Figures
- FIGURE 1
Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram, showing details of the search and study inclusion process, including reasons for exclusion of full-text articles reviewed. CINAHL: Cumulative Index to Nursing and Allied Health.
- FIGURE 2
Summary of quality assessment using the Effective Public Health Practice Project quality assessment tool for quantitative studies. The domains blinding, intervention integrity and analyses were not applicable for any of the studies and were therefore removed. n/a: not applicable. #: retrospective case–control studies were given a n/a score in the withdrawals and dropouts domain in accordance with the tool recommendations.
- FIGURE 3
Exacerbation outcomes, showing the number of studies that relate different categories of exacerbations at baseline to risk of future exacerbations. ED: emergency department; ICU: intensive care unit.
- FIGURE 4
Forest plot of studies assessing past exacerbation (unspecified severity) predicting future exacerbation. Data presented separately for five of the databases used in Engelkes et al. [38]. AUH: Aarhus University prescription database; CPRD: Clinical Practice Research Datalink; IPCI: Integrated Primary Care Information; SIDIAP: Sistema d'Informació per al Desenvolupament de la Investigació en Atenció Primària.
- FIGURE 5
Forest plots based on exacerbation severity definition and location of study publication. In d), data are presented separately for five of the databases used in Engelkes et al. [38]. PICU: paediatric intensive care unit.
Tables
- TABLE 1
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 visitChildren aged 2–17 years with ED visit for acute asthma
Mean age 7.99 years
59% male
19% White, 55% Black, 24% Hispanic, 2% other1184 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 yearChildren 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% other452 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-upChildren aged 4–18 years, hospitalised for asthma
Mean age 8.22 years
65% male
77% African American, 21% White, 2% other115 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 yearIndividuals aged 3–64 years with asthma
Children
62.2% male11 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% other2032 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-upCases: 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% other445
Cases: 152
Controls: 293Hospitalisation 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 weeksChildren 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 group285 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 visitChildren 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% other197 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 visitChildren aged 2–17 years visiting ED for acute asthma
70% aged <7 years
59% male269 enrolled;
247 completed 1-month follow-up,
220 completed 6-month follow-upSudden 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 studyUSA TENOR conducted from 2001 to 2004
Follow-up with semi-annual visits for 3 yearsChildren 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% other4756
637 children aged 6–11 years
Data available for 563 childrenUse 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 hospitalisationChildren aged 2–15 years admitted to hospital for asthma
Mean age 6.35 years
66% male
36% White, 35% Black, 24% Hispanic, 5% other298 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 yearsChildren aged 2–18 years with asthma admitted to ICU
Median age at admission 7.0 years
59% male410 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 yearsChildren aged 5–12 years with mild/moderate persistent asthma
60% male
68% White, 14% Black, 9% Hispanic, 9% other1041 enrolled
1019 completed daily diary cardsEpisode 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 dataOntario, Canada ED visit between 14 April 2006 and 28 February 2009
Follow-up for 1 yearChildren aged 2–17 years with prevalent asthma with unplanned visit to ED for asthma
38.4% aged 2–5 years
63.1% male29 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% White230
Cases: 66
Controls: 164PICU 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-upChildren aged ≤14 years admitted to hospital for acute asthma
50% aged >6 years
64.5% male76
1 admission: 56
Readmission: 20Acute 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-upChildren 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% other36 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% Black2451
With eosinophil count: 333Asthma 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 yearsChildren with asthma aged 5–18 years
Mean age: 10.5 years
58.7% male14 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 periodSARCA trial:
children aged 6–17 years, with poor asthma control, being treated with inhaled glucocorticoids
62% male
39% White, 49% Black, 12% other718 enrolled
Data for 295 from SARCARequirement 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 OCSCosta, 2018 [33]# Retrospective case–control Goiania, Brazil June 2012–August 2013
Data collected for 1 year prior to ED visitCases: 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% White153
Cases: 92
Controls: 61Increased 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 UniversityGeorgia, 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% other579
Cases: 170
Controls: 409PICU 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-upIndividuals 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 quintile58 366
18 352 aged 5–19 yearsED 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 yearChildren aged 5–12 years diagnosed with asthma
Mean age 9 years
57% male3776 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% male320
Cases: 72
Controls: 248PICU 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 databases212 060
Severe asthma 14 283Use 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.
- TABLE 2
Key results shown by index and subsequent exacerbation type
Category of subsequent exacerbation ED Hospital PICU Category of index exacerbation ED 5 studies
ORs 3.3 [23], 6.27 [21], 8.26 [14]
HR 2.03 [26]
Relative risk 1.13 [35]3 studies
ORs 3.1 [23], 3.28 [18]
HR 1.85 [26]No studies Hospital 2 studies
OR 0.85 [14]
HR 1.45 [26]6 studies
ORs 1.89 [18], 3.3 [24], 3.37 [16], 3.6 [29], 5.36 [15]
HR 2.87 [26]3 studies
ORs 5.4 [27], 8.19 [34], 16.9 [24]PICU No studies 2 studies
ORs 2.18 [18], 29.62 [28]2 studies
ORs 2.4 [24], 7.83 [37]ED: emergency department; PICU: paediatric intensive care unit; HR: hazard ratio.
Supplementary Material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary material 00174-2022.SUPPLEMENT
Vol 8 Issue 4
Table of Contents
Past asthma exacerbation in children predicting future exacerbation: a systematic review
