Detection of mostly viral pathogens and high proportion of antibiotic treatment initiation in hospitalised children with community-acquired pneumonia in Switzerland – baseline findings from the first two years of the KIDS-STEP trial

DOI: https://doi.org/10.57187/smw.2023.40040

Introduction

In Europe, community-acquired pneumonia accounts for 10 to 15% of paediatric hospital admissions [1]. Studies conducted in different settings since the introduction of conjugate vaccines for Streptococcus pneumoniae and Haemophilus influenzae type b consistently found the majority of admissions due to acute respiratory infections and more specifically for community-acquired pneumonia to be caused by respiratory viruses, most prominently respiratory syncytial virus (RSV) and influenza viruses [2–4]. Detection of individual respiratory viruses is not associated with increased severity or extended length of hospital stay [5], although some studies have shown increased length of stay for combinations of RSV and influenza or rhinoviruses [6, 7]. Current clinical guidelines for children hospitalised with CAP uniformly recommend antibiotic treatment in the presence of World Health Organization (WHO) danger signs but are incongruent regarding antibiotics in other children.

The KIDS-STEP trial enrols children at eight sites providing medical care to a large proportion of Switzerland’s paediatric population [8]. The ancillary microbiology study has the aim to provide data for pathogen subgroup analyses of the effect of betamethasone for community-acquired pneumonia and to differentiate pathogen-driven effects on length of hospital stay from medication effects. The COVID-19 pandemic has disrupted the seasonality of acute respiratory infections in Switzerland and may result in changed patterns of aetiology of community-acquired pneumonia in the coming years [9]. We therefore conducted a preliminary pathogen analysis for the trial participants enrolled up to autumn 2020 to provide representative, high-quality and mostly prepandemic data on pathogen detection in children in Switzerland hospitalised for community-acquired pneumonia. The aim of this study was to describe the pathogens detected and their association with clinical findings at baseline in Switzerland.

Methods

The KIDS-STEP Trial is a randomised controlled superiority trial on the effect of betamethasone on clinical stabilisation of children hospitalised with community-acquired pneumonia. The full trial protocol has been published [8]. In brief, children are screened following a clinical diagnosis or differential diagnosis of community-acquired pneumonia and the decision to admit as an inpatient and are eligible if they are between 6 months and 14 years of age and fulfil a clinical case definition of community-acquired pneumonia at eight paediatric emergency departments across Switzerland. The clinical case definition is a temperature ≥38°C and at least two from a list of signs and symptoms of a lower respiratory tract infection. Detailed eligibility criteria are presented in table 1. Co-primary outcomes of the main trial are (1) clinical stabilisation defined as normalisation of initially deranged vital signs or discharge from hospital and (2) re-admission to hospital within 4 weeks from randomisation. Enrolment into the main trial has been delayed owing to the COVID-19 pandemic and is currently ongoing [9].

All parents or legal guardians gave written informed consent including the ancillary study presented here. The trial and ancillary study were approved by the local ethics committee of the trial centre (Ethikkommission Nordwest- und Zentralschweiz (EKNZ), study no. 2018-00563), other local ethics committees in Switzerland for participating sites and the regulatory authority Swissmedic (2018 DR 3070). The trial is registered on https://clinicaltrials.gov (NCT03474991) and on the Swiss National Clinical Trials Portal (SNCTP000002864).

For this interim baseline analysis, we included participants enrolled over the first two years since opening of the trial, i.e., from September 2018 until calendar week 36 in 2020. Clinical data at presentation, management during the first 24 hours after admission and pathogen detection results from standard of care samples taken on the day of admission were collected prospectively for the main trial and were extracted from the trial database held at the University of Basel’s clinical trials unit. Positive pathogen detection results from standard of care testing were collected through the trial’s case report forms with specific fields for RSV, influenza viruses, S. pneumoniae, H. influenzae and Mycoplasma pneumoniae and free text for other respiratory pathogens. SARS-CoV-2 test results were collected in amended case report forms from 19 May 2020 onwards. Standard of care samples were obtained according to routine procedures at the trial sites, local materials were used and the analyses performed in local laboratories.

Randomised treatment allocation to betamethasone or placebo remained concealed. The interim analysis was exploratory and sample size and endpoints were not pre-specified. For the current report, use and release of clinical follow-up data from the main trial needed to be limited in order to prevent jeopardising the integrity of the trial.

As an ancillary study offered to all trial participants, nasopharyngeal swabs are taken on the day of admission. Nasopharyngeal swabs were collected on the day of randomisation using swabs (FLOQSwab^®) and transferred to 3 ml universal transport medium (UTM^®) by Copan (Brescia, Italy). The suspended swabs were processed within 48 hours. After vortexing for 30 seconds, UTM was transferred to sterile storage tubes in aliquots of 900 µl. Aliquots were stored at −80°C until analysis. One aliquot per patient was used for pathogen detection with Filmarray BIOFIRE^® Respiratory Panel 2.1 plus (bioMérieux, Marcy-l'Étoile, France). The panel detects the following targets: adenovirus, coronavirus 229E, coronavirus HKU1, coronavirus OC43, coronavirus NL63, Middle East respiratory syndrome coronavirus (Mers-CoV), human metapneumovirus, human rhinovirus/enterovirus, influenza A, influenza A/H1, influenza A/H1-2009, influenza A/H3, influenza B, parainfluenza 1, parainfluenza 2, parainfluenza 3, parainfluenza 4, RSV, Bordetella pertussis, Bordetella parapertussis, Chlamydophila pneumoniae and M. pneumoniae. The analyses were done before addition of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) to an updated version of the panel.

Data management and statistical operations were performed in Stata 15 (College Station, Texas). C-reactive protein values were grouped into lower than 80 mg/l and equal to or higher than 80 mg/l [12]. For comparisons between groups Wilcoxon rank-sum test or Kruskal-Wallis tests were used for continuous variables and chi-square tests for categorical variables. The alluvial plot for figure 1 was drawn using the online tool RawGraphs 2.0 beta.

Ethics

The trial and ancillary study were approved by the local ethics committee of the trial centre (Ethikkommission Nordwest- und Zentralschweiz (EKNZ), study no. 2018-00563), other local ethics committees in Switzerland for participating sites and the regulatory authority Swissmedic (2018 DR 3070).

Results

The first 138 children enrolled in the trial were included in the current analysis. The majority of participants were enrolled in autumn and winter 2019/20 (detailed in supplementary figure S1 in the appendix). Thirty (21.7%) were enrolled after detection of the first case of COVID-19 in Switzerland, mostly in March 2020 and only 6 were enrolled between April and September 2020.

Clinical characteristics of the participants at trial entry are presented in table 2. The median age was 3 years with an even distribution between sexes. A high proportion of participants had signs of more severe disease, including 65.2% with chest retractions and 31.2% with an oxygen saturation below 92%.

Forty-three participants (29.0%) were already on antibiotic treatment prior to admission and 104 participants (75.4%) received antibiotic treatment on admission. Figure 1 shows antibiotic treatment of trial participants before and after admission. The most common antibiotics used were aminopenicillins (58.7% of those receiving antibiotic treatment after admission). When antibiotics were used, 85.6% were Access group antibiotics according to the WHO’s Essential Medicines List AWaRe classification [13]. Six of the 27 participants (22.2%) initially treated with an aminopenicillin plus a beta-lactamase inhibitor were switched to an aminopenicillin only within the first 24 hours after admission.

Within the first 24 hours, 91 participants (65.9%) received supplemental oxygen and four (2.9%) were placed on some form of respiratory support.

Figure 1 Alluvial plot of antibiotic treatment prior to admission and on admission. No label: unknown; AMC: amoxicillin + clavulanic acid; AMX: amoxicillin; BL + M: beta-lactam + macrolide; CFX: cefuroxime; CLA: clarithromycin; CRO: ceftriaxone; DOX: doxycycline

Respiratory specimens for pathogen testing were obtained from 132 of 138 participants (95.7%). These were either part of the standard of care or nasopharyngeal swabs obtained as study samples. Eighty-four participants (60.9%) had both kinds of sample taken, 25 (18.1%) only standard of care samples, 23 (16.7%) only study samples, and 6 (4.3%) had no sample taken.

Standard of care samples were obtained from 109 participants (79.0%). Except for one tracheal aspirate, all other respiratory standard of care samples were nasopharyngeal swabs or nasopharyngeal aspirates. Seventy-two (66.1%) were tested for viruses and bacteria, 16 (14.7%) for viruses and 21 (19.3%) for bacteria only. Testing for bacteria was most commonly done by bacterial culture. The respiratory pathogens detected on standard of care samples are listed in table 3. No positive SARS-CoV-2 tests were reported. Blood cultures were obtained in 86 patients but yielded no positive results.

Study nasopharyngeal swabs at trial entry were obtained from 107 participants (77.5%). Table 3 shows the respiratory pathogens detected on nasopharyngeal swabs in these participants. The most frequently detected pathogen was human rhinovirus/enterovirus, followed by RSV and hMPV.

Twenty-one of the 24 participants where multiple pathogens were detected in study samples carried multiple (up to four) different viruses. In the other three participants, M. pneumoniae was found in combination with a virus (one adenovirus, two rhinovirus).

When information from standard of care and study samples was combined, 5 (3.8%) participants had co-detection of bacteria and viruses. Ninety-seven participants (73.5%) had at least one respiratory virus detected in any sample and 92 (69.7%) had one or more viruses but no bacteria detected. All co-detections of different pathogens including study and standard of care samples are detailed in supplementary table S1 in the appendix.

Detection of respiratory pathogens showed age- and season-related patterns (table 4). RSV was more commonly found in younger children and during the peak acute respiratory infection season from October to March. Influenza virus was equally more common during the acute respiratory infection season, and M. pneumoniae was more commonly detected in older children.

Patchy changes on chest X-ray were found in a higher proportion of children than lobar consolidations with any of the listed pathogens detected, but this finding was not supported by statistical evidence.

Discussion

In line with international data, RSV is the most commonly detected pathogen that has been found to be strongly associated with hospital admission [2, 4]. Compared with other European settings, hMPV was detected more frequently in the Swiss population. This may either reflect differences in local aetiology or the winter season 2019/20 may have seen uncommonly large numbers of hMPV infections. Both RSV and influenza showed the expected seasonality. Since very few participants were enrolled after the onset of the COVID-19 pandemic, we were unable to capture possible changes in seasonality of pathogens. As expected, M. pneumoniae was more commonly detected in older children [14]. Human rhinovirus was overall the most frequently detected pathogen, but previous studies have shown that detection of human rhinovirus in upper respiratory tract samples is only very weakly associated with pneumonia or hospital admission due to acute respiratory infection [2, 4].

Of the children admitted with community-acquired pneumonia, 75.4% received antibiotic treatment. Antibiotics were selected mostly in accordance with international treatment guidelines and from the AWaRe classification’s Access group [11, 13]. Assessed against the pathogen testing results, it is likely that a large proportion of these antibiotic prescriptions were unnecessary. German guidelines advise withholding antibiotics in children without WHO danger signs, whereas current UK guidelines advise treating and reviewing in due course, especially after receiving pathogen testing results. The proportion of children receiving antibiotics on admission in our study is broadly comparable to recent studies from the USA and Europe [4, 15, 16]. Future analyses of the trial will show if antibiotics were stopped during the participants’ hospitalisation. Earlier availability of results from rapid syndromic pathogen testing may prevent antibiotic prescriptions in children with acute respiratory infections including pneumonia, but results from single-centre or retrospective studies have so far been disappointing [17, 18]. Judicious prescribing can be aided by the implementation of Antimicrobial Stewardship programmes, which have so far not been extended to Swiss paediatric emergency departments [19].

Overall, 81.2% of patients had a chest X-ray. This is surprising because German language guidelines that were written with participation of Swiss members do not routinely recommend a chest X-ray in children without WHO danger signs [11]. The PERCH study demonstrated that abnormal chest X-ray findings were associated with severe or very severe pneumonia and slightly longer duration of symptoms [20]. However, 46% of children with clinically severe or very severe pneumonia had a normal chest X-ray, demonstrating that radiology alone is insufficient to identify children with more severe disease [20]. In a high-resource setting, the decision on antibiotic treatment taken before chest X-ray was altered in less than half of cases where radiological findings were discordant (i.e., not suggestive of pneumonia in children with a pre-X-ray plan for antibiotics and vice versa), indicating that care providers are aware of the limited predictive value of chest X-ray [21]. Pleura and lung sonography have demonstrated a high negative predictive value for pathological findings on chest X-ray in paediatric community-acquired pneumonia and can help to avoid unnecessary irradiation in settings where sufficient resources for ultrasound are available [22].

The presented study has some important limitations. Pathogen testing both on study and routine samples was mostly limited to upper respiratory tract samples. A strong association of detection of pathogens in upper respiratory tract samples with hospitalisation for acute respiratory infection is only seen with some respiratory viruses (RSV, influenza virus, hMPV and parainfluenza viruses) but not in bacteria or human rhinovirus/enterovirus [4]. Lower respiratory tract samples may be better suited for detection of aetiologically relevant bacteria [23, 24]. Additionally, for standard of care samples only positive findings were collected. However, in paediatric clinical routine, sampling is most commonly limited to upper respiratory tract samples [25]. Because more than a quarter of children had already received antibiotic treatment prior to inclusion in the trial, a higher proportion of detection of typical bacteria may have been missed [26]. The test method applied to study samples did not include S. pneumoniae or H. influenzae. It is likely that S. pneumoniae would otherwise have been detected in a substantial proportion of participants, but the relevance of this finding to establish a causal agent for the community-acquired pneumonia episode would have been questionable [2, 4, 27].

A second important limitation is the small sample size. Although we were able to sample more than 10% of children admitted for community-acquired pneumonia at the participating centres during the study period, we were not able to assess pathogen interactions or seasonality, age preponderance or association with length of hospital stay for rarer pathogens.

The trial was designed to capture a real-world patient population and is thus using a pragmatic clinical case definition [28]. Patients were selected for screening for the trial based on the clinician’s diagnosis of community-acquired pneumonia.

All patients admitted for community-acquired pneumonia at the participating sites were recorded as pre-screened patients and more than a third of these fulfilled the eligibility criteria. Although this proportion is arguably higher than in most medication trials, it is still likely that some patient groups are disproportionately affected by the exclusion criteria. These may likely include children with M. pneumoniae infection, who often present without fever, and older children who would often only be hospitalised in the presence of respiratory or immunological comorbidities that would result in exclusion from the trial. The demographics of included children nonetheless closely resemble those of all children admitted for community-acquired pneumonia in similar settings [1]. We therefore believe that the data we present are generalisable to children admitted with community-acquired pneumonia in Switzerland and largely transferable to similar European settings. Most participants were enrolled in the winter season before the start of the COVID-19 pandemic. In this way, the findings can provide information for pre-post pandemic comparisons on pathogen detection and management of paediatric community-acquired pneumonia.

Data availability statement

Raw data can be obtained from the sponsor (UKBB) upon request to the Trial Steering Committee via the corresponding author once the trial has been completed. Any requests for access to raw data will be welcomed as long as they are scientifically valid, appropriate consent for the requested level of data sharing has been obtained from participants and as long as the planned data use does not conflict with ongoing analyses by the trial team.

Acknowledgements

The authors thank the families participating in the KIDS-STEP trial and the contributing staff at all trial sites.

The authors thank Angela Huttner and Theoklis Zaoutis (Independent Data Monitoring Committee and the Trial Steering Committee) for advice on focus and delimitation of the current report against the planned final report on the clinical trial.

KIDS-STEP Trial Group:

Trial Management Group: Johannes van den Anker (Principal Investigator), Julia A. Bielicki (Co-Principal Investigator), Regina Santoro (Trial Manager), Malte Kohns Vasconcelos (Trial Physician), Michael Coslovsky (Trial Statistician)

Trial Steering Committee: Diana Gibb (chair), Theoklis Zaoutis, Olaf Neth, Henri van Werkhoven

Independent Data Manitoring Committee: Angelika Huttner (chair), Fiona Vanobhergen, Wolfgang Stöhr

University of Basel Clinical Trials Unit: Madeleine Vollmer, Patricia Arnaiz, Patrick Simon

Research Pharmacy, University Hospital Basel: Stefanie Deuster, Anne Henn

Microbiology Laboratory, University Hospital Basel: Adrian Egli

Trial Sites:

Aarau – Henrik Köhler (PI), Patrick Haberstich, Rachel Kusche, Dominik Müller-Suter

Basel – Ulrich Heininger (PI), Barbara Kern, Svetlana Beglinger, Michel Ramser, Claudia Werner, Linda Stamm, Aurora Frei

Bern – Kristina Keitel (PI), Daniel Garcia, Verena Wyss, PedNet Bern

Geneve – Anne Mornand (PI), Constance Barazzone, Klara Posfay Barbe, Natasha Loevy, Alban Glangetas, Sébastien Papis

Lausanne – Jean-Yves Pauchard (PI), Linda Guihard, Raquel Marques, Danielle Bally

Luzern – Marco Lurà (PI), Alex Donas, Michael Büttcher, Leopold Simma, Martina Bieri, Susanne Krieg, Diana Schirmann, Xenia Mandanis, Katja Hrup, Janine Stritt

St. Gallen – Christian Kahlert (PI), Konstanze Zöhrer, Anita Niederer-Loher, Tanja Wachinger

Zürich – Christoph Berger (PI), Patrick M. Meyer Sauteur, Michelle Seiler, Elena Pànisovà

References

1. Koshy E, Murray J, Bottle A, Sharland M, Saxena S. Impact of the seven-valent pneumococcal conjugate vaccination (PCV7) programme on childhood hospital admissions for bacterial pneumonia and empyema in England: national time-trends study, 1997-2008. Thorax. 2010 Sep;65(9):770–4. https://doi.org/10.1136/thx.2010.137802 

2. O’Brien KL, Levine OS, Knoll MD, Feikin DR, DeLuca AN, Driscoll AJ, et al. Causes of severe pneumonia requiring hospital admission in children without HIV infection from Africa and Asia: the PERCH multi-country case-control study. Lancet (London, England). 2019;394(10200):757-79. Epub 2019/07/02. doi: https://doi.org/10.1016/s0140-6736(19)30721-4. PubMed PMID: 31257127; PubMed Central PMCID: PMCPMC6727070. https://doi.org/10.1016/S0140-6736(19)30721-4 

3. Jain S, Finelli L; CDC EPIC Study Team. Community-acquired pneumonia among U.S. children. N Engl J Med. 2015 May;372(22):2167–8. https://doi.org/10.1056/NEJMc1504028 

4. Kohns Vasconcelos M, Loens K, Sigfrid L, Iosifidis E, Epalza C, Donà D, et al. Aetiology of acute respiratory infection in preschool children requiring hospitalisation in Europe-results from the PED-MERMAIDS multicentre case-control study. BMJ Open Respir Res. 2021;8(1). Epub 2021/07/31. doi: https://doi.org/10.1136/bmjresp-2021-000887. PubMed PMID: 34326154; PubMed Central PMCID: PMCPMC8323363. 

5. Williams DJ, Zhu Y, Grijalva CG, Self WH, Harrell FE, Jr., Reed C, et al. Predicting Severe Pneumonia Outcomes in Children. Pediatrics. 2016;138(4). Epub 2016/10/01. doi: https://doi.org/10.1542/peds.2016-1019. PubMed PMID: 27688362; PubMed Central PMCID: PMCPMC5051209 conflicts of interest to disclose. 

6. Mazur NI, Bont L, Cohen AL, Cohen C, von Gottberg A, Groome MJ, et al.; South African Severe Acute Respiratory Illness (SARI) Surveillance Group. Severity of Respiratory Syncytial Virus Lower Respiratory Tract Infection With Viral Coinfection in HIV-Uninfected Children. Clin Infect Dis. 2017 Feb;64(4):443–50. https://doi.org/10.1093/cid/ciw756 

7. Calvo C, García-García ML, Pozo F, Paula G, Molinero M, Calderón A, et al. Respiratory Syncytial Virus Coinfections With Rhinovirus and Human Bocavirus in Hospitalized Children. Medicine (Baltimore). 2015 Oct;94(42):e1788. https://doi.org/10.1097/md.0000000000001788 https://doi.org/10.1097/MD.0000000000001788 

8. Kohns Vasconcelos M, Meyer Sauteur PM, Santoro R, Coslovsky M, Lurà M, Keitel K, et al. Randomised placebo-controlled multicentre effectiveness trial of adjunct betamethasone therapy in hospitalised children with community-acquired pneumonia: a trial protocol for the KIDS-STEP trial. BMJ open. 2020;10(12):e041937. Epub 2020/12/31. doi: https://doi.org/10.1136/bmjopen-2020-041937. PubMed PMID: 33376176; PubMed Central PMCID: PMCPMC7778765. 

9. Kohns Vasconcelos M, Meyer Sauteur PM, Keitel K, Santoro R, Heininger U, van den Anker J, et al. Strikingly Decreased Community-acquired Pneumonia Admissions in Children Despite Open Schools and Day-care Facilities in Switzerland. Pediatr Infect Dis J. 2021 Apr;40(4):e171–2. https://doi.org/10.1097/inf.0000000000003026 https://doi.org/10.1097/INF.0000000000003026 

10. Harris M, Clark J, Coote N, Fletcher P, Harnden A, McKean M, et al.; British Thoracic Society Standards of Care Committee. British Thoracic Society guidelines for the management of community acquired pneumonia in children: update 2011. Thorax. 2011 Oct;66 Suppl 2:ii1–23. https://doi.org/10.1136/thoraxjnl-2011-200598 

11. Rose MA, Barker M, Liese J, Adams O, Ankermann T, Baumann U, et al. S2k-Leitlinie Management der ambulant erworbenen Pneumonie bei Kindern und Jugendlichen (pädiatrische ambulant erworbene Pneumonie, pCAP). Pneumologie. 2020 Aug;74(8):515–44. https://doi.org/10.1055/a-1139-5132 

12. Keitel K, Samaka J, Masimba J, Temba H, Said Z, Kagoro F, et al. Safety and Efficacy of C-reactive Protein-guided Antibiotic Use to Treat Acute Respiratory Infections in Tanzanian Children: A Planned Subgroup Analysis of a Randomized Controlled Noninferiority Trial Evaluating a Novel Electronic Clinical Decision Algorithm (ePOCT). Clin Infect Dis. 2019 Nov;69(11):1926–34. https://doi.org/10.1093/cid/ciz080 

13. Hsia Y, Lee BR, Versporten A, Yang Y, Bielicki J, Jackson C, et al.; GARPEC and Global-PPS networks. Use of the WHO Access, Watch, and Reserve classification to define patterns of hospital antibiotic use (AWaRe): an analysis of paediatric survey data from 56 countries. Lancet Glob Health. 2019 Jul;7(7):e861–71. https://doi.org/10.1016/s2214-109x(19)30071-3 https://doi.org/10.1016/S2214-109X(19)30071-3 

14. Dierig A, Hirsch HH, Decker ML, Bielicki JA, Heininger U, Ritz N. Mycoplasma pneumoniae detection in children with respiratory tract infections and influence on management - a retrospective cohort study in Switzerland. Acta Paediatr. 2020;109(2):375-80. Epub 2019/06/07. doi: https://doi.org/10.1111/apa.14891. PubMed PMID: 31168877; PubMed Central PMCID: PMCPMC7159768. 

15. Jain S, Williams DJ, Arnold SR, Ampofo K, Bramley AM, Reed C, et al. Community-acquired pneumonia requiring hospitalization among U.S. children. The New England journal of medicine. 2015;372(9):835-45. Epub 2015/02/26. doi: https://doi.org/10.1056/NEJMoa1405870. PubMed PMID: 25714161; PubMed Central PMCID: PMCPMC4697461. 

16. Williams DJ, Hall M, Gerber JS, Neuman MI, Hersh AL, Brogan TV, et al.; Pediatric Research in Inpatient Settings Network. Impact of a National Guideline on Antibiotic Selection for Hospitalized Pneumonia. Pediatrics. 2017 Apr;139(4):e20163231. https://doi.org/10.1542/peds.2016-3231 

17. Mattila S, Paalanne N, Honkila M, Pokka T, Tapiainen T. Effect of Point-of-Care Testing for Respiratory Pathogens on Antibiotic Use in Children: A Randomized Clinical Trial. JAMA Network Open. 2022;5(6):e2216162-e. doi: https://doi.org/10.1001/jamanetworkopen.2022.16162. 

18. Doan Q, Enarson P, Kissoon N, Klassen TP, Johnson DW. Rapid viral diagnosis for acute febrile respiratory illness in children in the Emergency Department. Cochrane Database of Systematic Reviews. 2014(9). doi: https://doi.org/10.1002/14651858.CD006452.pub4. PubMed PMID: CD006452. 

19. Donà D, Barbieri E, Daverio M, Lundin R, Giaquinto C, Zaoutis T, et al. Implementation and impact of pediatric antimicrobial stewardship programs: a systematic scoping review. Antimicrobial resistance and infection control. 2020;9:3. Epub 2020/01/09. doi: https://doi.org/10.1186/s13756-019-0659-3. PubMed PMID: 31911831; PubMed Central PMCID: PMCPMC6942341. 

20. Fancourt N, Deloria Knoll M, Baggett HC, Brooks WA, Feikin DR, Hammitt LL, et al. Chest Radiograph Findings in Childhood Pneumonia Cases From the Multisite PERCH Study. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America. 2017;64(suppl_3):S262-s70. Epub 2017/06/03. doi: https://doi.org/10.1093/cid/cix089. PubMed PMID: 28575361; PubMed Central PMCID: PMCPMC5447837. 

21. Nelson KA, Morrow C, Wingerter SL, Bachur RG, Neuman MI. Impact of Chest Radiography on Antibiotic Treatment for Children With Suspected Pneumonia. Pediatr Emerg Care. 2016 Aug;32(8):514–9. https://doi.org/10.1097/pec.0000000000000868 https://doi.org/10.1097/PEC.0000000000000868 

22. Claes AS, Clapuyt P, Menten R, Michoux N, Dumitriu D. Performance of chest ultrasound in pediatric pneumonia. Eur J Radiol. 2017 Mar;88:82–7. https://doi.org/10.1016/j.ejrad.2016.12.032 

23. Loens K, Van Heirstraeten L, Malhotra-Kumar S, Goossens H, Ieven M. Optimal sampling sites and methods for detection of pathogens possibly causing community-acquired lower respiratory tract infections. Journal of clinical microbiology. 2009;47(1):21-31. Epub 2008/11/21. doi: https://doi.org/10.1128/jcm.02037-08. PubMed PMID: 19020070; PubMed Central PMCID: PMCPMC2620840. https://doi.org/10.1128/JCM.02037-08 

24. Tschiedel E, Goralski A, Steinmann J, Rath PM, Olivier M, Mellies U, et al. Multiplex PCR of bronchoalveolar lavage fluid in children enhances the rate of pathogen detection. BMC pulmonary medicine. 2019;19(1):132. Epub 2019/07/20. doi: https://doi.org/10.1186/s12890-019-0894-7. PubMed PMID: 31319825; PubMed Central PMCID: PMCPMC6639929. 

25. Kohns Vasconcelos M, Renk H, Popielska J, Nyirenda Nyang'wa M, Burokiene S, Gkentzi D, et al. SARS-CoV-2 testing and infection control strategies in European paediatric emergency departments during the first wave of the pandemic. Eur J Pediatr. 2020:1-7. Epub 2020/10/15. doi: https://doi.org/10.1007/s00431-020-03843-w. PubMed PMID: 33051714; PubMed Central PMCID: PMCPMC7553380. 

26. Forster J, Piazza G, Goettler D, Kemmling D, Schoen C, Rose M, et al. Effect of Prehospital Antibiotic Therapy on Clinical Outcome and Pathogen Detection in Children With Parapneumonic Pleural Effusion/Pleural Empyema. Pediatr Infect Dis J. 2021 Jun;40(6):544–9. https://doi.org/10.1097/inf.0000000000003036 https://doi.org/10.1097/INF.0000000000003036 

27. Bogaert D, De Groot R, Hermans PW. Streptococcus pneumoniae colonisation: the key to pneumococcal disease. Lancet Infect Dis. 2004 Mar;4(3):144–54. https://doi.org/10.1016/s1473-3099(04)00938-7 https://doi.org/10.1016/S1473-3099(04)00938-7 

28. Dal-Ré R, Janiaud P, Ioannidis JP. Real-world evidence: how pragmatic are randomized controlled trials labeled as pragmatic? BMC Med. 2018 Apr;16(1):49. https://doi.org/10.1186/s12916-018-1038-2 

Appendix: Supplementary data

Figure S1 Inclusion over time.