Summary of key study characteristics of included studies
| First author, year [reference] | Study design | Study setting | Data collection period | Population/inclusion criteria | Sample size (n) | Definition of acute exacerbation | Results |
| Emerman, 2001 [13]# | Prospective cohort combining two studies with identical protocols | 44 EDs in USA and Canada | Studies performed 1997–1998 Follow-up 2 weeks after index ED visit | Children aged 2–17 years with ED visit for acute asthma Mean age 7.99 years 59% male 19% White, 55% Black, 24% Hispanic, 2% other | 1184 recruited; follow-up data available for 762 | ED visit with physician diagnosed acute asthma | Factors associated with acute asthma relapse: ED visits for asthma in past year (per 5 visits) OR 1.2 (1.0–1.5) Urgent clinic visits for asthma in past year (per 5 visits) OR 1.1 (0.9–1.3) |
| Lafata, 2002 [14] | Retrospective cohort study using routinely acquired data | Michigan, USA | 1992–1996 2-year observation: 1 baseline year and 1 follow-up year | Children aged 5–14 years, with 1 hospitalisation or 2 outpatient encounters for asthma and ≥1 paediatrician office visit for each year of inclusion Mean age 8.7 years 63% male 49% White, 44% African American, 7% other | 452 | ED visit or hospitalisation for asthma | Factors associated with ED use: ED visit for asthma in prior year OR 8.26 (4.79–14.25) Hospital admission for asthma in prior year OR 0.85 (0.32–2.22) Factors associated with ED or hospital admission for asthma: Prior ED visit for asthma OR 7.97 (4.64–13.71) |
| Chen, 2003 [15] | Prospective cohort | Children's Hospital, St Louis, MO, USA | Admissions between June and December 1999 1 year follow-up | Children aged 4–18 years, hospitalised for asthma Mean age 8.22 years 65% male 77% African American, 21% White, 2% other | 115 | Hospitalisation for asthma | Lifetime history of hospitalisations as a predictor of future hospitalisation: OR 5.36 (1.90–15.14) |
| Schatz, 2003 [16] | Retrospective cohort study using routinely acquired data | California, USA | 1998–1999 2-year observation: 1 baseline year and 1 follow-up year | Individuals aged 3–64 years with asthma Children 62.2% male | 11 101 in total; 6904 children aged 3–17 years | Hospitalisation or ED visit for asthma | In children aged 3–17 years, 1998 hospitalisations as a predictor for asthma hospitalisation in 1999 OR 3.37 (1.61–7.04) |
| McCoy, 2006 [17] | Data from RCT used as an observational study | 19 American Lung Association clinical research centres, USA | Recruitment from 15 September to 30 November 2000 Follow-up for 14 days after each injection (28 days total) | Volunteers aged 3–64 years with physician-diagnosed asthma Of children originally enrolled, 60% male 60.5% White, 28.7% Black, 5.6% Hispanic, 4.4% other | 2032 enrolled, 1949 completed trial; 353 children aged 3–10 years | New or increased OCS or an unscheduled healthcare encounter for asthma | History of intubation for asthma, hospitalisation ≥2 times for asthma, ≥3 courses of OCS for asthma in past year, or ≥2 unscheduled health contacts for asthma in past year in children aged 3–10 years as a predictor of exacerbations OR 2.19 (1.18–4.06) |
| Reznik, 2006 [18] | Retrospective case–control | Children's hospital, New York, USA | Admissions between January 1998–December 2004 30-day follow-up | Cases: children aged 0–21 years hospitalised for asthma and readmitted within 30 days of discharge for same reason Controls: children aged 0–21 years hospitalised for asthma, but not readmitted within 30 days of discharge Mean age 5.99 years 61% male 62.2% Hispanic, 34.2% African American, 3.6% other | 445 Cases: 152 Controls: 293 | Hospitalisation for asthma | Predictors of early asthma readmission: ED visit for asthma in past year OR 3.28 (1.55–6.94) Multivariate analysis of predictors of early asthma readmission: Hospital admission for asthma in past year OR 1.89 (1.10–3.25) Prior ICU admission for asthma OR 1.99 (0.93–4.27) |
| Covar, 2008 [19]# | Data from PACT RCT used as an observational study | USA | Recruited between October 2002 and January 2004 Trial period: 48 weeks | Children aged 6–14 years with documented mild–moderate persistent asthma, screening FEV1 ≥80% predicted and methacholine reactivity 61.4% male 44.9% from a minority ethnic group | 285 | Systemic corticosteroids or emergency care (ED visit or hospitalisation) for acute asthma | Logistic regression analysis of factors at baseline predictive of exacerbation: History of exacerbation requiring corticosteroid course in past year OR 2.28 (1.59–3.26) Multivariable model of factors associated with exacerbations: Prednisone course in year prior to study OR 2.10 (1.42–3.09) |
| Miller, 2008 [20] | Prospective cohort | Michigan, USA | Enrolment over 1-year period Follow-up at 2 weeks and 6 months post-ED visit | Children aged 2–17 years presenting to ED for acute asthma Mean age 8.1 years 61.5% male 71.7% White, 50% Black, 26% Hispanic, 7% American Indian or Alaska native, 2% Asian, 2% other | 197 enrolled; follow-up data available for 166 | Signs/symptoms compatible with asthma exacerbation (shortness of breath, coughing, wheezing, chest tightness) in a diagnosed asthmatic | Previous severe disease (e.g. systemic corticosteroids, ED visit or hospitalisation for asthma) as a predictor of 6-month morbidity (urgent care, ED or hospital admissions for asthma) Pearson correlation coefficient 0.17 |
| To, 2008 [21] | Prospective cohort | Children's hospital, Toronto, Canada | ED visit between January 2003 and June 2004 Follow-up at 1 and 6 months post-ED visit | Children aged 2–17 years visiting ED for acute asthma 70% aged <7 years 59% male | 269 enrolled; 247 completed 1-month follow-up, 220 completed 6-month follow-up | Sudden worsening of symptoms resulting in difficulty breathing often requiring extra medicine to relieve symptoms, with/without unscheduled ED/doctor visit | Predictors of acute asthma episode at 6-month follow-up: Acute asthma episode 6 months prior to baseline OR 4.73 (2.25–9.97) Predictors of ED visit at 6-month follow-up: ED visits in 12 months prior to baseline OR 6.27 (1.54–7.12) |
| Haselkorn, 2009 [22] | Prospective cohort Data from TENOR study | USA | TENOR conducted from 2001 to 2004 Follow-up with semi-annual visits for 3 years | Children with severe asthma or mild/moderate asthma considered difficult to treat Had ≥2 OCS bursts in past year; ≥2 unscheduled clinic or hospital visits for asthma in past year; requirement for chronic, daily high doses of ICS; or ≥5 mg oral prednisone or current use of ≥3 medications to control asthma 69% male 62% White, 38% other | 4756 637 children aged 6–11 years Data available for 563 children | Use of a corticosteroid burst | Multivariate model including 6- and 12-month events Recent exacerbation as a predictor of future exacerbation OR 1.99 (1.51–2.61) Multivariate model including only 6-month events: Recent exacerbation as a predictor of future exacerbation OR 3.08 (2.21–4.28) |
| Tolomeo, 2009 [23] | Retrospective cohort study using routinely acquired data | Children's hospital, New England, USA | Hospitalisation between January and December 2006 Data for 1 year before and after hospitalisation | Children aged 2–15 years admitted to hospital for asthma Mean age 6.35 years 66% male 36% White, 35% Black, 24% Hispanic, 5% other | 298 | Hospital admission with primary diagnosis of asthma | Previous asthma-related ED visit as a predictor of: Subsequent ED visit OR 3.3 (1.39–7.96) Subsequent hospitalisation for asthma OR 3.1 (1.17–8.33) |
| Triasih, 2011 [24]# | Retrospective cohort | Children's hospital, Melbourne, Australia | ICU admission between January 1990 and December 2004 Mean follow-up 10.3 years | Children aged 2–18 years with asthma admitted to ICU Median age at admission 7.0 years 59% male | 410 | ICU admission for asthma | Risk factors for readmission to hospital: Previous hospital admission OR 3.3 (2.1–5.3) Admission in year prior to index admission aOR 4.5 (2.5–8.4) Multiple previous hospital admissions OR 2.4 (1.3–4.2) Risk factors for readmission to ICU: Previous hospital admission OR 16.9 (4.1–70.4) Admission in year prior to index admission aOR 4.7 (2.4–9.3) Multiple previous hospital admissions OR 3.2 (1.6–6.7) Previous ICU admission aOR 2.4 (0.8–6.7) |
| Wu, 2011 [25]# | Data from CAMP RCT used as an observational study | USA | Enrolment between December 1993 and September 1995 Follow-up over 4 years | Children aged 5–12 years with mild/moderate persistent asthma 60% male 68% White, 14% Black, 9% Hispanic, 9% other | 1041 enrolled 1019 completed daily diary cards | Episode requiring ≥3 days use of OCS, hospitalisation, or ED visit due to asthma (ATS/ERS statement) | History of ED visits or hospitalisations in prior year as a predictor of having ≥1 severe exacerbations regression coefficient 0.73 (0.50–0.96) History of ≥3 days of treatment with OCS in prior 3 months as a predictor of having ≥1 severe exacerbations regression coefficient 0.40 (0.17–0.62) |
| Li, 2012 [26] | Retrospective cohort study using routinely acquired data | Ontario, Canada | ED visit between 14 April 2006 and 28 February 2009 Follow-up for 1 year | Children aged 2–17 years with prevalent asthma with unplanned visit to ED for asthma 38.4% aged 2–5 years 63.1% male | 29 391 | ED visit for asthma | Adjusted hazard ratios for ED re-visits: Asthma admission(s) in prior 2 years HR 1.45 (1.35–1.55) Asthma ED visit(s) in prior 2 years HR 2.03 (1.91–2.14) Adjusted hazard ratios for hospital admissions: Asthma admission(s) in prior 2 years HR 2.87 (2.43–3.39) Asthma ED visit(s) in prior 2 years HR 1.85 (1.57–2.19) |
| van den Bosch, 2012 [27]# | Retrospective case–control | 4 hospitals in the Netherlands | January 1994–October 2006 | Cases: children with doctor-diagnosed asthma admitted to PICU for acute asthma Controls: patients with asthma who never needed PICU admission for any reason Median age at PICU admission 5.2 years 77% White | 230 Cases: 66 Controls: 164 | PICU admission for acute asthma | Earlier hospitalisation for asthma (non-PICU) as a risk factor for PICU admission OR 5.4 (1.34–21.45) |
| Visitsunthorn, 2013 [28] | Retrospective case–control | Children's hospital, Bangkok, Thailand | January 2006–December 2007 1 year follow-up | Children aged ≤14 years admitted to hospital for acute asthma 50% aged >6 years 64.5% male | 76 1 admission: 56 Readmission: 20 | Acute asthma that was severe or did not improve after 3 doses of bronchodilator nebulisation | ICU admission at first admission as a risk factor for readmission OR 29.62 (3.35–262.18) |
| Kenyon, 2014 [29] | Retrospective cohort study using routinely acquired data | USA | Discharges between 1 July 2008 and 30 June 2010 1 year follow-up | Children aged ≥2 years discharged from hospital after admission for asthma 61% aged 5–18 years 61.1% male 46.6% Black, 27.3% White, 16.8% Hispanic, 7.9% other | 36 601, contributing 44 203 hospitalisations | Hospital admission for asthma | Prior-year admission as a risk factor for asthma rehospitalisation 7 days OR 2.0 (1.4–2.7) 15 days OR 2.7 (2.1–3.3) 30 days OR 2.9 (2.5–3.4) 60 days OR 3.5 (3.1–3.9) 180 days OR 3.5 (3.3–3.8) 365 days OR 3.6 (3.4–3.8) |
| Zeiger, 2015 [30] | Retrospective cohort study using routinely acquired data | California, USA | 2010–2011: 1 baseline year and 1 outcome year | Children aged 5–11 years who met HEDIS criteria for persistent asthma Blood eosinophil level determined in 2010 With eosinophil level: Mean age 7.7 years 61.9% male 45.3% Hispanic, 24.6% White, 17.4% Black | 2451 With eosinophil count: 333 | Asthma outpatient visits requiring systemic corticosteroids within ±7 days or asthma ED visits or hospitalisation | History of exacerbation as a risk factor for exacerbation: Adjusted rate ratio 2.35 (1.61–3.44) Adjusted risk ratio 1.94 (1.37–2.73) |
| Engelkes, 2016 [31]# | Retrospective cohort study using routinely acquired data | The Netherlands | 1 January 2000 to 1 January 2012 Mean follow-up 2.46 years | Children with asthma aged 5–18 years Mean age: 10.5 years 58.7% male | 14 303 | Hospitalisation, ED visit or prescription of systemic corticosteroids for ≥3 days for asthma | Prior exacerbations as a risk factor for exacerbation: Total cohort model 1: relative rate 1.99 (1.40–2.83), model 2: relative rate 2.17 (1.30–3.60) Children with ≥1 exacerbation ever model 1: relative rate 1.60 (1.37–1.88), model 2: relative rate 1.52 (1.19–1.94) Exacerbations as a risk factor for nonfrequent exacerbations compared to frequent exacerbations <2 versus ≥2 OR 2.11 (1.66–2.68) versus OR 1.93 (1.42–2.63) <3 versus ≥3 OR 2.43 (1.84–3.23) versus OR 1.99 (1.35–2.94) |
| Quezada, 2016 [32]# | Data from SARCA and SARA RCTs used for an observational study | USA | 2007–2011 24-week treatment period | SARCA trial: children aged 6–17 years, with poor asthma control, being treated with inhaled glucocorticoids 62% male 39% White, 49% Black, 12% other | 718 enrolled Data for 295 from SARCA | Requirement for OCS or urgent healthcare visit for asthma symptoms | 83% of children with an exacerbation had an unscheduled visit for asthma in the previous year and 80% had been treated with OCS 69% of children without an exacerbation had an unscheduled visit for asthma in the previous year and 61% had been treated with OCS |
| Costa, 2018 [33]# | Retrospective case–control | Goiania, Brazil | June 2012–August 2013 Data collected for 1 year prior to ED visit | Cases: children aged 4–14 years, admitted to ED for asthma who had ≥3 previous episodes of bronchospasm Controls: asthmatic children without exacerbation recruited during outpatient appointment Cases: Median age 7 years 56% male 62% White Controls: Median age 8 years 42% male 28% White | 153 Cases: 92 Controls: 61 | Increased symptoms requiring change in medication, judged by physician according to ATS/ERS statement Severe exacerbation: hospital admission or course of OCS for asthma | ≥3 ED visits in past year for asthma as a risk factor for asthma exacerbation incidence risk ratio 1.40 (1.01–1.95) |
| Grunwell, 2018 [34] | Retrospective case–control Data from outpatient asthma clinical research studies at Emory University | Georgia, USA | January 2004–December 2015 | Cases: children aged 6–18 years with historical admission to PICU for acute asthma Controls: asthmatic children without prior PICU admission Cases: Median age 10 years 56.7% male 52.3% Black, 35.2% White, 12.5% other Controls: Median age 12 years 61.8% male 71.8% Black, 17.1% White, 11.2% other | 579 Cases: 170 Controls: 409 | PICU admission, hospitalisation or ED visit for asthma | Hospitalisation for asthma in year prior to data collection was associated with increased odds of PICU admission OR 8.19 (4.83–13.89) |
| To, 2018 [35] | Retrospective cohort study using routinely acquired data | Toronto, Canada | April 2008–March 2014 1 year follow-up | Individuals aged 5–99 years with ED visit for asthma Aged 5–19 years Mean age 12.15 years 57% male 31% in most marginalised ethnic concentration quintile | 58 366 18 352 aged 5–19 years | ED visit for asthma | Adjusted relative risk of asthma ED return visit within 1 year of ED discharge in children aged 5–19 years 1.13 (1.03–1.25) |
| Turner, 2018 [36]# | Retrospective cohort study using routinely acquired data | UK | January 1999–December 2012 1 baseline year and 1 outcome year | Children aged 5–12 years diagnosed with asthma Mean age 9 years 57% male | 3776 | Hospitalisation, ED admission or OCS for asthma (ATS/ERS) | Previous asthma attack as a risk factor for ≥1 attack 1 OR 3.74 (2.92–4.80) ≥2 OR 7.72 (5.55–10.74) |
| Alsheri, 2020 [37]# | Retrospective case–control | Abha Hospital, Saudi Arabia | January 2014–December 2018 | Cases: children aged 2–14 years admitted to PICU for acute asthma Controls: children admitted to the ward for acute asthma Cases Mean age 6.3 years 41.7% male Controls Mean age 4.6 years 42.7% male | 320 Cases: 72 Controls: 248 | PICU or hospital admission for asthma | Previous admission to PICU as a risk factor for PICU admission OR 7.83 (2.58–23.76) |
| Engelkes, 2020 [38]# | Retrospective cohort study using routinely acquired data from the Netherlands, Italy, UK, Denmark and Spain | The Netherlands | January 2008–December 2013 | Patients aged 5–17 years with asthma Subcohort with severe asthma (requiring high dose ICS+second controller and/or systematic corticosteroids for ≥120 consecutive days) Mean age 10.4 years (7.2–14.8 years across databases) Male preponderance across all databases | 212 060 Severe asthma 14 283 | Use of systemic corticosteroids, ED visit and/or hospitalisation, for worsening asthma | Relative rates of exacerbation in those with history of exacerbation CPRD 5.76 (5.25–6.33) SIDIAP 2.53 (2.27–2.81) IPCI 20.04 (12.91–31.10) AUH 45.71 (31.2–66.92) PEDIANET 29.36 (16.25–53.05) HSD 10.07 (4.56–22.20) |
Data are presented with 95% confidence intervals in brackets where provided. ED: emergency department; RCT: randomised controlled trial; OCS: oral corticosteroids; ICU: intensive care unit; PACT: Pediatric Asthma Controller Trial; FEV1: forced expiratory volume in 1 s; TENOR: The Epidemiology and Natural History of Asthma; ICS: inhaled corticosteroids; aOR: adjusted odds ratio; CAMP: Childhood Asthma Management Program; ATS: American Thoracic Society; ERS: European Respiratory Society; HR: hazard ratio; PICU: paediatric intensive care unit; HEDIS: Healthcare Effectiveness Data and Information Set; SARCA: Study of Acid Reflux in Children with Asthma; SARA: Study of Acid Reflux in Adults with Asthma; CPRD: Clinical Practice Research Datalink; SIDIAP: Sistema d'Informació per al Desenvolupament de la Investigació en Atenció Primària; IPCI: Integrated Primary Care Information; AUH: Aarhus University prescription database; HSD: Health Search Database. #: included in meta-analysis.