Respiratory pathogens and clinical outcomes in children with an asthma exacerbation: A systematic review

Abstract

Background

In asthmatic children, respiratory pathogens are identified in 60%–80% of asthma exacerbations, contributing to a significant burden of illness. The role of pathogens in the clinical evolution of exacerbations is unknown.

Objective

We systematically reviewed the association between the presence of pathogens and clinical outcomes in children with an asthma exacerbation.

Method

PubMed, EMBASE, BIOSIS, and the Cochrane Central Register of Controlled Trials were searched up to October 2016 for studies reporting on respiratory pathogen exposure and clinical outcome. The Risk of Bias in Non-Randomized Studies of Interventions tool was used for quality assessment.

Results

Twenty-eight observational studies (N = 4,224 children) reported on 112 different associations between exposure to any pathogen (n = 45), human rhinovirus (HRV; n = 34), atypical bacteria (n = 21), specific virus (n = 11), or bacteria (n = 1) and outcomes of exacerbation severity (n = 26), health care use (n = 38), treatment response (n = 19), and morbidity (n = 29). Restricting the analysis only to comparisons with a low to moderate risk of bias, we observed an association between HRV and higher exacerbation severity on presentation (regression p = .016) and between the presence of any pathogen and emergency department treatment failure (odds ratio [OR] = 1.57; 95% CI 1.04% to 2.37%). High-quality evidence for effect on morbidity or health care use is lacking.

Conclusions

Further research on the role of pathogen–treatment interaction and outcomes is required to inform the need for point-of-care, real-time testing for pathogens. Studies with a sufficiently large sample size that address selection bias, correctly adjust for confounding, and rigorously report core patient-centred outcomes are necessary to improve knowledge.

Introduction

Asthma is a chronic inflammatory disorder of the airways, and it is the single most common chronic disease of childhood (1,2). Asthma morbidity and costs are mainly due to acute exacerbations and associated hospital admissions (3). Respiratory pathogens, along with host–pathogen interactions, play a role in the onset of asthma (4), and both viral and bacterial pathogens are also known triggers of asthma exacerbations (5,6). Other endogenous and exogenous stimuli that may trigger or worsen asthma symptoms include allergens, tobacco smoke, exercise, and stress (7,8). Among children, 60%–80% of exacerbations are associated with viral pathogens and respiratory infections (4,5). Moreover, multiplex real-time polymerase chain reaction (RT-PCR) has enhanced the detection of a larger number of different pathogens, given its improved sensitivity, and sequencing has allowed for the identification of specific subtypes, resulting in a renewed interest in studying specific pathogens and their role in asthma exacerbations (9).

Human rhinovirus (HRV) has been identified as a pathogen prevalent among children with asthma exacerbations that follows seasonal peaks (10). To understand this phenomenon, studies have investigated the pathogenicity and interaction of HRV with the epithelial cells of the respiratory tract and with immunological pathways (11–13). The role, pathogenicity, and impact on clinical severity of individual HRV subtypes, including the recently discovered HRV-C, among children with and without underlying asthma, are still debated (14–16). Moreover, recent outbreaks of specific pathogens, such as enterovirus D68, have been linked with severe respiratory complications in children with asthma (17–19). The role of influenza and its association with the severity of asthma exacerbation and health care use remain controversial (6,20). The atypical bacteria, Mycoplasma pneumoniae and Chlamydia pneumoniae, have been associated with exacerbations; however, their impact on clinical evolution is not clear (21).

The evidence related to the strength of the association, if any, between the presence of a respiratory pathogen and severity of asthma exacerbation has not been systematically evaluated, nor has its clinical evolution. In addition, the actual importance of specific respiratory pathogens and their impact on the clinical picture, exacerbation evolution, and response to treatment (steroids and bronchodilators) remains unclear. Understanding the role of a particular respiratory pathogen among children with an exacerbation would help to evaluate the utility of infection prevention, pathogen identification at presentation, pathogen-adjusted treatment, and follow-up regimens.

In this systematic review, we summarize and critically analyze the available evidence from studies reporting on the association between respiratory pathogens with one or more clinical relevant outcomes—namely, the severity of asthma exacerbation, health care use, treatment response, morbidity, and mortality. Our objective was to compare short-term clinical outcomes of asthma exacerbations in children with and without an associated respiratory pathogen.

Methods

We used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for the design, performance, and reporting of this review (22). A research protocol using the PRISMA for Systematic Review Protocols guidelines specifying our methods was developed in advance (23).

Eligibility criteria: types of studies, participants, exposures, and outcomes

The review included studies providing original data examining the association between the presence of respiratory pathogens and patient-related outcomes of an existing asthma exacerbation in children aged younger than 18 years. Quantitative study designs, including all types of observational studies (cohort, case-control, cross-sectional, and longitudinal), and randomized controlled trials (RCT) were considered. Case reports, case series, ecological studies, and studies for which only an abstract was available were excluded. Studies investigating the association of respiratory pathogens and the onset of asthma and studies aiming only to determine infection or viral incidence in people with asthma were also excluded.

Participants were restricted to children with pre-existing and diagnosed asthma. Studies of children with two or more previous wheezing episodes were eligible, on the basis of their criterion validity in predicting the presence of asthma, which has been proposed by the Canadian Thoracic Society and Canadian Paediatric Society and used in previously published peer-reviewed literature (24–27). We excluded asthmatic patients without an acute exacerbation, non-asthmatic control groups, and adults.

The presence of respiratory pathogens—bacterial or viral—or of one specific respiratory pathogen had to be investigated systematically among the study population. For our systematic review, all diagnostic methods, including PCR, the current gold standard for diagnosis of the presence of viral respiratory infection, and other methods such as serology testing, enzyme-linked immunosorbent assay (ELISA), and rapid tests were accepted. Studies looking at sensitization to fungal elements were not within the scope of this review.

Defining a single primary outcome in pediatric asthma studies is complex. In 2009, a group of experts from the American Thoracic Society/European Respiratory Society defined a list of core health outcomes that had to be present in asthma research (28). Similarly, a wide consortium of governmental and non-governmental organizations held an Asthma Outcomes workshop in 2010 (29). We used the outcomes agreed on in both documents as guidance for the evaluation of asthma outcomes in the studies included in our systematic review. The outcomes were grouped under (a) severity of the asthma exacerbation, (b) health care use, (c) response to treatment, and (d) other indicators of morbidity and mortality (Box 1). Studies were included if they reported absolute numbers (means, proportions), inference testing with absolute numbers, or effect measures (odds ratios [ORs], relative risk, or model coefficients) with confidence intervals. Each study could provide more than one exposure–outcome comparison for assessment.

Box 1: Description of Outcome Domains.

1. Severity of asthma exacerbation

   • Severity of asthma exacerbation often expressed on severity scale (mild, moderate, or severe)
   • Hypoxia or oxygen therapy needed
   • Duration of oxygen treatment
2. Health care use

   • Need for systemic treatment
   • Hospital admission for asthma
   • Duration of hospitalization
   • ICU admission for asthma
   • Duration of ICU admission
   • Mechanical ventilation
   • Relapse or new consultation
3. Response to treatment

   • Length of active treatment
   • Duration of symptoms after treatment
   • Number of bronchodilator doses needed
   • Need for corticosteroids
   • Change in lung function after treatment
   • Treatment failure
4. Morbidity or mortality

   • Death
   • Sequelae
   • Asthma-related school absenteeism
   • Quality of life: CHSA, PAQLQ, pediatric caregiver AQLQ, PedsQL 3.0, Asthma Module PACD (daily diary; self-perception of health status—symptom-free days)
   • Duration of symptoms
   • Consolidation/pneumonia
   • Lung function test

ICU = intensive care unit; CHSA = Children’s Health Survey for Asthma; PAQLQ = Pediatric Asthma Quality of Life Questionnaire; AQLQ = Asthma Quality of Life Questionnaire; PedsQL = Paediatric Quality of Life; PACD = Pediatric Asthma Caregiver Diary.

Search strategy and study selection

An electronic search strategy was developed in collaboration with a librarian (GG) with expertise in knowledge synthesis. PubMed, EMBASE, BIOSIS, and the Cochrane Central Register of Controlled Trials (CENTRAL) were systematically searched without date or language restrictions. A primary search was done on December 11, 2015, with an update on October 19, 2016. Our search strategies are available in supplemental Table S1. Four domains were searched: asthma, child, respiratory pathogens, and outcomes. Citations of included articles were hand searched. EndNote (Version X7.5; Clarivate Analytics, Philadelphia, Pennsylvania) was used for the merging and de-duplicating of the articles. Distiller (Distiller SR v2; Evidence Partners, Ottawa, Ontario) was used for title and abstract screen, full text screen, and data extraction. Articles were restricted to English and French. Studies providing insufficient information for the evaluation of study quality (e.g., abstracts, poster presentations), studies discussing non-original data, and studies with inadequate exposure assessment and ineligible outcomes were excluded. Two reviewers (HK-M and JM) independently screened citations (titles and abstracts) identified through the search strategy. Potentially relevant articles were retrieved in full and assessed for eligibility for inclusion by the two independent reviewers. Disagreement was resolved by consensus.

Risk-of-bias assessment and data collection

We used the updated and validated Cochrane Collaboration’s Risk of Bias In Non-Randomized Studies – of Interventions tool (ROBINS–I) (30) to assess the risk of bias for each outcome in the studies. The various bias domains (bias due to confounding, patient selection, exposure classification, missing data, outcome measurement, and selective reporting of results) were explored per outcome in each study. We considered age a critical confounder for which adjustment was required. Each study was judged to be at low, moderate, serious, or critical risk of bias for each domain and, overall, for each specific outcome. Quality assessment was not a judgement of the overall methodological quality of the study, because in many cases our objectives differed from those of the initial study, but rather of how well the measurement of each exposure–outcome association reflected the association between the investigated respiratory pathogens and the chosen outcome. Data and quality assessments were extracted using a pre-piloted data extraction form by a single reviewer (JM), with quality control of all extracted data and quality assessment performed by a second reviewer (HK-M). Disagreement was resolved by consensus. Study authors were contacted by email when necessary to address issues of insufficient clarity or missing information.

Data analysis

Descriptive statistics were used for analysis. Summary tables present the characteristics of each study and outcomes for each exposure–outcome comparison, with more than one comparison per study. Outcomes are presented as means and proportions in groups exposed and unexposed to respiratory pathogens. We report p values when no other effect measures or confidence interval are reported, acknowledging their limited value in assessing the clinical significance of the findings, the association strength, or the presence of a causal effect (31). Conclusions were based only on associations of interest assessed at low or moderate risk of bias.

Only outcomes with non-critical overall risk of bias are included in the detailed description of the outcome domains. Given the heterogeneity of outcome domains and measures assessed, we could not carry out a meta-analysis. Data were analyzed using R (version 3.2.1; R Foundation for Statistical Computing, Vienna, Austria) and STATA 13 (StataCorp, College Station, Texas).

Results

Search results

A total of 14,424 records were identified on December 11, 2015, with an updated search on October 19, 2016, yielding an additional 906 records. A total of 10,329 unique records were screened for eligibility by title and abstract, with 232 assessed as full text. Finally, 28 studies were retained for inclusion (26,32–58). The flow diagram (Figure 1) details reasons for exclusion. Searches of included studies’ reference lists did not identify any additional records.

Figure 1:

Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram

Characteristics of the included studies

Of the 28 included studies, 20 were cohort studies; 3, cross-sectional studies; 4, case-control studies; and 1, an RCT. All studies were published in English; 4 were conducted in Australia; 2, China (Hong Kong); 3, France; 3, Japan; 2, Turkey; 6, the United States; and 1 each in Argentina, Canada, Colombia, Egypt, Finland, India, Portugal, and Taiwan. All studies, except for 1 published in 1970, were performed between 1999 and 2016 (38). Study characteristics are provided in Table 1.

Table 1:

Characteristics of included studies (N = 28)

Extracted information	n (%)
Study design
Cohort study	20 (71.4)
Case-control study	4 (14.3)
RCT	1 (3.6)
Cross-sectional study	3 (10.7)
Study season
Year-round	20 (71.4)
Fall and winter	4 (14.3)
Fall, winter, spring	1 (3.6)
Winter, early spring	1 (3.6)
Not reported	2 (7.1)
Study setting*
Emergency department	13 (46.4)
Outpatient clinic	6 (21.4)
Hospitalized patients	12 (42.9)
Sub-study of asthma study in community	1 (3.6)
Asthma definition used*
Institutional guidelines†	9 (32.1)
Physician diagnosis	11 (39.3)
Hospitalized for asthma	1 (3.6)
Pulmonary function	1 (3.6)
Clinical diagnosis	6 (21.4)
Asthma control medication use	4 (14.3)
Response to bronchodilator therapy	4 (14.3)
Recurrent wheezing symptoms (≥3 wheezing episodes including presenting episode)	11 (39.3)
Patient self-diagnosis of presence of past symptoms or diagnosis according to parents	1 (3.6)
Not defined	3 (10.7)
Asthma exacerbation definition*
Clinical symptoms	9 (32.1)
Institutional guidelines	8 (28.6)
Physician diagnosis	3 (10.7)
Asthma medication use	3 (10.7)
Health service utilization	1 (3.6)
Not defined	7 (25.0)
Participant sex (n = 24)
>50% male	21 (75.0)
>65% male‡	9 (32.1)
Not reported	4 (14.3)
Participant age, y, mean or median
<5	4 (14.3)
5–12	11 (39.3)
>12	0 (0)
Not reported	13 (46.4)
Exposure measurement*
PCR on respiratory specimen	21 (75.0)
Point-of-care test on respiratory specimen	2 (7.1)
Viral culture of respiratory specimen	6 (21.4)
DFA on respiratory specimen	5 (17.9)
IFA on respiratory specimen	5 (17.9)
Bacterial culture on respiratory specimen	3 (10.7)
Serology on blood or serum	8 (28.6)
Molecular typing	2 (7.1)
Industry sponsored
Yes	3 (10.7)
No	18 (64.3)
Not reported	7 (25.0)

^* Not mutually exclusive; studies could have more than one

† Global Initiative for Asthma guidelines, National Institutes of Health guidelines, British Thoracic Society guidelines, or American Thoracic Society guidelines

RCT = randomized controlled trial; PCR = polymerase chain reaction; DFA = direct fluorescent antibody; IFA = indirect immunofluorescence assay

‡ These 9 studies are included in the 21 with more than 50% males

For 15 of 28 included studies, we included only a portion of the original study population to focus on children with acute asthma and, when indicated, we excluded asthmatic patients without an acute exacerbation, non-asthmatic control groups, and adults. The total number of included participants was 4,224, with almost all studies (24/28) including, but not limited to, patients aged younger than 5 years. In studies reporting participants’ sex, the average proportion of males was 60.6%. Asthma diagnosis leading to participant inclusion was primarily (39.3%; 11/28) defined by physician diagnosis and secondarily (32%; 9/28) by using guidelines-defined criteria (Global Initiative for Asthma [GINA], National Institutes of Health [NIH], British Thoracic Society, and American Thoracic Society). The main criteria used to diagnose an asthma exacerbation were clinical symptoms (32%; 9/28) and the GINA guidelines (29%; 8/28).

Twenty-one (75.0%) studies used PCR to detect the presence of a respiratory pathogen in the upper respiratory tract; 24 investigated viral pathogens; 11, atypical bacteria; and 2, bacteria. Among viral pathogens, the most frequently investigated was respiratory syncytial virus (RSV) (23 studies). Fifteen studies investigated the presence of HRV, with subtyping being completed in 4 studies. Detailed characteristics for each study are presented in Table 2. Nineteen studies investigated the severity of asthma exacerbation; 19, the use of the health care system; 8, the response to treatment; and 16, the morbidity and mortality of children with and without any respiratory pathogen or a specific pathogen. In total, 112 exposure–outcome associations were investigated.

Table 2:

Detailed study description

Study and country

Study design

Study setting

Season and study duration

Study definition of asthma

Study definition of exacerbation

Sample size

Age of study population

Diagnostic method used for pathogen exposure diagnosis

Pathogens investigated

Proportion positive for pathogens, n (%)

Most frequently diagnosed pathogen, n (%)

Co-infection n (%)

Akturk et al (32), Turkey

Case-control

ED

Winter, early spring; 12/2013–4/2014

≥3 episodes of wheezing or dyspnea, clinically improved with beta2-agonist

Clinical symptoms

125

2.5–15 y; median 5.5 y

RT-PCR

HRV, RSV, hMPV, ADV, HBoV, M. pneumoniae, INF-A, INF-B, H1N1, PIV (types 1,2,3 and 4), CoV-NL63, CoV-229R, CoV-OC43, CoV-HKU1, EV, para-echovirus

106/125 (85%)

HRV 38/125 (30%)

17/125 (14%)

Amin et al (33), Egypt

Cohort

Hospitalized patients

Year-round; 1/2011–12/2011

Repeated episodes of wheezing, with bronchodilator response or on long-term controller therapy

Not defined

130

2–12 y; mean 2.95 (SD 2.46) y

RT-PCR

RSV, hMPV, ADV, INF-A, PIV

54/130 (42%)

RSV 28/130 (22%)

0/130 (0%)

Arden et al (34), Australia

Cohort

ED

Year-round; 3/2005–2/2007

>2 episodes of wheezing or dyspnea with doctor diagnosis of clinical response to beta-2 agonist

Asthma medication use

51 (N = 78)

Median 5.1 (IQR 2.5–7.3) y

RT-PCR with DNA sequencing analysis

HRV-A, HRV-B, HRV-C, RSV, hMPV, ADV, HBoV, INF-A, INF-B, EV, CoV, PIV, Echovirus, para-echovirus

51/78 (65%)

HRV 41/78 (53% )

20/78 (26%)

Awasthi et al (35), India

Cross-sectional

Hospitalized patients

Year-round; 8/2010–8/2011

GINA guidelines

GINA guidelines

44

1–12 y

Serology

C. pneumoniae

11/44 (25%)

C. pneumoniae 11/44 (25%)

NA

Bebear et al (36), France

Cohort

ED and hospitalized patients

Year-round; 3/2007–10/2010

GINA guidelines

GINA guidelines

168

Infected, 5.4 y (Q1, 3.3; Q3, 10.4) vs non-infected 3.3 y (Q1, 1.5 y; Q3, 5.8 y)

RT-PCR; bacterial culture; serology

HRV, RSV, hMPV, ADV, HBoV, INF-A, INF-B, EV, CoV, PIV, M. pneumoniae, C. pneumoniae

73/168 (44%)

NR

7/168 (4%)

Belessis et al (37), Australia

Case-control

ED and hospitalized patients

Year-round; 2/2000–6/2001

Diagnostic criteria based on ISAAC questionnaire and National Asthma Council Australia classification

Not defined

92

1–16 y

RT-PCR; viral culture; IFA ; serology

HRV, RSV, hMPV, ADV, M. pneumoniae, C. pneumoniae, INF-A, INF-B, PIV type 1, 2, 3

79/92 (86%)

HRV 67/92 (73%)

6/92 (7%)

Berkovich et al (38), United States

Cohort

Outpatient clinic

Fall, winter; 9/1967–2/1968

Not defined

Not defined

79 (108 asthma exacerbation episodes)

0.5–16 y

Viral culture; bacterial culture; serology

RSV, ADV, M. pneumoniae, INF-A, INF-B, PIV type 1, 2, 3, S. pneumoniae, beta-haemolytic streptococci

27/79 patients(34); 33/108 episodes (31)

INF-A 12/108 episodes (11%)

Not clear

Bizzintino et al (39), Australia

Cohort

ED

Year-round; 4/2003–2/2010

Doctor diagnosis of wheezing with increased difficulty breathing, based on American Thoracic Society definitions

American Thoracic Society guidelines

126

2–16 y; mean 6.4 (SD 3.3) y

RT-PCR ; DFA testing on fresh specimen; DNA sequencing analysis

HRV, RSV, hMPV, ADV, HBoV, INF-A, INF-B, PIV (types 1, 2, 3, 4), EV, CoV

118/128 (92%)

HRV 112/128 (88%) with HRV-C 76/128 (59%)

14/128 (11%)

Coleman et al (40), United States

Cohort

Not reported

Not reported

≥ 1 of (a) physician diagnosis of asthma, (b) use of albuterol for coughing or wheezing, (c) use of a daily controller medication, (d) step-up plan including use of albuterol or short-term use of inhaled corticosteroids during illness, (e) use of prednisone for asthma exacerbation

Asthma medication use

192 exacerbations in 102 patients

6–11 y

RT-PCR

HRV, RSV, hMPV, ADV, INF, EV, CoV, PIV

69% of 192 exacerbations

HRV 34%

NR

Ducharme et al (26), Canada

Cohort

ED

Year-round; 2/2011–12/2013

Physician diagnosis of asthma, based on previous wheezing episode with signs of airflow obstruction and response to bronchodilators, ≥3 asthma-like episodes (if <2 y old), or previous diagnostic lung function tests

Clinical symptoms

965 (N = 973)

1–17 y; 75% 1–5 y

RT-PCR

HRV-A, HRV-B, RSV, hMPV, ADV, INF, INF-A, INF-B, H1N1, H3N2, H5N1, PIV (types 1, 2, 3, 4), CoV, HKU1, CoV-NL63, CoV-229E, CoV-OC43, EV/HRV, EV A, B, C, D

579/933 (62%)

HRV-A, HRV-B, or RV/EV, 54%

166/933 (18%)

Duenas Meza et al (41), Colombia

Cross-sectional

ED

Year-round; 12/2010–3/2012

GINA guidelines; asthma diagnosis ≥6 mo; pediatrician or pulmonologist confirmed asthma

Clinical symptoms

169

2–15 y

RT-PCR; DNA sequencing analysis; molecular typing

HRV, HRV-A, HRV-B, HRV-C, RSV, ADV, M. pneumoniae, INF-A, INF-B, EV-D68, PIV types 1-3, picornavirus

141/169 (83%)

HRV 125/169 (74%); HRV-C 83 (49%)

21/169 dual M. pneumoniae and virus infection; 18/169 more than 1 viral pathogen

Hanhan et al (42), United States

Cohort

Hospitalized patients

Year-round; 9/1997–10/1998

Status asthmaticus definition: failure to respond to usual appropriate initial emergency room treatment necessitating PICU admission

Clinical symptoms and asthma medication use

35

1.5–19 y

Serology

M. pneumoniae

15/35 (42%)

NA

NA

Jartti et al (43), Finland

RCT

Hospitalized patients

Year-round; 9/2000–5/2002

NIH guidelines

Clinical symptoms

58

0.42–6.1 y; mean 2.6 (SD 1.3) y

RT-PCR; viral culture; DFA testing on fresh specimen; serology

HRV, RSV, hMPV, ADV, INF-A, INF-B, PIV (types 1– 4), EV, CoV

58,158 (100%; presence of pathogen was inclusion criterion)

EV 21/58 (36%)

17/58 (29%)

Joao Silva (44), Portugal

Cohort

ED

Year-round; 1/2003–12/2003

GINA guidelines

GINA guidelines

37

6–13 y; (mean 8.5 y)

RT-PCR; IFA

HRV, RSV, ADV, INF, M. pneumoniae, C. pneumoniae, EV, PIV

78%

HRV 70.3%

22%

Kantor et al (45), United States

Cohort

ED

Year-round; Summer 2011–Spring 2015

History of physician-diagnosed asthma

Clinical symptoms, management of exacerbation and treatment response

155 (N = 183)

6–17 y; mean 9.9 (SD 3.2) y

RT-PCR; molecular typing

HRV, RSV, hMPV, ADV, INF-B, RV/EV, PIV types 1–3, INF, INF-A types H1 and H3

123/183 (67%)

HRV/EV (61% of 155)

4/183 (2%)

Kato et al (46), Japan

Cohort

Hospitalized patients

Year-round; 11/2003–11/2006

History of ≥3 episodes recurrent wheezing and documented wheezing by auscultation; based on Japanese Society of Pediatric Allergy and Clinical Immunology guidelines

Doctor diagnosis

33

0.3–8.1 y

RT-PCR; antigen rapid testing; DNA sequencing analysis

HRV, RSV, EV-D68, EV, Coxsackie/echovirus, PIV

33/33 (100%; pathogen positive was inclusion criterion)

HRV 21/33 (64% )

7/58 (12%); analysis population NR

Leung et al (47), China (Hong Kong)

Cohort

Hospitalized patients and outpatient clinic

Year-round; 1/2007–2/2008

Hyperresponsiveness to methacholine or reversible airflow limitation or ≥3 episodes of cough, shortness of breath, and wheezing during previous 12 months, based on British Thoracic Society guidelines or criteria

GINA guidelines

209

3–18 y; mean 7.6 (SD 4.1) y

RT-PCR

HRV, RSV, hMPV, ADV, HBoV, INF-A (H1N1, H3N2, H5N1), INF-B, PIV (types 1–4), CoV-229E, CoV-OC43, SARS–CoV, EV, M. pneumoniae, C. pneumoniae

105/206 (51.0%)

54/206 (26.2%)

22/206 (10.7%)

Maffey et al (48), Argentina

Cohort

Hospitalized patients

Year-round; 1/2006–12/2006

GINA guidelines with history of ≥2 previous wheezing episodes diagnosed by a physician, presenting with a new episode requiring hospitalization

Not defined

137 (N = 209)

0.4–16 y

RT-PCR; IFA

HRV, RSV, hMPV, ADV, bocavirus, INF-A, INF-B, CoV, echovirus, ParaINF 1–3, M. pneumoniae, C. pneumoniae

162/209 (78%)

RSV 85/209 (40.7%)

47/209 (22.5%); 43/209 dual infection, 4/209 triple infection

Mak et al (49), China (Hong Kong)

Case-control

Hospitalized patients

Fall and winter; 10/2008–3/2009

Discharge diagnosis of asthma in hospital computerized database

GINA guidelines

126

mean 5.6 (SD 3.6) y

RT-PCR; DNA sequencing analysis

HRV-A, HRV-B, HRV-C

NR; only reported for HRV

HRV 107/126 (85%); HRV-C 88/126 (70%)

NR

Malka et al (50), United States

Cross-sectional

Outpatient clinic

Year-round; 6/2010–5/2011

Not defined

Not defined

66

7–18 y; mean 11 (SD 3.2) y

RT-PCR

RSV, MPV, ADV, HBoV, INF-A, INF-B, PIV (types 1–4), EV/RV, CoV, Coxsackie/echovirus

39/66 (59%)

HRV (62%)

11/39 (28%)

Mandelcwajg et al (51), France

Cohort

ED

Fall, winter; 11/2005–3/2009

Previous doctor diagnosis of asthma or history of ≥1 acute asthma episode

Not defined

339

1.5–8.92 y

RT-PCR; viral culture; DFA testing on fresh specimen

RSV, hMPV, ADV, HBoV, INF-A, INF-B, PIV types 1–3

123/339 (36.3%)

RSV 50/339 (14.8%)

NR

Nagayama et al (52), Japan

Cohort

Hospitalized patients

Not reported

≥2 episodes of wheezy distress

Clinical symptoms and health service use

212 (no. of exacerbation episodes)

0.5–14 y

Bacterial culture

H. influenzae, S. pneumoniae

43/212 (20.3%)

H. influenzae 17/212 episodes (8.0%)

5/212 (2.4%)

Ou et al (54), Taiwan

Cohort

ED and outpatient clinic

Year-round; 1/2000–12/2005

history of ≥1 clinically evident asthma attacks after age 2 y

Clinical symptoms and asthma medication use

316

2–14 y

Serology

M. pneumoniae

99/316 (31.3%)

NA

NA

Ozcan et al (53), Turkey

Cohort

Outpatient clinic

Year-round; 9/2009–9/2010

GINA guidelines

GINA guidelines

104

mean 8.78 (SD 3.4) y

RT-PCR

HRV, RSV, hMPV, ADV, INF-A, INF-B, PIV (types 1–4), CoV-OC49, CoV-NL63, CoV-229E

56/104 (54%)

HRV 37/104 (36%)

9/104 (9%)

Rueter et al (55), Australia

Cohort

ED

Year-round; 7/2002–9/2004

Physician diagnosis of asthma attack based on clinical symptoms on presentation

Doctor diagnosis

135 (N = 168)

2–16 y; mean 6.6 (SD 3.5) y

RT-PCR; viral culture; DFA testing on fresh specimen

HRV, RSV, hMPV, ADV, INF-A, INF-B, PIV

All investigated patients: 64% (107/168); NR for 135 included

NR

NR

Smith et al (56), United States

Case-control

ED

Year-round; 7/1996–5/1997

Medical history of at least mild intermittent asthma

Clinical symptoms

101

3–18 y

Viral culture; DFA fresh specimen; IFA

RSV, ADV, INF-A, INF-B, PIV 1–3

15/101 (15%)

RSV 7/101 (7%)

0 (0%)

Thumerelle et al (57), France

Cohort

Hospitalized patients

Fall, winter, spring; 10/1998–6/1999

≥3 recurrent episodes of reversible wheezing within 2 y preceding study

Guidelines

82

2.1–15.33 y; mean 7.75 (SD 4.8) y

RT-PCR; IFA; serology

HRV, RSV, ADV, INF-A, INF-B, CoV-229E, EV, PIV types 1–3, M. pneumoniae, C. pneumoniae

37/82 (45%)

EV 13/82 (16%)

6/82 (7%)

Zhao et al (58), Japan

Cohort

Outpatient clinic

Fall, winter; 10/1999–3/2000

NIH guidelines

Not defined

28 (N = 64)

0.33–15 y; mean 4.14 (SD 3.55) y

Antigen rapid testing

RSV, ADV, INF-A

28/64 (44%)

RSV 17/64 (27%)

0/64 (0%)

ED = emergency department; RT-PCR = real-time polymerase chain reaction; HRV = human rhinovirus; RSV = respiratory syncytial virus; hMPV = human metapneumovirus; HBoV = bocavirus; ADV = adenovirus; M. pneumoniae = Mycoplasma pneumoniae; INF-A = influenza A virus; INF-B = influenza B virus; PIV = parainfluenza virus; CoV = coronavirus; EV = enterovirus; IQR = inter-quartile range; HRV-A = human rhinovirus A; HRV-B = human rhinovirus B; HRV-C = human rhinovirus C; EV = enterovirus; GINA = Global Initiative for Asthma; C. pneumoniae = Chlamydia pneumoniae; N/A = not applicable; Q1 = first quartile; Q3 = third quartile; S. pneumoniae = Streptococcus pneumoniae; NR = not reported; ISAAC = International Study of Asthma and Allergies in Childhood; IFA = indirect immunofluorescence assay; DFA = direct fluorescent antibody; INF = influenza virus; SARS = severe acute respiratory system; H. influenza = Haemophilus influenza

Quality and risk-of-bias assessment of included studies

Figure 2 summarizes the risk of judgement bias per outcome domain as extracted from 112 comparisons. The quality assessment is presented per outcome domain and per bias domain, with most single studies contributing to more than one outcome within and between outcome domains. In supplemental Table S2, the detailed risk-of-bias assessment is presented for each exposure–outcome assessed. None of the outcome assessments had a low overall risk of bias; 7 had a moderate risk of bias, 85 had a serious risk of bias, and 20 had a critical risk of bias. Lack of investigation of confounding, selection bias, and multiple testing with reporting bias were the major problems with the quality of the investigated exposure–outcome associations.

Figure 2:

ROBINS-I risk-of-bias assessment

ROBINS-I risk-of-bias assessment is presented per outcome domain: severity of asthma exacerbation, health care use, response to treatment, and morbidity or mortality. The y-axis shows the proportion of exposure–outcome associations with low, moderate, high, or serious risk of bias within the different risk of bias domains.

ROBINS-I = Risk of Bias In Non-Randomized Studies – of Interventions

Exposure–outcome results

Of the 112 exposure–outcome comparisons, the primary exposure was the presence of any pathogen in 45 (40%) comparisons, any type of HRV in 34 (30%), atypical bacteria in 21 (19%), and specific viruses in 11 (10%); only 1 looked at the presence of any bacteria. Detailed results are organized per outcome domain in supplemental Table S3.

In 23% (26/112) of associations, the outcome was exacerbation severity. In 34% (38/112) of associations, the relationship between exposure and health care use was investigated, whereas in 26% (29/112), the clinically relevant outcome was morbidity or mortality. Seventeen percent (19/112) had response to treatment as the investigated outcome. The variety of exposures and outcomes precluded data aggregation. Very few outcomes were assessed using methods other than inference testing, and the presence of an association was often only suggested by the presence of a so-called statistically significant p value (p < .05). In the next section, we describe the data derived from the 7 associations considered at moderate risk of bias (Table 3) and pertaining to 112 comparisons, with added evidence from serious non-critical risk of bias when appropriate.

Table 3:

Associations with moderate risk of bias

Study	No. of patients	Exposure	Comparator	Outcome	Effect measure	Statistical method	Outcome numbers	Crude EM	Adjustment and interaction	Overall risk of bias
Bizzintino et al (39)	110 (76 + 34); subpopulation	HRV-C positive	HRV-A or HRV-B positive	Acute Asthma Severity Score (modified NIH score); on presentation to the hospital	Presentation of means, confidence intervals	Inference testing: t-test, multivariate linear regression	HRV-C exposed 10.4 vs HRV-A/HRV-B exposed 9.5	HRV-C positive 95% CI 10.0 to 10.9 vs HRV-A/HRV-B positive 95% CI 8.7 to 10.3 (p = .028)	Adjusted for age and sex; linear regression coefficient not reported (p = .018); no interaction terms	Moderate
Bizzintino et al (39)	116 (76 + 40); subpopulation	HRV-C positive	HRV-C negative and other viral pathogen positive	Acute Asthma Severity Score (modified NIH score); on presentation	Presentation of means, confidence intervals	Inference testing: t-test, multivariate linear regression	HRV-C exposed 10.4 vs viral pathogen HRV-C non-exposed 9.4	HRV-C exposed 95% CI 10.0 to 10.9 vs viral pathogen non–HRV-C exposed 95% CI 8.7 to 10.1 (p = .013)	Adjusted for age and sex; linear regression coefficient not reported (p = .009); no interaction terms	Moderate
Bizzintino et al (39)	126; subpopulation	HRV-C positive	HRV-C negative	Acute Asthma Severity Score (modified NIH score); on presentation	Presentation of means, confidence intervals	Inference testing: t-test, multivariate linear regression	HRV-C exposed 10.4 vs HRV-C non-exposed 9.4	HRV-C exposed 95% CI 10.0 to 10.9 vs HRV-C non-exposed 95% CI 8.7 to 10.2 (p = .015)	Adjusted for age and sex; linear regression coefficient not reported (p = .016); no interaction terms used	Moderate
Ducharme et al (26)	965; entire population	Pathogen positive	Pathogen negative	Duration of symptoms (in ED; proportion of children with PRAM severity score ≥ 4, at disposition or 4 hr after oral corticosteroids)	Presentation of proportions, OR	Multivariate logistic regression	Exposed 119 (20.7%) vs non-exposed 55 (15.5%)	—	Adjustment for baseline PRAM, ipratropium bromide received, Caucasian ethnicity, and viral detection; OR = 1.53, 95% CI 1.05 to 2.22	Moderate
Ducharme et al (26)	965; entire population	Pathogen positive	Pathogen negative	Duration of symptoms (in ED): Time to PRAM ≤3 = consideration for discharge	HR	Cox proportional hazards model	—	—	Adjustment for baseline PRAM, perfect adherence to oral corticosteroids, viral detection, and sites; HR = 0.86, 95% CI 0.74% to 0.99%	Moderate
Ducharme et al (26)	965; entire population	Pathogen positive	Pathogen negative	Length of active treatment inhalation albuterol (in ED) (in hr)	Presentation of means; regression coefficients	Linear regression model	Exposed, mean 4.00 (SD 4.16) h vs non-exposed, mean 3.32 (SD 2.48) h	—	Adjustment for baseline PRAM, delay between triage and oral corticosteroids, fever, no. of albuterol doses received in the first hour, viral detection, and sites; coefficient = .10, 95% CI 0.03% to 0.17%	Moderate
Ducharme et al (26)	965; entire population	Pathogen positive	Pathogen negative	Failure of ED treatment management (composite outcome)	Presentation of proportions, OR	Logistic regression model	Exposed, 110/579 (19%) vs non-exposed, 46/354 (13%)	—	Adjusted for age, sex, baseline PRAM, viral detection (yes or no), salivary cotinine (values of <1 ng/mL, 1–<4 ng/mL, and ≥4 ng/mL), and oral corticosteroid dose (mg/kg): OR = 1.61, 95% CI 1.06% to 2.45%; adjusted for symptoms between exacerbations, fever, baseline PRAM, oxygen saturation, viral trigger, and sites: OR = 1.57 , 95% CI 1.04% to 2.37%, p = .0312	Moderate

EM = effect measure; HRV-C = human rhinovirus C; HRV-A = human rhinovirus A; HRV-AB = human rhinovirus B; NIH = National Institutes of Health; ED = emergency department; PRAM = Pediatric Respiratory Assessment Measure; HR = hazard ratio

Association between presence of respiratory pathogens and exacerbation severity

When stratified by exposure, five comparisons were in favour of an association between HRV and increased severity of the exacerbation (34,39,45), and two were not (44,46), based on statistically significant differences (including p < .05). In the only published article at low to moderate risk of bias for this outcome, patients with HRV-C were associated with a higher severity score than patients without HRV-C on presentation to the emergency department (ED); the mean score for exposed patients was 10.4 (95% CI 10.0 to 10.9) compared with 9.4 (95% CI 8.7 to 10.2) for non-exposed patients (linear regression coefficient adjusted for age and sex, p = .016) (39). This study used a modified NIH severity score (range = 0–15), with a score between 8 and 11 defined as a moderate exacerbation and a score between 12 and 15 defined as a severe exacerbation (59). This positive association also remained when participants positive for HRV-C were compared with those infected with other viral pathogens (score = 10.4; 95% CI 8.7 to 10.1) and with HRV-A and HRV-B subtypes (score = 9.4; 95% CI 8.7 to 10.2) (39).

When comparing asthma exacerbation severity at presentation among patients harbouring any pathogen versus pathogen-negative patients, only one (32) of five comparisons (40,46,53,57) favoured a positive association. The presence of atypical bacteria was not associated with higher severity scores (35,41). The evidence for an association between any pathogen and hypoxia or the need for or duration of oxygen therapy is insufficient or of low quality (36,42,45,47,57).

Association between the presence of respiratory pathogens and health care utilization

No comparison with a low or moderate risk of bias was available for health care use, such that we cannot make any sound conclusion regarding the association between pathogens and health care use. Six of 30 comparisons at serious risk of bias investigated the association of respiratory pathogens and the need for hospital admission (26,32,41,44,53); another 6 comparisons investigated the presence of HRV or of any pathogen and its association with length of hospitalization (43,45–47,57). Four studies explored the effect of atypical bacteria on the length of hospitalization (35,36,41,42). Only low-quality evidence exists for the association of intensive care admission, length of stay, or mechanical ventilation with RSV or HRV (45,56), but not with pathogen exposure in general (32) or with atypical bacterial infection (42,60). Four studies investigated clinical deterioration leading to a return visit or readmission (26,43,54,60).

Association between the presence of respiratory pathogens and treatment effect

Only one study with a moderate risk of bias was identified with regard to treatment response. This study was conducted with children with a moderate or severe asthma exacerbation who were treated according to evidence-based guidelines in the ED (26). The authors observed, as their primary outcome, an association between the presence of any respiratory pathogen and an increased risk of ED treatment failure (odds ratio [OR] = 1.57; 95% CI 1.04 to 2.37), after adjusting for fever, baseline Pediatric Respiratory Assessment Measure clinical score (61), oxygen saturation, study site, and symptoms between exacerbations. No increased odds of treatment failure were observed when comparing HRV-positive and HRV-negative patients in this study (OR = 0.96; 95% CI 0.67 to 1.39). The latter comparison was, however, assessed as having a serious risk of bias. As secondary outcomes, the authors reported an association between the presence of any pathogen and prolonged duration of symptoms after discharge from the ED or hospital despite oral corticosteroid treatment. Pathogen presence was also associated with increased length of active treatment in the ED and length of rescue β2-agonist use over the next 10 days. Other evidence for an association between the presence of a pathogen and treatment response to bronchodilators (45), bronchodilator use (42,55), need of steroid treatment (53,60), and duration of steroid treatment (46) was assessed as being at high risk of bias.

Association between the presence of respiratory pathogens and morbidity and mortality

High-quality data are lacking on the association between pathogens and other indicators of morbidity. This includes the absence of high-quality evidence for an association between lung consolidation or infiltrate and the presence of any pathogen (32,57,60) or of atypical bacteria (35,36,42,54) or bacteria (52) at time of exacerbation. Duration of symptoms (32,43,47,53,57,62), symptom scores (34), respiratory function evaluation (45,47), and quality of life (26,34) were investigated, but comparisons were at a serious risk of bias, leading to inconclusive results.

Discussion

The large quantity of different exposures and comparisons, the heterogeneity of outcomes, and a high or critical risk of bias in most exposure–outcome comparisons in the 28 identified studies prevent firm conclusions regarding the potential impact and magnitude of effect of respiratory pathogens on asthma exacerbation outcomes. Our systematic review concludes that there is a possible association between HRV—more specifically, HRV-C—and more severe asthma exacerbation at ED presentation. It also established that treatment response to maximized bronchodilators and systemic corticosteroid in the ED was poorer in the presence of any respiratory pathogen, but not specifically HRV, among children with a moderate to severe exacerbation. However, the impact of specific respiratory pathogens or of pathogens in general on health care use and other indicators of morbidity is not clear. Our conclusions are based on only two studies that provided an outcome assessment with a low or moderate overall risk of bias (26,39).

Specific respiratory pathogens, more specifically HRV and RSV, are suspected to influence clinical presentation of acute asthma exacerbations (5,63). However, our review had an insufficient number of adequately powered high-quality studies to provide conclusive evidence regarding pathogen-specific effects, with the exception of a possible effect of HRV on exacerbation severity. HRV and its subtypes are highly prevalent in pediatric asthma exacerbations (64), with a higher prevalence among asthmatic children presenting with an exacerbation than among those without exacerbation (65). Lower respiratory tract symptoms persist for a longer duration and are more severe in asthmatic than non-asthmatic children (10,66).

Bizzintino et al described that children infected with HRV and particularly HRV-C had a higher asthma severity (by 1 point on the average clinical severity score) at presentation (39). Perhaps a defect in the HRV-induced innate immune response may be of importance, leading to more severe infection, particularly with the HRV-C subtype. Indeed, interferon type I and III responses are modified in asthmatic children infected with HRV (67,68). In addition, evidence shows that HRV leads to an aberrant adaptive immune response through rapid recruitment of circulating CD4 and CD8 (69). Among people with asthma, a more pronounced expression of cytokines associated with Th2-lymphocytes is postulated. Accordingly, the observed positive association between HVR and asthma severity might be explained by the impaired immune response to respiratory pathogens, and more specifically to HRV, which drives exacerbation development and presentation (70).

Extensive literature has emerged investigating the interaction between respiratory viral infection and allergic sensitization. Clinical studies reported a positive interaction between atopy and viral pathogens in the onset of asthma exacerbations (71,72). Whether this interaction leads to clinically important differences between infected and non-infected children once an asthma exacerbation is established remains unclear. The only study investigating the interaction between atopy and viral pathogens included in our review reported an increased length of active treatment among HRV-infected children (45). The risk of bias in this study was, however, assessed as serious.

Although several reports have suggested a role of HRV co-infections and infection with multiple pathogens in the severity of exacerbation (73,74), studies included in our review did not provide sufficient data to come to a conclusion regarding the impact of co-infections on clinical outcomes. After HRV, RSV was the exposure most sought for and the second most frequent pathogen, but its specific role with regard to exacerbation severity and clinical exacerbation evolution has not been extensively investigated.

In their large prospective cohort, Ducharme et al showed that the presence of a respiratory pathogen was associated with increased ED treatment failure (prolonged ED stay, hospitalization, and relapse), which suggests an infection-mediated altered response to maximized bronchodilator and oral corticosteroid therapy, which are the cornerstone of ED treatment of moderate to severe exacerbations among children (8,26,75). The decreased response was not attributable to HRV. However, with only one published study, more data are needed to further explore the clinical importance of specific respiratory pathogens associated with exacerbations and their treatment response.

This review has several limitations. First, we could only comment on associations with, and not on any causal role of respiratory pathogens on, clinical outcomes of interest. PCR is the current gold standard in the diagnosis of respiratory pathogens from the upper respiratory tract, replacing viral and bacterial cultures. The use of PCR has led to wider diagnosis of the presence of (multiple) respiratory pathogens (9,76). However, PCR cannot differentiate between dead pathogens and actively replicating infectious pathogens, nor can it determine whether the exacerbation was indeed triggered by the pathogen. Kling et al reported that more than 40% of children with a PCR-positive specimen for HRV during an asthma exacerbation still had HRV detectable by PCR 6 weeks after the exacerbation, suggesting the possibility of prolonged shedding (77). Alternatively, Engelmann et al proposed that re-infection, rather than pathogen persistence, may better explain this finding (78).

In clinical practice, there is no way to differentiate colonization from infection. In all included studies, it is likely that some patients colonized with a respiratory pathogen were misclassified as patients with an exacerbation triggered by the identified pathogen. In practice, less sensitive ELISA tests and rapid tests have also been widely used to diagnose respiratory pathogens, and this review included studies that used diagnostic methods other than PCR to define the presence of a pathogen. Second, most studies were too small and lacked sufficient power to make firm conclusions regarding the potential effect of less prevalent pathogens on various exacerbation outcomes. Third, given the differences in outcomes and effect measures reported in the different studies, it was not possible to investigate a potential publication bias in any formal way.

Fourth, the poor methodological quality of most included studies (85/112 and 20/112 comparisons at serious and critical risk of bias, respectively) prohibits firm conclusions. The lack of formal statistical investigation and adjustment for confounding variables were the main methodological issues. The risk of selection bias is important when included children were all hospitalized or presenting to the ED. Selective reporting of the numerous possible exposure comparators and outcome associations was also identified as an important source of bias. Non-significant results were often either not reported or reported but without their estimates of effect.

To improve the quality of studies investigating causal and mechanistic associations between specific respiratory pathogen exposure and clinical important outcomes, we suggest that researchers

1. use validated instruments for asthma and asthma exacerbation diagnosis and use core health outcomes with emphasis on patient-centred outcomes as proposed by consensus groups (28,29);

2. be cognizant of selection bias during study design and analysis (79,80);

3. rigorously report all the investigated outcomes, including statistically non-significant results;

4. investigate, report, and, where necessary, adjust for known confounders (81);

5. investigate the interaction between pathogens and between pathogens and atopy–allergy (82);

6. use formal statistical investigation and methods, more and above solely the calculation of p values, with reporting of absolute effect measures (83); and

7. consciously design studies with a large enough sample size, taking into account seasonality and time-varying changes in respiratory virus presence, to investigate the species specific role of respiratory pathogens.

Following these suggestions could lead to firmer conclusions regarding the impact of specific pathogens and co-infection on outcomes of interest. Reporting study results using the CONSORT and STROBE guidelines would also improve the assessment of study quality and facilitate their inclusion in systematic reviews and in meta-analyses (84,85).

Finally, we acknowledge that our systematic search and rigorous data extraction and analysis dates from end of 2016. On July 7, 2018, we performed a limited updated search in PubMed only. No other systematic review or meta-analysis had been published since our last search. This more recent search picked up our peer-reviewed secondary analysis and sub-study of Ducharme et al (26,86).

We investigated the association between the presence of a laboratory-confirmed specific respiratory pathogen and both risk of ED treatment failure and baseline exacerbation severity among children presenting to the ED with a moderate or severe asthma exacerbation. We found that although the presence of rhinovirus and its species was not associated with treatment failure, there was a higher failure risk in children with a non-rhinovirus infection (absolute risk = 25.4%; 95% CI 19.8% to 31.0%) compared with no pathogen, resulting in an adjusted risk difference of 13.1% (95% CI 6.4% to 19.8%). More specifically, RSV, influenza, and parainfluenza virus were associated with a higher risk of treatment failure of 21.4% (95% CI 14.1% to 28.7%), 37.5% (95% CI 17.8% to 57.2%), and 46.7% (95% CI 20.4% to 73.0%), respectively. Human metapneumovirus had an RD of 8.0% (95% CI 1.6% to 17.6%). Coronavirus, adenovirus, and enterovirus D68 and the presence of a co-infection were not associated with an increased risk of failure. Presence of a pathogen was not associated with increased exacerbation severity on presentation.

The implication of these study results for further treatment development or for application of current available strategies needs to be further assessed. Therapies with inhaled anticholinergics or magnesium sulfate, currently reserved for severe exacerbations, may be tested for their efficacy in exacerbations of any severity triggered by viruses that have been associated with poor treatment response. In addition, the risks, benefits, and safety of an alternative pathogen non-specific therapy targeting anti-neutrophilic inflammation, such as azithromycin, could be explored. Any pathogen-specific anti-viral or treatment intensification first requires identification of pathogens associated with each exacerbation.

Conclusions

In summary, the evidence suggests a potential association between HRV and higher severity of asthma exacerbations in children presenting to the ED. In children with a moderate or severe asthma exacerbation, the presence of one or more respiratory pathogens is associated with a lower treatment response. Although these findings deserve replication, further investigation of the effect of specific pathogens seems appropriate, mainly in light of growing evidence of the importance of specific pathogens such as HRV and HRV-subtype infections. Increased knowledge of host–pathogen and pathogen–pathogen interactions and their role during an asthma exacerbation may enable us to better identify and perhaps manage children with acute asthma at higher risk of treatment failure. Identified key pathogens could thus be targeted by specific preventive approaches and rapidly identified with point-of-care diagnosis.

Competing Interests:

Dr Ducharme reports receiving unrestricted donations from Novartis, Astra-Zeneca, Merck Canada, and Trudell Medical outside the submitted work. Dr. Merckx reports current employment with bioMérieux Canada Inc.; the article originated before this employment.

Ethics Approval:

N/A

Informed Consent:

N/A

Registry and the Registration No. of the Study/Trial:

N/A

Animal Studies:

N/A

Funding:

No external funding was received for this article.

Peer Review:

This article has been peer reviewed.