma severities. We did not use any additional methods to test for differences between subgroups. The quality of evidence for each outcome was evaluated using four domains of the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) tool: study limitations, precision, consistency, and directness (15). For primary outcomes, we specified two a priori clinical decision thresholds for precision based on expert opinion and GRADE guidance: 25% decrease in hospital admission, and 30-min decrease in ED LOS (16). Quality was assessed independently by two reviewers (MP, LH), with discrepancies resolved through consensus. Results Results of the search The literature search retrieved 2428 unique references. Of 130 potentially eligible SRs, 13 were included (Fig. 1) (17 29). Description of included reviews Tables 1 and 2 describe the characteristics of the included SRs published in the CDSR ( Cochrane SRs ) or in other sources, respectively. All SRs were published between 1997 and Ten SRs searched for RCTs only, two searched for RCTs and CCTs (19, 25), and one did not report study designs (29). The SRs contained 189 trials and patients, of which 82 trials (43%) and 8040 patients (49%) matched our eligibility criteria and were included in the overview. Thirty-two trials were included in a single SR, 22 were included in two SRs, and two were included in three SRs. After accounting for overlapping trials contained within multiple SRs, 56 unique trials (5526 patients) remained (Table 3; Appendix S2). Search methods used in the included SRs All SRs searched MEDLINE, 11 searched CENTRAL (17, 19 28), and nine searched EMBASE (17 22, 25 28). Four or less searched AMED (17, 19, 20, 22), PsycINFO (17, 19, 20, 22), PubMed (23, 25), and clinicaltrials.gov (19), and all Cochrane SRs searched the Cochrane Airways Group Specialized Register of Trials. Common additional sources searched included the following: handsearching reference lists, journals, meeting abstracts, and/or conference proceedings (17 22, 24 29); and contacting authors, clinicians, collaborators, and/or drug companies (17 19, 21, 22, 24, 27, 29). Participants Age ranges and clinical definitions of asthma and wheezing in the included SRs varied slightly. Two SRs included children aged 0 2 years diagnosed with acute wheeze (21, 23), one included children aged 0 5 years diagnosed with acute wheeze or asthma (28), and 10 SRs included children between 1.5 and 18 years of age (or of unspecified age) diagnosed with acute asthma (17 20, 22, 24 27, 29). Interventions and comparators The included SRs provided outcome data relating to nine comparisons: SABA vs placebo (23), SABA vs SAAC (20), SABA vs adrenaline (26), SABA vs SABA (24), SABA and SAAC vs SABA alone (18, 21, 27), SABA and SAAC vs SAAC alone (20), SABA and MgSO 4 vs SABA alone (19, 25), SABA delivered via MDI with spacer or VHC vs nebulizer (17, 28, 29), and SABA delivered by continuous vs intermittent nebulization (22). No SR provided data examining Allergy 72 (2017) John Wiley & Sons A/S. Published by John Wiley & Sons Ltd 185

4 Bronchodilators for emergency childhood asthma Pollock et al. Records identified through electronic database search (n = 3329) Unique records remaining after duplicates removed (n = 2428) Records excluded after title and abstract screening (n = 2300) Records included after hand search (n = 2) Full-text articles assessed for eligibility (n = 130) Systematic reviews meeting full eligibility criteria (n = 13) Full-text articles excluded (n = 117) Not SR (n = 68) Not acute asthma/wheeze (n = 11) Not inhaled bronchodilator (n = 26) Not ED (n = 10) Protocol (n = 1) Outside of scope (n = 1) Figure 1 Flow diagram of references through the overview. SAAC vs placebo. Co-interventions were permitted in four SRs (17, 19, 21, 22) and not reported in nine SRs. Outcome measures All 13 included SRs provided data on outcomes that were identified a priori as being of interest for the overview. Eleven SRs provided data on hospital admission (17 20, 22 25, 27 29), four provided data on ED LOS (17, 22, 24, 29), and no SRs provided data on ICU admission with or without ventilator support. Eight SRs provided data on vital signs (17, 18, 21, 23, 24, 26, 28, 29), eight provided data on clinical severity scores (18, 20, 22 24, 27 29), and nine provided data on adverse effects (17 20, 22, 24, 26, 27, 29). Nine SRs also provided data on pulmonary function (17 20, 22, 24, 25, 27, 29). Methodological quality of included reviews and their trials The quality of the included SRs was good. Scores on the AMSTAR tool ranged from 4 (29) to 10 (19, 20, 24), with a mean of 7.8/11, and more than 50% of the SRs met the criteria for 9/11 domains (Appendix S3). Quality of the trials contained within the included SRs was adequate: More than 40% of trials were at low RoB for random sequence generation and allocation concealment; more than 60% of trials were at low RoB for blinding and incomplete outcome data; and more than 70% of trials were at low RoB for selective outcome reporting (Appendix S4). Effects of interventions For younger children (below 3 years of age), outcome data were available for three comparisons (Tables 4 and 5). For older children (3 18 years of age), data were available for seven comparisons (Tables 6 8). Summary of findings tables for primary outcomes are located in Appendix S5. Younger children aged 0 3 years SABA vs placebo. One trial (28 children) found no effect on hospital admission (P = 0.51; low-quality evidence). At 30 min, however, younger children treated with SABA showed a large decrease in clinical severity score (SMD: 186 Allergy 72 (2017) John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

5 Pollock et al. Bronchodilators for emergency childhood asthma Table 1 Characteristics of the seven included Cochrane* systematic reviews Systematic review title First author, year of publication (reference) Total number of studies (child, ED, and inhaled therapy only) Pooled sample size (child, ED, and inhaled therapy only) Population Definition of acute asthma Setting Intervention Comparator Holding chambers (spacers) versus nebulizers for betaagonist treatment of acute asthma Cates, 2013 (17) 39 (19) 2556 (1503) Children and adults 2 years of age Acute asthma Community setting, ED, and patients admitted to hospital Any beta2-agonist given by any nebulizer. Co-interventions permitted The same beta2- agonist given by MDI with any spacer Combined inhaled anticholinergics and short-acting beta2- agonists for initial treatment of acute asthma in children Griffiths, 2013 (18) Inhaled magnesium sulfate in the treatment of acute asthma Powell, 2012 (19) Anticholinergic therapy for acute asthma in children Teoh, 2012 (20) 20 (20) 2695 (2695) Children between 18 months and 18 years of age Acute exacerbation of asthma 16 (2) 824 (79) All ages Acute asthma (any reasonable diagnosis, namely clinical and guidelinebased criteria). Patients with chronic or stable asthma were excluded 6 (3) 411 (130) Children between 2 and 18 years of age Acute asthma All settings (ED, observation unit, inpatient, outpatient, general practice and home) ED Single or repeated doses of nebulized or inhaled SAAC plus SABA NR Inhaled MgSO 4 alone or in combination with inhaled beta2- agonist. Co-interventions were permitted Single or repeated doses of nebulized or inhaled placebo plus SABAs Inhaled beta2- agonist alone, placebo alone, or placebo in combination with inhaled beta2- agonist and ipratropium bromide Inhaled SAAC Placebo, beta2- agonist, SAAC plus beta2-agonists, or any other drugs or drug combinations Overview outcomes for which data are reported Hospital admission; ED LOS; respiratory rate; heart rate; oxygen saturation; adverse effects; PEF; FEV1 Hospital admission; clinical severity score; oxygen saturation; adverse effects; PEF; FEV 1 Hospital admission; adverse effects; FEV 1 Hospital admission; clinical severity score; adverse effects; PEF; FEV 1 Allergy 72 (2017) John Wiley & Sons A/S. Published by John Wiley & Sons Ltd 187

6 Bronchodilators for emergency childhood asthma Pollock et al. Table 1 (continued) Systematic review title First author, year of publication (reference) Total number of studies (child, ED, and inhaled therapy only) Pooled sample size (child, ED, and inhaled therapy only) Population Definition of acute asthma Setting Intervention Comparator Overview outcomes for which data are reported Anticholinergic drugs for wheeze in children under the age of two years Everard, 2005 (21) Continuous versus intermittent betaagonists for acute asthma Camargo, 2003 (22) Short-acting beta2- agonists for recurrent wheeze in children under two years of age Chavasse, 2002 (23) 6 (2) 322 (130) Children under 2 years of age 8 (1) 664 (70) Children and adults 8 (1) 313 (28) Children under 2 years of age Wheeze due to reversible airway obstruction. Studies recruiting premature infants or subjects with chronic lung disease or widespread crepitations on auscultation of the chest were excluded Home, ED, and patients admitted to hospital SAAC therapy alone or combined with additional therapy Acute asthma ED or equivalent Continuous inhaled beta2-agonists. Co-interventions were permitted Two or more episodes of wheeze which was not related to another form of chronic lung disease. Studies of first episode of wheeze or neonatal respiratory illness were excluded Home, ED, hospital ward, or pulmonary function laboratory A single or multiple doses of nebulized, inhaled, or oral beta-agonist Placebo or beta2-agonist Intermittent inhaled beta2- agonists Respiratory rate; oxygen saturation Hospital admission; ED LOS; clinical severity score; adverse effects; PEF Placebo Hospital admission; clinical severity score; respiratory rate; oxygen saturation ED, emergency department; FEV 1, forced expiratory volume in 1-s; LOS, length of stay; MDI, metered-dose inhaler; NR, not reported; PEF, peak expiratory flow; SAAC, short-acting anticholinergic; SABA, short-acting beta-agonist. *Published in the Cochrane Database of Systematic Reviews. 188 Allergy 72 (2017) John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

7 Pollock et al. Bronchodilators for emergency childhood asthma Table 2 Characteristics of the six included non-cochrane* systematic reviews Systematic review title First author, year of publication (reference) Total number of studies (child, ED, and inhaled therapy only) Pooled sample size (child, ED, and inhaled therapy only) Population Levalbuterol versus albuterol for acute asthma: a systematic review and meta-analysis Jat, 2013 (24) 7 (4) 1625 (381) Adults or children Intravenous and nebulized magnesium sulfate for treating acute asthma in adults and children: a systematic review and meta-analysis Shan, 2013 (25) Comparison between nebulized adrenaline and beta-2 agonists for the treatment of acute asthma. A meta-analysis of randomized trials Rodrigo, 2006 (26) 25 (1) 1766 (62) Adults or children 6 (1) 282 (121) Children (18 months 17 years) and adults ( 18 years) Anticholinergics in the treatment of children and adults with acute asthma: a systematic review with meta-analysis Rodrigo, 2005 (27) 32 (15) 4045 (1918) Children (18 months to 17 years) and adults ( 18 years) Definition of acute asthma Setting Intervention Comparator Acute asthma ED and patients admitted to hospital Levosalbutamol (levalbuterol) Salbutamol (albuterol) Acute asthma NR Intravenous or nebulized MgSO 4 as an adjuvant in combination with beta2-agonists Beta2-agonists and placebo Acute asthma exacerbation according to accepted criteria (e.g. American Thoracic Society) Acute exacerbation of asthma ED or equivalent care setting ED or equivalent care setting Single or repeated doses of inhaled adrenaline Single or repeated doses of inhaled SAAC agents given in combination with inhaled beta2- agonists Inhaled beta2- agonist Inhaled beta2- agonist alone Overview outcomes for which data are reported Hospital admission; ED LOS; clinical severity score; respiratory rate; heart rate; oxygen saturation; adverse effects; PEF; FEV 1 Hospital admission; FEV 1 Heart rate; adverse effects Hospital admission; clinical severity score; adverse effects; FEV 1 Allergy 72 (2017) John Wiley & Sons A/S. Published by John Wiley & Sons Ltd 189

8 Bronchodilators for emergency childhood asthma Pollock et al. Table 2 (continued) Systematic review title First author, year of publication (reference) Total number of studies (child, ED, and inhaled therapy only) Pooled sample size (child, ED, and inhaled therapy only) Population Definition of acute asthma Setting Intervention Comparator Overview outcomes for which data are reported Beta-agonists through metered-dose inhaler with valved holding chamber versus nebulizer for acute exacerbation of wheezing or asthma in children under 5 years of age: a systematic review with meta-analysis Castro-Rodriguez, 2004 (28) Metered-dose inhaler accessory devices in acute asthma Amirav, 1997 (29) 6 (6) 491 (491) Infants or preschool children (<5 years of age) Acute exacerbation of wheezing or asthma ED or equivalent care setting Any beta-agonist given by MDI with any VHC 10 (7) 552 (432) Children Acute asthma NR Inhaled medications by MDI and accessory device The same beta-agonist given by any nebulizer Inhaled medications by current standard of care, SVNs Hospital admission; clinical severity score; respiratory rate; heart rate; oxygen saturation Unable to extract usable data ED, emergency department; FEV 1, forced expiratory volume in 1-s; LOS, length of stay; MDI, metered-dose inhaler; NR, not reported; PEF, peak expiratory flow; SAAC, short-acting anticholinergic; SVN, small-volume nebulizer; VHC, valved holding chamber. *Published in a source other than the Cochrane Database of Systematic Reviews. 190 Allergy 72 (2017) John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

9 Pollock et al. Bronchodilators for emergency childhood asthma Table 3 Map of primary studies contained within each included review* Study ID Cochrane systematic reviews Non-Cochrane systematic reviews Cates 2013 Griffiths 2013 Powell 2012 Teoh 2012 Everard 2005 Camargo 2003 Chavasse 2002 Jat 2013 Shan 2013 Rodrigo 2006 Rodrigo 2005 Castro-Rodriguez 2004 Amirav 1997 Total number of reviews (/13) Andrews 2009 Yes 1 Ashtekar 2008 Yes 1 Batra 1997 Yes 1 Beck 1985 Yes Yes 2 Benito Fernandez 2000 Yes Yes 2 Bentur 1992 Yes 1 BI 2009 Yes 1 Calvo 1998 Yes 1 Chakraborti 2006 Yes 1 Chong-Neto 2005 Yes 1 Chou 1995 Yes Yes 2 Closa 1998 Yes 1 Cook 1985 Yes Yes Yes 3 Delgado 2003 Yes 1 Direkwatanachai 2011 Yes 1 Duarte 2002 Yes 1 Ducharme 1998 Yes Yes 2 Ferres 1989 Yes 1 Freelander 1984 Yes Yes 2 Fuglsang 1986 Yes 1 Guill 1987 Yes Yes 2 Iramain 2011 Yes 1 Jamalvi 2006 Yes 1 Kerem 1993 Yes Yes 2 Khine 1996 Yes 1 Leveshra 2000 Yes Yes 2 Lin 1995 Yes Yes 2 Mahajan 2004 Yes Yes 2 Maldano-Alanis 1997 Yes 1 Mandelberg 2000 Yes 1 Naspitz 1992 Yes 1 Pendergast 1989 Yes Yes 2 Peterson 1994 Yes Yes 2 Phanichyakam 1990 Yes Yes 2 Plint 2000 Yes 1 Ploin 2000 Yes Yes 2 Punj 2009 Yes 1 Quershi 1997 Yes Yes 2 Allergy 72 (2017) John Wiley & Sons A/S. Published by John Wiley & Sons Ltd 191

10 Bronchodilators for emergency childhood asthma Pollock et al. Table 3 (continued) Study ID Cochrane systematic reviews Non-Cochrane systematic reviews Cates 2013 Griffiths 2013 Powell 2012 Teoh 2012 Everard 2005 Camargo 2003 Chavasse 2002 Jat 2013 Shan 2013 Rodrigo 2006 Rodrigo 2005 Castro-Rodriguez 2004 Amirav 1997 Total number of reviews (/13) Quershi 1998 Yes Yes 2 Quershi 2005 Yes 1 Reisman 1988 Yes Yes 2 Robertson 1998 Yes 1 Rubilar 2000 Yes 1 Sannier 2007 Yes 1 Schuh 1992 Yes 1 Schuh 1995 Yes Yes 2 Sharma 2004 Yes Yes 2 Sienra Monge 2000 Yes Yes 2 Timsit 2002 Yes 1 Valencia 1999 Yes 1 Vasquez 1992 Yes 1 Watanasomsiri 2006 Yes 1 Watson 1988 Yes Yes Yes 3 Wilkinson 2011 Yes 1 Williams 1996 Yes Yes 2 Zorc 1999 Yes Yes 2 BI, Boehringer Ingelheim Pharmaceuticals. *Included primary studies are those conducted in children presenting to the emergency department who received treatment with an inhaled bronchodilator. Cochrane systematic reviews are those published in the Cochrane Database of Systematic Reviews; non-cochrane systematic reviews are those published in other sources. 192 Allergy 72 (2017) John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

11 Pollock et al. Bronchodilators for emergency childhood asthma Table 4 Primary outcomes in younger children aged 0 3 years (or with a mean age between 0 and 3 years) Outcome Comparison Number of subjects (trials) Measure of effect (95% CI) Direction of effect I 2 (%) Quality of evidence (GRADE)* Hospital admission SABA vs placebo 28 (1) RR = 1.73 ( ) NS NA Low SABA delivered by MDI with spacer or VHC vs nebulizer 490 (6) RR = 0.56 ( ) Favors MDI with spacer or VHC 0 High CI, confidence interval; GRADE, Grading of Recommendations Assessment, Development, and Evaluation; MDI, metered-dose inhaler; NA, not applicable; NS, not significant; RR, risk ratio; SABA, short-acting beta-agonist; VHC, valved holding chamber. *See Appendix S5 for a breakdown of GRADE assessments by domain. Table 5 Secondary outcomes in younger children aged 0 3 years (or with a mean age between 0 and 3 years) Outcome Description of outcome Comparison Number of subjects (trials) Measure of effect (95% CI) Direction of effect I 2 (%) Quality of evidence (GRADE)* Outcomes measured up to and including 1 h Clinical severity score Mean change (30 min) SABA vs placebo 28 (1) SMD = 1.31 ( 2.14 to 0.48) Respiratory rate Heart rate Oxygen saturation Absolute values (measured at multiple time points ) Mean change (30 min) Mean change (20 min) % change (20 min) Mean change (30 min) % change (20 min) Outcomes measured after 1 h Respiratory rate Oxygen saturation Mean change (2 h) Mean change (2 h) SABA delivered by MDI with spacer or VHC vs nebulizer 367 (4) SMD = 0.48 ( 0.77 to 0.19) SABA vs placebo 28 (1) MD = 5.10 ( 9.45 to 0.75) SABA delivered by MDI 60 (1) MD = 1.34 with spacer or ( 2.71 to 5.39) VHC vs nebulizer SABA delivered by MDI with spacer or VHC vs nebulizer 60 (1) MD = 7.15 ( to 3.51) SABA vs placebo 28 (1) MD = 1.60 (0.33 to 2.87) SABA delivered by MDI with spacer or VHC vs nebulizer SABA and SAAC vs SABA alone SABA and SAAC vs SABA alone 60 (1) MD = 0.33 ( 1.52 to 0.86) 69 (1) MD = 2.00 ( 6.77 to 2.77) 69 (1) MD = 0.08 ( 0.84 to 1.00) Favors SABA NA Very low Favors MDI with spacer or VHC 40 Low Favors SABA NA Very low NS NA Very low Favors MDI with spacer or VHC NA Very low Favors SABA NA Very low NS NA Very low NS NA Very low NS NA Very low CI, confidence interval; GRADE, Grading of Recommendations Assessment, Development, and Evaluation; MD, mean difference; MDI, metered-dose inhaler; NA, not applicable; NS, not significant; SAAC, short-acting anticholinergic; SABA, short-acting beta-agonist; SMD, standardized mean difference; VHC, valved holding chamber. *GRADE assessments for all comparisons were as follows: no serious study limitations ; imprecise ( 1) ; inconsistent ( 1) ; and indirect ( 1). The one exception was absolute clinical severity score (measured at multiple time points), which was rated as consistent. The four trials included in this meta-analysis measured clinical severity score at four different time points (10 min, 15 min, 20 min, and 1 h). 1.31; 95% CI: 2.14 to 0.48), a decrease in respiratory rate (MD: 5.10; 95% CI: 9.45 to 0.75), and an increase in oxygen saturation (MD: 1.60; 95% CI: ). No data on adverse effects were reported. SABA delivered by MDI with spacer or VHC vs nebulizer. High-quality evidence from six trials (490 children) found that younger children receiving SABA administered via MDI with spacer or VHC compared to nebulizer had a 44% decrease in hospital admission (RR: 0.56; 95% CI: ). Absolute event rates were 11.3% (MDI) and 21.7% (nebulizer), and the NNT to prevent one additional hospital admission was 10 (95% CI: 6 27). Subgroup analyses showed that the decrease in hospital admission was Allergy 72 (2017) John Wiley & Sons A/S. Published by John Wiley & Sons Ltd 193

12 Bronchodilators for emergency childhood asthma Pollock et al. Table 6 Primary outcomes in older children aged 3 18 years (or with a mean age between 3 and 18 years) Outcome Comparison Number of subjects (trials) Measure of effect (95% CI) Direction of effect I 2 (%) Quality of evidence (GRADE)* Hospital admission ED LOS (in minutes) SABA vs SAAC 59 (2) RR = 0.27 (0.08 to 0.98) Favors SABA 0 Low SABA (levosalbutamol) vs 240 (2) RR = 0.98 (0.63 to 1.51) NS 0 Moderate SABA (salbutamol) SABA and SAAC vs 2497 (19) RR = 0.73 (0.63 to 0.85) Favors SABA 0 High SABA alone and SAAC SABA and SAAC vs 61 (2) RR = 0.26 (0.07 to 0.92) Favors SABA 0 Moderate SAAC alone and SAAC SABA and MgSO 4 vs 62 (1) RR = 2.00 (0.19 to 20.93) NS NA Low SABA alone SABA delivered by MDI 784 (10) RR = 0.73 (0.45 to 1.21) NS 0 Low with spacer or VHC vs nebulizer SABA delivered by continuous vs intermittent nebulization 70 (2) RR = 0.71 (0.36 to 1.40) NS 0 Low SABA (levosalbutamol) vs 139 (1) MD = 4.00 ( to 81.50) NS NA Low SABA (salbutamol) SABA delivered by MDI 396 (3) MD = ( to 23.65) Favors MDI 66 Low with spacer or VHC with spacer vs nebulizer or VHC SABA delivered by 70 (1) MD = 1.00 ( to 11.50) NS NA Very low continuous vs intermittent nebulization CI, confidence interval; ED, emergency department; GRADE, Grading of Recommendations Assessment, Development, and Evaluation; LOS, length of stay; MD, mean difference; MDI, metered-dose inhaler; MgSO 4, magnesium sulfate; NA, not applicable; NS, not significant; RR, risk ratio; SAAC, short-acting anticholinergic; SABA, short-acting beta-agonist; VHC, valved holding chamber. *See Appendix S5 for a breakdown of GRADE assessments by domain. limited to younger children with moderate and severe asthma only, with no differences shown in trials with participants with moderate asthma (Appendix S6). Results for secondary outcomes (clinical severity score, heart rate, respiratory rate, oxygen saturation) were inconsistent (Table 5). No data on adverse effects were reported. SABA and SAAC vs SABA alone. One trial (69 children) examined this comparison. No data were available for primary outcomes, and no statistically significant differences in secondary outcomes (respiratory rate, oxygen saturation) were observed (Table 5). No data on adverse effects were reported. Older children aged 3 18 years SABA vs SAAC. Low-quality evidence from two trials (59 children) found that treating older children with SABA (fenoterol, metaproterenol) compared to SAAC (ipratropium bromide, atropine sulfate) resulted in a 73% reduction in hospital admission (RR: 0.27; 95% CI: ). Absolute event rates were 7.4% (SABA) and 27.6% (SAAC), and the NNT was 5 (95% CI: 3 71). Moderate and large decreases in clinical severity score were observed at 30 min (two trials, 59 children; SMD: 0.56; 95% CI: 1.09 to 0.04) and 2 h (one trial, 31 children; SMD: 1.09; 95% CI: 1.86 to 0.33), respectively. There were no statistically significant group differences in unspecified side effects, PEF, or FEV 1 (Appendix S7). SABA vs adrenaline. One trial (121 children) examined this comparison. No data were available for primary outcomes. No statistically significant difference in heart rate was observed between groups (Table 7). Adverse effects data were inconsistent (no difference in serious adverse effects, but higher incidence of nasal symptoms in the adrenaline group). SABA (levosalbutamol) vs SABA (salbutamol). When comparing levosalbutamol to salbutamol, moderate-quality evidence from two trials (240 children) found no difference in hospital admission (RR = 0.98; 95% CI: ), and lowquality evidence from one trial (139 children) found no difference in ED LOS (P = 0.93). Outcome data for secondary outcomes (clinical severity score, heart rate, respiratory rate, oxygen saturation) were inconsistent (Table 5). There were no statistically significant differences between groups in adverse effects (tremor, nausea/vomiting, and headache/nervousness), PEF, or FEV 1 (Table 8; Appendix S7). 194 Allergy 72 (2017) John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

13 Pollock et al. Bronchodilators for emergency childhood asthma Table 7 Secondary outcomes in older children aged 3 18 years (or with a mean age between 3 and 18 years) Outcome Name of outcome Comparison Number of subjects (trials) Measure of effect (95% CI) Direction of effect I 2 (%) Quality of evidence (GRADE)* Outcomes measured up to and including 1 h Clinical severity score Absolute values (30 min) Absolute SABA vs SAAC SABA (levosalbutamol) 59 (2) 60 (1) SMD = 0.56 ( 1.09 to 0.04) SMD = 2.08 values (1 h) vs SABA (salbutamol) ( 2.72 to 1.45) % Change (1 h) SABA (levosalbutamol) 240 (2) SMD = 0.12 vs SABA (salbutamol) ( 0.13 to 0.38) Absolute values SABA and SAAC vs 61 (2) SMD = 0.56 (30 min) SAAC alone ( 1.07 to 0.04) Respiratory Absolute SABA (levosalbutamol) 60 (1) MD = 0.77 rate values (1 h) vs SABA (salbutamol) ( 2.65 to 1.11) % Change (1 h) SABA (levosalbutamol) 240 (2) MD = 0.42 vs SABA (salbutamol) ( 9.29 to 8.46) Mean change SABA delivered by MDI 609 (7) MD = 0.49 (measured with spacer or VHC ( 2.20 to 1.22) at multiple vs nebulizer time points ) Heart rate Oxygen saturation Mean change (15 min) Absolute values (1 h) SABA vs adrenaline 121 (1) MD = 0.65 ( 4.63 to 5.93) SABA (levosalbutamol) vs SABA (salbutamol) % Change (1 h) SABA (levosalbutamol) vs SABA (salbutamol) % Change (measured at multiple time points ) Absolute values (1 h) No. of subjects with oxygen saturation <95% (1 h) % Change (measured at multiple time points ) Outcomes measured after 1 h Clinical severity score SABA delivered by MDI with spacer or VHC vs nebulizer SABA (levosalbutamol) vs SABA (salbutamol) % Change (1 h) SABA (levosalbutamol) vs SABA (salbutamol) SABA and SAAC vs SABA alone SABA delivered by MDI with spacer or VHC vs nebulizer 60 (1) MD = ( to 7.47) 240 (2) MD = 3.61 ( 5.75 to 12.96) 704 (9) MD = 5.81 ( 9.38 to 2.24) 60 (1) MD = 0.50 ( 4.21 to 5.21) 240 (2) MD = 0.38 ( 2.98 to 2.23) 416 (2) RR = 0.73 (0.55 to 0.96) 366 (4) MD = 0.08 ( 0.56 to 0.40) Absolute value (2 h) SABA vs SAAC 31 (1) SMD = 1.09 ( 1.86 to 0.33) Mean change (measured at multiple time points ) Absolute value (2 h) Mean change (2 h) SABA and SAAC vs SABA alone SABA and SAAC vs SAAC alone SABA delivered by continuous vs intermittent nebulization 934 (3) SMD = 0.23 ( 0.42 to 0.04) 32 (1) SMD = 1.39 ( 2.17 to 0.60) 70 (1) SMD = 0.66 (0.18 to 1.14) Favors SABA 0 Low Favors SABA NA Very low (levosalbutamol) NS 0 Low Favors SABA 0 Low and SAAC NS NA Very low NA 0 Low NS 49 Very low NS NA Very low Favors SABA NA Very low (levosalbutamol) NS 0 Low Favors MDI with spacer or VHC 55 Very low NS NA Very low NS 0 Low Favors SABA and SAAC 0 Low NS 0 Very low Favors SABA NA Very low Favors SABA and SAAC Favors SABA and SAAC Favors continuous nebulization 0 Low NA NA Very low Very low Allergy 72 (2017) John Wiley & Sons A/S. Published by John Wiley & Sons Ltd 195

14 Bronchodilators for emergency childhood asthma Pollock et al. Table 7 (continued) Outcome Name of outcome Comparison Number of subjects (trials) Measure of effect (95% CI) Direction of effect I 2 (%) Quality of evidence (GRADE)* Oxygen saturation No. of subjects with oxygen saturation <95% (2 h) SABA and SAAC vs SABA alone 185 (2) RR = 1.08 (0.63 to 1.83) NS 50 Very low CI, confidence interval; GRADE, Grading of Recommendations Assessment, Development, and Evaluation; MD, mean difference; MDI, metered-dose inhaler; NA, not applicable; NS, not significant; RR, risk ratio; SAAC, short-acting anticholinergic; SABA, short-acting beta-agonist; SMD, standardized mean difference; VHC, valved holding chamber. *GRADE assessments for all comparisons were as follows: no serious study limitations ; imprecise ( 1) ; consistent ; and indirect ( 1). Exceptions were as follows: The four comparisons examining mode of delivery of SABAs had serious study limitations ( 1), all comparisons including one trial were inconsistent ( 1), and % change in heart rate (measured at multiple time points) and No. of subjects with oxygen saturation <95% (2 h) were inconsistent ( 1). The trials included in these meta-analyses measured outcomes at different time points. SABA and SAAC vs SABA alone or SAAC alone. High-quality evidence from nineteen trials (2497 children) and moderate-quality evidence from two trials (61 children) found that combined treatment with SABA and SAAC reduced hospital admissions in older children by 27% and 74% when compared with SABA alone (RR: 0.73; 95% CI: ) and SAAC alone (RR: 0.26; 95% CI: ), respectively. For combined therapy compared to SABA alone, the absolute event rates were 16.9% (combined SABA and SAAC) and 23.2% (SABA alone), and the NNT was 18 (95% CI: 11 41). For combined therapy compared to SAAC alone, absolute event rates were 6.3% (combined SABA and SAAC) and 24.4% (SAAC alone), and the NNT was 5 (95% CI: 2 42). The most common drug combination used in these trials was salbutamol and ipratropium bromide (18 trials), the most common mode of administration was via nebulizer (18 trials), and the most common frequency of administration was either two or three doses of SABA and SAAC administered min apart (14 trials). Subgroup analyses for combined treatment compared to SABA alone showed that the decrease in hospital admission was limited to older children with severe and moderate and severe asthma only (Appendix S6). Compared to SABA alone, combined treatment had an inconsistent effect on secondary outcomes (clinical severity score, oxygen saturation) and adverse effects (tremor, nausea, vomiting). Compared to SAAC alone, combined treatment led to moderate and large decreases in clinical severity score at 30 min (two trials, 61 children; SMD: 0.56; 95% CI: 1.07 to 0.04) and 2 h (one trial, 32 children; SMD: 1.39; 95% CI: 2.17 to 0.60), respectively, and had no statistically significant effect on unspecified side effects. For both comparisons, results for PEF and FEV 1 were inconsistent (Appendix S7). SABA and MgSO 4 vs SABA alone. One trial (62 children) found no difference in hospital admission (P = 0.56; lowquality evidence) and no statistically significant difference in adverse effects (tremor, nausea, vomiting, hypotension, headache, deep tendon jerks) or FEV 1 (Table 8; Appendix S7). SABA delivered by MDI with spacer or VHC vs nebulizer. Ten trials (784 children) failed to identify a difference in hospital admission for older children receiving SABA administered via MDIs with spacers or VHCs compared to nebulizers (RR = 0.73; 95% CI: ; low-quality evidence). Three trials (396 children) with significant (P = 0.05) and substantial (I 2 = 66%) heterogeneity found that the use of MDIs with spacers or VHCs led to a decrease in ED LOS of about half an hour (MD: 33.48; 95% CI: to 23.65; low-quality evidence). The decreased LOS occurred for older children regardless of asthma severity (Appendix S6). Results for secondary outcomes (heart rate, respiratory rate, and oxygen saturation) were inconsistent (Table 5). Short-acting beta-agonist delivered by MDI compared to nebulizers decreased incidence of tremor by 37% (four trials, 237 children; RR: 0.63; 95% CI: ). Results for PEF and FEV 1 were inconsistent (Appendix S7). SABA delivered by continuous vs intermittent nebulization. There was low- and very low-quality evidence for older children receiving SABA delivered by continuous compared to intermittent nebulization: No differences in hospital admission (two trials, 70 children; RR = 0.71; 95% CI: ) or ED LOS (one trial, 70 children; P = 0.88) were identified, respectively. One trial (70 children) found that children receiving SABA delivered by continuous compared to intermittent nebulization showed a large decrease in clinical severity score at 2 h (SMD: 0.66; 95% CI: ). There were no statistically significant differences between groups for adverse events (tremor, and nausea/vomiting) or PEF (Table 8; Appendix S7). Discussion This overview represents a systematic, comprehensive, and thorough review of the evidence regarding inhaled bronchodilators for the management of acute asthma in children presenting to the ED. The overview revealed high-quality evidence that the use of MDI to deliver SABA in children 196 Allergy 72 (2017) John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

15 Pollock et al. Bronchodilators for emergency childhood asthma Table 8 Adverse effects in older children aged 3 18 years (or with a mean age between 3 and 18 years)* Outcome Comparison Number of subjects (trials) Measure of effect (95% CI) Direction of effect I 2 (%) Quality of evidence (GRADE) Tremor Nausea Vomiting Nausea/vomiting SABA (levosalbutamol) vs SABA (salbutamol) SABA and MgSO 4 vs SABA alone SABA delivered by MDI with spacer or VHC vs nebulizer SABA delivered by continuous vs intermittent nebulization SABA and SAAC vs SABA alone SABA and MgSO 4 vs SABA alone SABA and SAAC vs SABA alone SABA and MgSO 4 vs SABA alone SABA (levosalbutamol) vs SABA (salbutamol) SABA delivered by continuous vs intermittent nebulization Hypotension SABA and MgSO 4 vs SABA alone Headache/ nervousness Deep tendon jerks SABA (levosalbutamol) vs SABA (salbutamol) SABA and MgSO 4 vs SABA alone SABA and MgSO 4 vs SABA alone 240 (2) RR = 0.94 (0.69 to 1.30) NS 0 Moderate 62 (1) Not estimable (no events in either group) NA NA NA 237 (4) RR = 0.63 (0.43 to 0.92) Favors MDI 0 Low with spacer or VHC 70 (1) RR = 0.56 (0.21 to 1.49) NS NA Very low 757 (7) RR = 0.59 (0.37 to 0.93) Favors SABA and SAAC 0 Moderate 62 (1) Not estimable NA NA NA (no events in either group) 1230 (8) RR = 0.81 (0.44 to 1.49) NS 0 Moderate 62 (1) Not estimable (no NA NA NA events in either group) 240 (2) RR = 0.65 (0.29 to 1.48) NS 64 Low 70 (1) RR = 0.20 (0.01 to 4.02) NS NA Very low 79 (2) RR = 4.13 (0.19 to 88.71) NA NA Low 240 (2) RR = 1.82 (0.82 to 4.05) NS 0 Moderate 62 (1) Not estimable (no events in either group) 62 (1) Not estimable (no events in either group) NA NA NA NA NA NA CI, confidence interval; GRADE, Grading of Recommendations Assessment, Development, and Evaluation; MDI, metered-dose inhaler; MgSO 4, magnesium sulfate; NA, not applicable; NS, not significant; RR, risk ratio; SAAC, Short-acting anticholinergic; SABA, short-acting beta-agonist; VHC, valved holding chamber. *Not all adverse effects data could be meta-analyzed. See Results text for additional adverse effects data for the following four comparisons: SABA vs SAAC, SABA vs adrenaline, SABA and SAAC vs SABA alone, and SABA and SAAC vs SAAC alone. GRADE assessments for all comparisons were as follows: no serious study limitations ; imprecise ( 1) ; consistent ; and direct. Exceptions were as follows: The two comparisons examining continuous nebulization had serious study limitations ( 1) and were inconsistent ( 1), and the comparison examining SABA and MgSO 4 had serious study limitations ( 1) for the outcome of tremor and was inconsistent ( 1) for the outcome of hypotension. younger than 3 years reduces hospital admission compared with nebulizers. In children older than 3 years, low-quality evidence failed to identify any difference between these methods of delivery on hospital admission; however, MDI and spacer appear to be beneficial with regard to the other primary outcome, ED LOS. There is evidence that in children over 3 years of age, SABA should be combined with SAAC, as opposed to giving either treatment alone, for moderate and severe exacerbations. Continuous nebulization does not appear to reduce the risk of hospital admission, based on low-quality evidence in children over 3 years of age. The British Thoracic Society (BTS) and the Global Initiative for Asthma (GINA) are the two most widely recognized guidelines in the United Kingdom for the emergency management of childhood asthma and wheeze. Both guidelines have converging recommendations around the use of bronchodilators for children with acute asthma in the ED, despite a few relevant discrepancies (30, 31). The BTS guidelines suggest Allergy 72 (2017) John Wiley & Sons A/S. Published by John Wiley & Sons Ltd 197

16 Bronchodilators for emergency childhood asthma Pollock et al. that SABA should be the first bronchodilator for both younger (below 2 years of age) and older children, preferably given via MDI and spacer for mild or moderate attacks, or oxygen-driven nebulizers in case of hypoxia (SpO 2 < 92%). Nebulized ipratropium bromide should be added early in all children who are poorly responsive to SABA, and given repeatedly in the first hours, and nebulized MgSO 4 can be considered in older children with severe exacerbations of short duration. The general approach of the GINA guidelines is mostly similar, with both guidelines emphasizing the need to individualize dosing according to severity and response. However, there are differences in the recommended use of ipratropium bromide, via either MDI with spacer or nebulizer, for moderate-to-severe exacerbations restricted to a first hour in younger children (below 5 years of age). We should note that grading of severity of exacerbations is distinct between guidelines. Further, continuous nebulized SABA use is considered in GINA, while the BTS guidelines found no advantage when compared to frequent intermittent doses. On the basis of the evidence reviewed here, SABA should be given as the first-line bronchodilator, with the use of MDI and spacer supported in children with mild to moderate attacks. No specific SABA was favored in terms of efficacy nor safety. Combining ipratropium bromide and SABA may be helpful in children with moderate-to-severe disease. These results are in line with the current guideline recommendations. On the contrary, there is no clear evidence favoring the use of nebulizers based on the severity of the exacerbation. Most nebulizers allow the provision of oxygen in case of hypoxia, and may reduce the risk of ventilation/perfusion mismatching that may occur with SABA administration. Our results, however, do not suggest any clinically relevant benefit from their routine use even for severe exacerbations. In addition, the use of nebulization increases the risk of transmission of infectious agents through respiratory droplet spread, which is generally unmeasured in trial data. Further, no evidence was identified to support the recommendations for continuous doses of SABA therapy or the addition of MgSO 4, and these should be restricted to children with severe or refractory asthma attacks. A recent survey of practice among EDs in the UK identified variations in the management of acute asthma in children (3). While most clinicians surveyed used inhaled salbutamol as first-line therapy, other practices did not reflect the results of this overview. Firstly, 43% of clinicians used ipratropium bromide immediately; however, we suggest that best practice would be to combine this drug with salbutamol only if their asthma attack is moderate to severe or a child is not responding to standard care. Secondly, although it is commonplace for clinicians to use three consecutive doses initially, followed by reassessment, this does not appear to be needed routinely and may in fact lead to increased risk of adverse effects of therapy. Individualized care should be encouraged. Thirdly, 64% of UK clinicians used nebulizers for hypoxic children and MDIs for nonhypoxic children, although the overview findings suggest MDI and spacer should be used whenever possible, regardless of severity. In two Canadian surveys of practice (32, 33), which also found high rates of use of nebulizers, some barriers to changing practice toward using MDI and spacer were concerns regarding safety (of sterilizing and reusing spacers); concerns regarding increased costs and workload; perceived patient and parental resistance (which appear from three studies to be unfounded) (34 36); lack of a local clinician who would champion the need for change; and lack of perceived effectiveness of MDI and spacer. This overview provides a comprehensive synthesis of an extensive body of evidence that was conducted using systematic, rigorous methods designed to avoid bias and enhance reliability and validity of results. We have also been able to extract and appraise evidence specifically around the use of bronchodilators in children. The data upon which these conclusions are derived arise largely from moderate-quality trials contained within moderate-quality SRs, and the conclusions are based on clinically important primary outcomes. Although outcome data for one of our primary outcomes LOS may be influenced by extraneous variables unrelated to our research question, the data from primary outcomes are supported by data on multiple relevant secondary outcomes. Moreover, while these overview findings are consistent with another recent review (37), by re-extracting outcome data from included SRs (and primary trials, as required) and re-analyzing these data in the overview, we were able to ensure adherence to our inclusion criteria at the study level; standardize method of presentation of outcome data; and conduct detailed analyses by age, timing of outcome measurement, and asthma severity. Efforts were used throughout the overview to avoid publication and selection bias; however, we recognize that these cannot always be completely eliminated. This overview pertains specifically to the treatment of asthma and wheeze in the ED, and conclusions should not be extrapolated to management in hospital or critical care settings. Future research should address the comparisons for which there is low-quality evidence, as well as methodological issues such as standardization of exacerbation severity and core outcome set development. Conclusions Based on current evidence, mild and moderate asthma should be treated by administering SABA through MDI and spacer (or VHC) rather than nebulizer, particularly in younger children. In children over 3 years of age with refractory, severe, or life-threatening acute asthma requiring nebulization, SABA should be combined with SAAC; however, there is no evidence that continuous nebulization is helpful at initial presentation to the ED. Further research is needed to evaluate the use of bronchodilators in children under 3 years, and to examine the role of nebulized MgSO 4 alone or in combination. Acknowledgments The authors wish to thank Robin Featherstone, Catherene Joseph, and Aireen Wingert for assistance with searching, screening, and inclusion, respectively. MP is supported by an Alberta Innovates Health Solutions Graduate Studentship and a Knowledge Translation Canada Graduate Student Fellowship. BHR is supported by a Tier I Canada Research Chair in Evidence-Based Emergency Medicine from the 198 Allergy 72 (2017) John Wiley & Sons A/S. Published by John Wiley & Sons Ltd