Impact factor (WEB OF SCIENCE - Clarivate)

2 year: 7.2 | 5 year: 6.6

Ongoing Research Themes

Can child pneumonia in low-resource settings be treated without antibiotics? A systematic review & meta-analysis

Patrick JB Walker1, Chris Wilkes1, Trevor Duke1,2, Hamish R Graham1,2, ARI Review group

1 Centre for International Child Health, Murdoch Children’s Research Institute, University of Melbourne, Royal Children’s Hospital, Parkville, Victoria, Australia
2 Department of Paediatrics, University of Melbourne, Parkville, Victoria, Australia

DOI: 10.7189/jogh.12.10007

Share:

Facebook
Twitter
LinkedIn
Abstract

Background

WHO guidelines recommend the use of antibiotics for all cases of pneumonia in children, despite the majority being caused by viruses. We performed a systematic review and meta-analysis to determine which children aged 2-59 months with WHO-defined fast breathing pneumonia, if any, can be safely treated without antibiotics.

Methods

We systematically searched medical databases for articles published in the last 20 years. We included both observational and interventional studies that compared antibiotics to no antibiotics in children aged 2-59 months diagnosed with fast breathing pneumonia in low- and middle-income countries (LMICs). We screened articles according to specified inclusion and exclusion criteria, and assessed for risk of bias using the Effective Public Health Practice Project (EPHPP) framework. Overall, we included 13 studies in this review. We performed a meta-analysis of four included studies comparing amoxicillin to placebo.

Results

Most children with fast breathing pneumonia will have a good outcome, regardless of whether or not they are treated with antibiotics. Meta-analysis of four RCTs comparing amoxicillin to placebo for children with pneumonia showed higher risk of treatment failure in the placebo group (odds ratio OR 1.40, 95% confidence interval CI = 1.00-1.96). We did not identify any child pneumonia subgroups in whom antibiotics can be safely omitted. Limited data suggest that infants with clinically-diagnosed bronchiolitis are a particular low-mortality group who may be safely treated without antibiotics in some contexts.

Conclusions

Children with WHO-defined fast breathing pneumonia in LMICs should continue to be treated with antibiotics. Future studies should seek to identify which children stand to benefit most from antibiotic therapy, and identify those in whom antibiotics may not be required, and in which circumstances.

Print Friendly, PDF & Email

Pneumonia is the leading single cause of death in children under five, killing an estimated 740 000 children in 2019 [1]. WHO guidance recommends antibiotics for all children meeting WHO’s broad clinical definition for pneumonia (cough and fast or difficult breathing) [2,3]. WHO guidance does not identify any patient groups in whom antibiotics may not be needed despite the majority of pneumonia episodes being caused by viruses [2,3]. This approach is designed to ensure children with bacterial pneumonia at risk of death do not go without antibiotic therapy, but has meant that many children with viral infections are being treated with antibiotics. This treatment paradigm its in tension with antimicrobial stewardship efforts, intended to optimise the use of antimicrobials, improve patient outcomes, prevent antimicrobial resistance, and save health care costs [48]. In the era of widespread uptake of vaccination against Streptococcus pneumoniae and Haemophilus influenzae this has become even more important, as the proportion of cases caused by bacterial pathogens continues to falll [9].

Differentiation between bacterial and viral pneumonia is challenging and clinical signs alone are often unreliable [10]. A number of studies have examined the utility of biomarkers including white cell count (WCC), C-reactive protein (CRP), and procalcitonin (PCT), in conjunction with clinical symptoms and examination findings, in determining likely aetiology. These studies have shown somewhat inconsistent findings, generally indicating that biomarkers may have a role in identifying children with bacterial rather than viral infection but are insufficiently sensitive to “rule out” bacterial infection [1113]. The usefulness of these biomarkers to health care workers (and patients) using WHO guidelines in low- and middle-income countries (LMICs) is limited further by low availability and high relative cost [1113].

For infants with clinically-diagnosed bronchiolitis, WHO hospital guidelines acknowledge that, in the absence of signs of severe illness or danger signs, these children are unlikely to benefit from antibiotics [3]. However, the WHO pneumonia case definition is broad, including all children with “cough and fast or difficult breathing”. This means that many children in LMICs with uncomplicated bronchiolitis are likely to be treated with antibiotics, as they fulfil WHO criteria for pneumonia despite having a viral infection which is unlikely to require antibiotics [14].

We conducted a systematic review and meta-analysis to examine current evidence on the effectiveness of antibiotics in children with WHO-defined fast breathing pneumonia in LMICs and determine which children, if any, can be safely managed without antibiotics, and in which contexts.

METHODS

Types of studies

We included studies that: 1) were published in English in the year 2000 or later; 2) included children with WHO-defined fast breathing pneumonia; 3) included children who were treated with antibiotics and children who were treated without antibiotics; and 4) were wholly or partially undertaken in LMICs. We included both interventional and observational studies. Full inclusion and exclusion criteria are detailed in Table 1.

Table 1.  Inclusion and exclusion criteria

WordPress Data Table

WHO – World Health Organization, LMICs – low- and middle-income countries, ICU – intensive care unit

Search strategy

We developed the protocol for this systematic review and meta-analysis in accordance with PRISMA reporting guidelines [15,16]. We conducted a systematic search of medical databases Medline, Embase, and PubMed for all relevant articles. We mapped search terms to medical subject headings where possible, using Boolean operators to combine searches into our final systematic search query. We used synonyms of ‘pneumonia’, ‘antibiotics’, and ‘child’ to target our search strategy, with oversight from an experienced Health Service Librarian to ensure all relevant papers were identified. The specific search terms used for our Medline search are included in Appendix S1 IN the Online Supplementary Document. We also searched reference lists of all included references for eligible studies.

Assessment of study eligibility

Two reviewers, PW and CW, independently screened the titles and abstracts of all returned studies. We obtained full-text for studies that were screened in by either reviewer, and the same two reviewers independently assessed them for inclusion. We resolved disagreements by discussion and, where appropriate, review by a third reviewer, HG. None of the reviewers were blind to the journal titles, study authors, or affiliated institutions.

Data management, extraction and synthesis

We used a standardised data extraction form to extract data relevant to our review. Two reviewers, PW and CW, independently extracted data from each eligible study and entered data into an Excel spreadsheet (Microsoft, Redmond, US). We resolved disagreements by discussion, and contacted study authors where appropriate to resolve any uncertainties. Types of data extracted are listed in Table S1 in the Online Supplementary Document.

Meta-analysis

We conducted a meta-analysis of four RCTs [1720] comparing oral amoxicillin to placebo in children diagnosed with fast breathing pneumonia. These studies had similar patient populations, similar methodologies, and similar primary outcomes (treatment failure at day 3 or 4). We used Stata 17.0 (StataCorp, College Station, TX, USA) to perform the meta-analysis, using raw data from each of the four studies imported into Stata using Microsoft Excel (Microsoft Corporation, NM, USA). We used a random effects model, and reported outcomes both as odds ratios (using exponentiated effect sizes) We set the confidence level at 95%.

Assessment of study quality and risk of bias

We assessed the quality and risk of bias of all included studies by using the Effective Public Health Practice Project (EPHPP) Quality Assessment Tool [21,22]. Using this tool, two reviewers, PW and CW, independently rated studies as strong, moderate or weak with respect to selection bias, study design, confounders, blinding, data collection method, withdrawals and dropouts, and a global rating. Where disagreements occurred, a third reviewer, HG, carried out a final assessment. Risk of bias of each study is included in Table S2 in the Online Supplementary Document.

RESULTS

Our database search returned 841 records, and review of references returned an additional five papers. After removing duplicates, 649 papers were eligible for screening. We excluded 608 papers based on title and abstract screening, and a further 28 papers based on full-text review. Overall, we included 13 papers in our review (Figure 1).

Figure 1.  PRISMA flow chart for inclusion of studies. LMIC – low- or middle-income country, WHO – World Health Organization.

We included four randomised-controlled trials (RCTs) which compared antibiotic therapy with placebo in children with WHO-defined fast breathing pneumonia (previously known as non-severe pneumonia) [1720], and five RCTs which compared antibiotic therapy to no antibiotic therapy in children with clinically diagnosed bronchiolitis [2327]. We also included two RCT sub-analyses [28,29], one cohort study [30], and one systematic review [31], all of which looked at the utility of antibiotic therapy in children with fast breathing pneumonia. Detailed characteristics of included studies can be found in Table 2.

Table 2.  Detailed characteristics of included studies

WordPress Data Table

WHO – World Health Organization, RCT – randomised controlled trial, SEARO – Regional Office for South-East Asia, LMIC – lower-middle income country, PCV – pneumococcal conjugate vaccine, Hib – Haemophilus influenzae type B, RR – respiratory rate, I – intervention, C – control, TDS – three times per day, AFRO – Regional Office for Africa, LIC – low-income country, HIV – human immunodeficiency virus, SAM – severe acute malnutrition, TB – tuberculosis, EMRO – Regional Office for the Eastern Mediterranean, BD – twice per day, PAHO – Regional Office for the Americas, UMIC – upper-middle income country, LRTI – lower respiratory tract infection, EURO – Regional Office for Europe, TF – treatment failure, CHW – community health-worker

Antibiotics vs placebo in children with WHO-defined fast breathing pneumonia

All of the four RCTs comparing antibiotics to placebo in children with pneumonia took place in low- or lower-middle income countries (two in Pakistan, one in India, and one in Malawi). Three took place in hospital outpatient departments and one was conducted in primary health centres. Each trial included children with fast breathing pneumonia as defined by WHO (cough or difficult breathing with chest indrawing or fast breathing for age) [32], and excluded children who responded to inhaled bronchodilators and those with WHO general danger signs (inability to drink or breastfeed, vomiting everything, convulsions, lethargy, or unconsciousness). All four trials used oral amoxicillin in the intervention arm and evaluated “treatment failure” at day 3 or 4 as the main clinical outcome and relapse at day 14 as a secondary outcome. Treatment failure definitions varied slightly between studies, typically including death, hypoxaemia, WHO emergency or severe pneumonia signs, chest wall indrawing, or admission to hospital. Notably, Awasthi et al included presence of wheeze despite treatment in their definition of treatment failure [17]. Primary outcomes reported were treatment failure at day 3 [19] or day 4 [17,18.20].

Three trials, which together included 6799 children, found that treatment with placebo was associated with a significantly higher rate of treatment failure by day 4 (OR range 1.28-1.92, 95% CI range = 1.01-2.70) [17,18,20]. The remaining trial, which included 873 children, found no difference (OR 0.88, 95% CI = 0.57-1.39) (Table 3) [19]. Treatment failure rates were low in both antibiotic and placebo groups with moderate variability between studies: median 7.6% (range 2.5-19.9) vs 8.5% (range 4.8 to 24.1). Meta-analysis showed a significantly higher rate of treatment failure in children treated with placebo than those treated with amoxicillin (OR 1.40, 95% CI = 1.00-1.96; RD 2%, 95% CI = 1%-3%) (Figure 2). Overall, 3551/3840 (92.5%) of children who received amoxicillin and 3448/3828 (90.1%) of children who received placebo had good day 4 outcome and, 3448 (90.1%) had a good outcome and, of those cured on day 4, 3.9% (136/3524) of the amoxicillin group and 3.4% (116/3408) of the placebo group had relapsed at day 14.

Table 3.  Key results from randomised controlled trials comparing antibiotics to no antibiotics in children with WHO-defined fast breathing pneumonia

WordPress Data Table

OR – odds ratio, 95% CI – 95% confidence interval, RD – risk difference

Intention-to-treat results used unless otherwise stated; statistically significant results underlined

95% confidence intervals reported in parentheses following odds ratios unless otherwise stated

*Per-protocol figures (ITT not reported).

Figures from original paper adjusted (inverted) to enable comparison of studies.

Figure 2.  Meta-analysis forest plot describing risk of treatment failure by day 3 or 4 in children with WHO-defined fast breathing pneumonia. 95% CI – 95% confidence interval.

Mortality was exceptionally low and similar between antibiotic and placebo groups (1/3840 vs 1/3828). A secondary analysis of Ginsburg et al’s trial [29] found that few participants reported any severe adverse effects (SAEs) in either the placebo (9.6%) or amoxycillin (7.8%) group and only 5/102 (4.9%) SAEs were possibly related to amoxicillin [29]. Awasthi et al reported 30 hospitalisations due to adverse effects with no deaths [17], and Jehan et al reported low rates of adverse effects (3.0% and 2.1% and 3.0% of participants in the amoxicillin and placebo groups, respectively) [20].

None of the included trials identified any population subgroups in whom placebo was as effective as amoxicillin. Jehan et al found increased rates of treatment failure for those without antibiotics were still evident after defining subgroups according to age, presence or absence of fever, and presence or absence of wheeze [20]. Similarly, Ginsburg et al found that none of age, respiratory rate, malnutrition, malaria, or pneumococcal conjugate or pentavalent vaccine status affected rates of treatment failure [18]. However, a number of factors were associated with an increased risk of treatment failure across both antibiotic and placebo groups: presence of wheeze, tachypnoea, fever, vomiting, and diarrhoea [19,20].

All the included trials specified strict exclusion criteria, including clinical signs of severity (eg, WHO general danger signs) as well as comorbid conditions and malnutrition. All trials included regular clinical review, typically 2-3 times during the first week and 1-2 times in the second week, and most provided parents with a phone number to call if concerned. All trials used pulse oximetry routinely (except Hazir et al., [19] which did not report pulse oximetry use), excluding children who had hypoxaemia at baseline and using hypoxaemia as a key marker of treatment failure.

A single secondary analysis of an RCT evaluated the effect of using C-reactive protein (CRP) in addition to pulse oximetry as part of a treatment algorithm to guide antibiotic use. This study, conducted in hospital outpatient departments and urban health centres in Tanzania, reported a lower rate of day 7 treatment failure compared to usual care (OR 0.60, 95% CI = 0.37-0.98) and significantly reduced antibiotic use [28]. A sub-analysis of a cohort study of rural Malawian children with fast-breathing pneumonia seen by community health workers compared children who were treated with co-trimoxazole with those who were not. On univariate analysis co-trimoxazole was associated with lower levels of treatment failure on day 5 (OR 0.34, 95% CI = 0.16-0.74) however, this was not significant on multivariate analysis [30]. One Cochrane review conducted in 2014 looked at the effectiveness of antibiotics in children with fast breathing pneumonia and wheeze but did not identify any studies which met the authors’ inclusion criteria [31].

Mortality was low in all studies included in this review, with four children out of 8383 dying (overall case fatality rate = 0.05%). Jehan et al reported one death in each group [20], and Keitel et al reported two deaths in the control (standard of care) group [28]. The remaining studies all reported zero deaths.

Antibiotics vs no antibiotics in children with clinically-diagnosed bronchiolitis

We included five RCTs looking at the use of antibiotics in children aged less than 7-24 months with bronchiolitis. These studies all took place in hospital settings in middle-income countries (Bangladesh x3, Brazil, Turkey), and mostly included admitted patients. Definitions used of bronchiolitis varied, however all involved a clinical diagnosis by treating clinicians based on symptoms such as coryza, cough, difficulty in breathing, chest indrawing, and in one study [27] RSV positivity on nasopharyngeal aspiration. While these studies restricted inclusion to children with bronchiolitis rather than pneumonia, the vast majority would have met the broad WHO diagnostic criteria for pneumonia (or severe pneumonia). Antibiotic regimens varied between the studies, with three trials comparing both intravenous and oral antibiotics to no antibiotics, and two trials comparing oral macrolide antibiotics with placebo. Aside from Kabir et al’s trial, which excluded 146 out of 441 recruited participants [23], exclusion rates were low. Reasons for exclusions were not well reported.

Results of these studies were mixed (Table 4). Kabir et al [23] found that treatment without antibiotics was associated with a reduced hospital length of stay compared to both IV and oral antibiotics (3.67 ± 1.45 days vs 4.29 ± 1.89 and 4.44 ± 1.93 days respectively, P < 0.001), whereas Tahan et al [27] found that oral clarithromycin was associated with a reduced length of stay compared with placebo (51 vs 88 hours, IQR 48-68 and 71-100h respectively). The two other trials that reported length of stay found no significant difference between antibiotics and no antibiotics. Only Tahan et al. found a significant difference in clinical improvement rates, finding better outcomes with use of clarithromycin (P < 0.05), though this study only included 21 children. Only two trials reported mortality, both reporting zero deaths [24,26]. All participants of all trials underwent pulse oximetry, and were not discharged from hospital or outpatient follow-up until specified discharge criteria were met. Frequency of review, safety net and escalation procedure for deterioration post-discharge was not reported in any trial.

Table 4.  Key results from randomised controlled trials comparing antibiotics to no antibiotics in children with bronchiolitis

WordPress Data Table

Statistically significant results underlined; P values are reported as they were reported in individual studies. Ranges are reported only where more specific P values were not available.

*Mean ± standard deviation.

Median (interquartile range).

Defined as no re-admission to hospital post-discharge.

DISCUSSION

This review included 13 papers which assessed the efficacy of antibiotics in children with WHO defined fast breathing pneumonia and bronchiolitis. We found that treatment failure rates were low regardless of antibiotic use. However, in three of four high-quality RCTs directly comparing amoxicillin to placebo, amoxicillin was associated with lower rates of treatment failure. Our meta-analysis showed that children treated with amoxicillin have 40% higher odds of treatment success than those treated with placebo (OR 1.40, 95% CI = 1.00-1.96). In children with bronchiolitis, antibiotics do not appear to have a significant benefit. Mortality rates, where reported, were very low regardless of antibiotic use.

Our results indicate that the majority of children with WHO-defined fast breathing pneumonia will have a good outcome regardless of treatment with or without antibiotics, though the rate of treatment failure is likely to be lower with use of amoxicillin. Included studies did not identify population subgroups in whom antibiotics are less likely to be beneficial. Specifically, wheeze, malaria, malnutrition, age, and pneumococcal conjugate and pentavalent vaccination status were not associated with any difference in response to antibiotics. As such, current evidence does not support the use of these factors to discriminate between patients who should or should not receive antibiotics. Further research in this area is needed to reliably determine which patients are most likely to benefit from antibiotics, and in which children antibiotics may be safely withheld.

We included one study [28] which examined a CRP-informed strategy (ePOCT) to guide antibiotic use, in which non-hypoxaemic children with cough and tachypnoea and a CRP over 80mg/L were given amoxicillin, and those without an elevated CRP were not. This trial found a significantly lower rate of treatment failure on day 7 among children treated using the CRP-informed strategy (RR 0.60, 95% CI = 0.37-0.98), and significantly lower antibiotic use. A 2016 study of patients with acute respiratory infection in Vietnam which included both children and adults similarly showed reduced antibiotic use in the group in which treatment was guided by CRP, without significant difference in treatment failure or death [33]. Sub-analysis of the PERCH trial, published in 2017, similarly found that elevated CRP was associated with bacterial infection, highlighting it as a potentially useful biomarker to identify children with bacterial infection [11]. More recent data from Malaysia involving 300 children with very severe pneumonia found that male sex, crepitations and elevated CRP were associated with higher risk of bacterial infection [34]. These data suggest that strategies which involve CRP and other inflammatory biomarkers may be helpful in determining which children stand to benefit most from antibiotics, particularly when used in conjunction with clinical signs. However, more research is needed in this area, and particularly in children with fast breathing pneumonia, including data on cost implications. Given their imperfect sensitivity and specificity, care must also be taken to ensure that these biomarkers are used as an adjunct to clinical signs rather than used as a standalone diagnostic tool, to ensure serious bacterial infections in children with low markers are not missed [35].

In children with clinically-diagnosed bronchiolitis, current research from LMICs included in this review supports current WHO guidance that antibiotics are not necessary in the absence of signs of pneumonia. There is, however, considerable overlap in the presentation of bronchiolitis and pneumonia, with many children satisfying criteria for both conditions. In these cases, where the treating clinician cannot be certain that the correct diagnosis is bronchiolitis, it may be safest to treat with antibiotics.

Evidence suggests that viral-bacterial co-infection is common in children with both pneumonia and bronchiolitis, meaning antibiotic use may be useful even in children with suspected or known viral infection [36,37]. This complicates the question of when antibiotics may be safely withheld, as despite the fact that viruses cause the majority of cases, bacterial co-infection is challenging to confidently rule out [2]. Given this uncertainty, treating these children with antibiotics remains the safest option in many settings.

While the studies included in this review that examined antibiotic use in children with WHO-defined fast breathing pneumonia are of high quality and have a low risk of bias, care needs to be taken when applying these results to a real-world context. Included studies had strict inclusion criteria, generally excluding children with chronic illnesses, including malnutrition, HIV, congenital heart disease, and chronic respiratory illnesses (Table 2). These children are likely to experience higher rates of treatment failure than other children, and the studies included in this review are unlikely to be helpful in determining when to use antibiotics in these children. Exclusion rates were generally high, with some studies excluding more than half of children screened. In the four RCTs which included children with WHO-defined pneumonia, more than 7000 children were excluded, mostly due to not meeting strict inclusion criteria (Table 2). This potentially limits clinicians’ ability to confidently translate the studies’ results to their clinical practice. Included studies also reported universal pulse oximetry use, allowing children with hypoxaemia to be reliably identified. Pulse oximetry is superior to clinical signs in detecting hypoxaemia and is an essential clinical tool in caring for sick children [38]. Previous research indicates many children with pneumonia in LMICs do not have access to pulse oximetry [38], and as such many cases of hypoxaemia, particularly if mild, are likely to be unrecorded. Hypoxaemia is a major risk factor for death among children with pneumonia [39,40], and absence of universal pulse oximetry may therefore limit the generalisability of the studies included in this review. Further, additional factors such as housing, proximity and access to health care facilities, maternal literacy, and financial position may also influence children’s ability to seek medical attention should they deteriorate. The studies included in this review which examined the efficacy of antibiotics in fast breathing pneumonia generally reported escalation and referral procedures and an ability to carry out home visits where participants failed to report as planned. This degree of safety net is unlikely to be present in many settings in LMICs, meaning deterioration may not be able to be detected as easily. This should be taken into consideration when deciding whether antibiotics can be safely withheld, to avoid harm in children who cannot be reviewed at home or referred easily should they deteriorate.

CONCLUSIONS

Most children with mild acute lower respiratory tract infection characterised only by fast breathing, but without chest-indrawing or danger signs – those WHO defines as having fast breathing pneumonia – will experience a good outcome regardless of whether they are treated with antibiotics or not. However, current evidence suggests that use of oral amoxicillin is associated with a lower rate of treatment failure, and as such should be used when treating these children in LMICs. Children in whom a diagnosis of bronchiolitis can confidently be made, and who do not satisfy WHO criteria for pneumonia, can be safely treated without antibiotics. Further research is needed to identify subgroups of children with pneumonia in whom antibiotics may be safely withheld, and in which circumstances.

Additional material

Online Supplementary Document

Acknowledgements

Disclaimer: The authors alone are responsible for the views expressed in this publication and they do not necessarily represent the views, decisions or policies of the World Health Organization.

Acknowledgements: Thanks to Poh Chua, research librarian, for substantial technical support in setting up and running the database searches, and Helen Thomson and Haset Samuel for administrative support.

[1] Funding: This work was funded by a grant from the World Health Organization (WHO) to the Murdoch Children’s Research Institute (MCRI). Employees of WHO contributed to the design and oversight of the reviews. Any views or opinions presented are solely those of the authors and do not necessarily represent those of the World Health Organization, unless otherwise specifically stated.

[2] Authorship contributions: TD, HG and members of the ARI Review group conceived the study and initiated the study design. PW and CW led the conduct of searches and data extraction. Data analysis was conducted by PW, CW and HG. The manuscript was drafted by PW and CW, with input from TD and HG. All authors contributed to revisions and approved the final manuscript.

[3] Competing interests: The authors completed the ICMJE Unified Competing Interest form (available upon request from the corresponding author), and declare no conflicts of interest.

references

[1] J Perin, A Mulick, D Yeung, F Villavicencio, G Lopez, and KL Strong. Global, regional, and national causes of under-5 mortality in 2000-19: an updated systematic analysis with implications for the Sustainable Development Goals. Lancet Child Adolesc Health. 2022;6:106-15. DOI: 10.1016/S2352-4642(21)00311-4. [PMID:34800370]

[2] KL O’Brien, HC Baggett, WA Brooks, DR Feikin, LL Hammitt, and MM Higdon. Causes of severe pneumonia requiring hospital admission in children without HIV infection from Africa and Asia: the PERCH multi-country case-control study. Lancet. 2019;394:757-79. DOI: 10.1016/S0140-6736(19)30721-4. [PMID:31257127]

[3] World Health Organization. Pocket book of hospital care for children: Guidelines for the management of common childhood illnesses (2nd edition). Geneva: World Health Organization; 2013.

[4] R Colgan and JH Powers. Appropriate antimicrobial prescribing: approaches that limit antibiotic resistance. Am Fam Physician. 2001;64:999 [PMID:11578036]

[5] KG Kristinsson. Effect of antimicrobial use and other risk factors on antimicrobial resistance in pneumococci. Microb Drug Resist. 1997;3:117-23. DOI: 10.1089/mdr.1997.3.117. [PMID:9185137]

[6] C Llor and L Bjerrum. Antimicrobial resistance: risk associated with antibiotic overuse and initiatives to reduce the problem. Ther Adv Drug Saf. 2014;5:229-41. DOI: 10.1177/2042098614554919. [PMID:25436105]

[7] World Health Organization. Antimicrobial stewardship programmes in health-care facilities in low- and middle-income countries: a WHO practical toolkit. Geneva: World Health Organization; 2019.

[8] P Davey, E Brown, E Charani, L Fenelon, IM Gould, and A Holmes. Interventions to improve antibiotic prescribing practices for hospital inpatients. Cochrane Database Syst Rev. 2013;CD003543. DOI: 10.1002/14651858.CD003543.pub3. [PMID:23633313]

[9] B Wahl, KL O’Brien, A Greenbaum, A Majumder, L Liu, and Y Chu. Burden of Streptococcus pneumoniae and Haemophilus influenzae type b disease in children in the era of conjugate vaccines: global, regional, and national estimates for 2000–15. Lancet Glob Health. 2018;6:e744-57. DOI: 10.1016/S2214-109X(18)30247-X. [PMID:29903376]

[10] IJ Haq, AC Battersby, K Eastham, and M McKean. Community acquired pneumonia in children. BMJ. 2017;356:j686 [PMID:28255071]

[11] MM Higdon, T Le, KL O’Brien, DR Murdoch, C Prosperi, and HC Baggett. Association of C-Reactive Protein With Bacterial and Respiratory Syncytial Virus–Associated Pneumonia Among Children Aged <5 Years in the PERCH Study. Clin Infect Dis. 2017;64:S378-S386. DOI: 10.1093/cid/cix150. [PMID:28575375]

[12] IS Kamat, V Ramachandran, H Eswaran, D Guffey, and DM Musher. Procalcitonin to Distinguish Viral From Bacterial Pneumonia: A Systematic Review and Meta-analysis. Clin Infect Dis. 2020;70:538-42. DOI: 10.1093/cid/ciz545. [PMID:31241140]

[13] MU Bhuiyan, CC Blyth, R West, J Lang, T Rahman, and C Granland. Combination of clinical symptoms and blood biomarkers can improve discrimination between bacterial or viral community-acquired pneumonia in children. BMC Pulm Med. 2019;19:71 DOI: 10.1186/s12890-019-0835-5. [PMID:30940126]

[14] JM Mansbach, PA Piedra, SJ Teach, AF Sullivan, T Forgey, and S Clark. Prospective Multicenter Study of Viral Etiology and Hospital Length of Stay in Children With Severe Bronchiolitis. Arch Pediatr Adolesc Med. 2012;166:700-6. DOI: 10.1001/archpediatrics.2011.1669. [PMID:22473882]

[15] D Moher, L Shamseer, M Clarke, D Ghersi, A Liberati, and M Petticrew. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev. 2015;4:1-9. DOI: 10.1186/2046-4053-4-1. [PMID:25554246]

[16] MJ Page, JE McKenzie, PM Bossuyt, I Boutron, TC Hoffmann, and CD Mulrow. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Int J Surg. 2021;88:105906. DOI: 10.1016/j.ijsu.2021.105906. [PMID:33789826]

[17] S Awasthi, G Agarwal, SK Kabra, S Singhi, M Kulkarni, and V More. Does 3-day course of oral amoxycillin benefit children of non-severe pneumonia with wheeze: A multicentric randomised controlled trial. PLoS One. 2008;3:e1991. DOI: 10.1371/journal.pone.0001991. [PMID:18431478]

[18] AS Ginsburg, T Mvalo, E Nkwopara, ED McCollum, CB Ndamala, and R Schmicker. Placebo vs amoxicillin for nonsevere fast-breathing pneumonia in Malawian children aged 2 to 59 Months: a double-blind, randomized clinical noninferiority trial. JAMA Pediatr. 2019;173:21-8. DOI: 10.1001/jamapediatrics.2018.3407. [PMID:30419120]

[19] T Hazir, YB Nisar, S Abbasi, YP Ashraf, J Khurshid, and P Tariq. Comparison of oral amoxicillin with placebo for the treatment of world health organization-defined nonsevere pneumonia in children aged 2-59 months: a multicenter, double-blind, randomized, placebo-controlled trial in Pakistan. Clin Infect Dis. 2011;52:293-300. DOI: 10.1093/cid/ciq142. [PMID:21189270]

[20] F Jehan, I Nisar, S Kerai, B Balouch, N Brown, and N Rahman. Randomized Trial of Amoxicillin for Pneumonia in Pakistan. N Engl J Med. 2020;383:24-34. DOI: 10.1056/NEJMoa1911998. [PMID:32609980]

[21] S Armijo-Olivo, CR Stiles, NA Hagen, PD Biondo, and GG Cummings. Assessment of study quality for systematic reviews: a comparison of the Cochrane Collaboration Risk of Bias Tool and the Effective Public Health Practice Project Quality Assessment Tool: methodological research. J Eval Clin Pract. 2012;18:12-8. DOI: 10.1111/j.1365-2753.2010.01516.x. [PMID:20698919]

[22] BH Thomas, D Ciliska, M Dobbins, and S Micucci. A process for systematically reviewing the literature: providing the research evidence for public health nursing interventions. Worldviews Evid Based Nurs. 2004;1:176-84. DOI: 10.1111/j.1524-475X.2004.04006.x. [PMID:17163895]

[23] AR Kabir, AH Mollah, KS Anwar, AK Rahman, R Amin, and ME Rahman. Management of bronchiolitis without antibiotics: a multicentre randomized control trial in Bangladesh. Acta Paediatr. 2009;98:1593-9. DOI: 10.1111/j.1651-2227.2009.01389.x. [PMID:19572992]

[24] MJU Mazumder, MM Hossain, and AL Kabir. Management of bronchiolitis with or without antibiotics a randomized control trial. J Bangladesh Coll Phys Surg. 2009;27:63-9. DOI: 10.3329/jbcps.v27i2.4248

[25] LA Pinto, PM Pitrez, F Luisi, PP de Mello, M Gerhardt, and R Ferlini. Azithromycin therapy in hospitalized infants with acute bronchiolitis is not associated with better clinical outcomes: a randomized, double-blinded, and placebo-controlled clinical trial. J Pediatr. 2012;161:1104-8. DOI: 10.1016/j.jpeds.2012.05.053. [PMID:22748516]

[26] CH Rasul, AR Kabir, AK Rashid, AA Mahboob, and MA Hassan. Role of antibiotic in the outcome of bronchiolitis. Bronchiolitis Compendium. 2008;6:41

[27] F Tahan, A Ozcan, and N Koc. Clarithromycin in the treatment of RSV bronchiolitis: a double-blind, randomised, placebo-controlled trial. Eur Respir J. 2007;29:91-7. DOI: 10.1183/09031936.00029206. [PMID:17050564]

[28] K Keitel, J Samaka, J Masimba, H Temba, Z Said, and F Kagoro. 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;69:1926-34. DOI: 10.1093/cid/ciz080. [PMID:30715250]

[29] E Nkwopara, R Schmicker, T Mvalo, M Phiri, A Phiri, and M Couasnon. Analysis of serious adverse events in a paediatric fast breathing pneumonia clinical trial in Malawi. BMJ Open Respir Res. 2019;6:e000415. DOI: 10.1136/bmjresp-2019-000415. [PMID:31548894]

[30] C King, T Colbourn, L Mankhambo, J Beard, DC Hay Burgess, and A Costello. Non-treatment of children with community health worker-diagnosed fast-breathing pneumonia in rural Malawi: exploratory subanalysis of a prospective cohort study. BMJ Open. 2016;6:e011636. DOI: 10.1136/bmjopen-2016-011636. [PMID:27852705]

[31] ZS Lassi, R Kumar, JK Das, RA Salam, and ZA Bhutta. Antibiotic therapy versus no antibiotic therapy for children aged two to 59 months with WHO-defined non-severe pneumonia and wheeze. Cochrane Database Syst Rev. 2014;CD009576. DOI: 10.1002/14651858.CD009576.pub2. [PMID:24859388]

[32] World Health Organization. Integrated Management of Childhood Illness: Chart Booklet. Geneva: World Health Organization; 2014.

[33] NT Do, NT Ta, NT Tran, HM Than, BT Vu, and LB Hoang. Point-of-care C-reactive protein testing to reduce inappropriate use of antibiotics for non-severe acute respiratory infections in Vietnamese primary health care: a randomised controlled trial. Lancet Glob Health. 2016;4:e633-41. DOI: 10.1016/S2214-109X(16)30142-5. [PMID:27495137]

[34] AM Nathan, CSJ Teh, KA Jabar, BT Teoh, A Tangaperumal, and C Westerhout. Bacterial pneumonia and its associated factors in children from a developing country: A prospective cohort study. PLoS One. 2020;15:e0228056. DOI: 10.1371/journal.pone.0228056. [PMID:32059033]

[35] MA Elemraid, SP Rushton, MF Thomas, DA Spencer, AR Gennery, and JE Clark. Utility of inflammatory markers in predicting the aetiology of pneumonia in children. Diagn Microbiol Infect Dis. 2014;79:458-62. DOI: 10.1016/j.diagmicrobio.2014.04.006. [PMID:24857169]

[36] PG Bezerra, MC Britto, JB Correia, MdCM Duarte, AM Fonceca, and K Rose. Viral and atypical bacterial detection in acute respiratory infection in children under five years. PLoS One. 2011;6:e18928. DOI: 10.1371/journal.pone.0018928. [PMID:21533115]

[37] K Thorburn, S Harigopal, V Reddy, N Taylor, and HKF van Saene. High incidence of pulmonary bacterial co-infection in children with severe respiratory syncytial virus (RSV) bronchiolitis. Thorax. 2006;61:611-5. DOI: 10.1136/thx.2005.048397. [PMID:16537670]

[38] HR Graham, AA Bakare, A Gray, AI Ayede, S Qazi, and B McPake. Adoption of paediatric and neonatal pulse oximetry by 12 hospitals in Nigeria: a mixed-methods realist evaluation. BMJ Glob Health. 2018;3:e000812. DOI: 10.1136/bmjgh-2018-000812. [PMID:29989086]

[39] R Subhi, M Adamson, H Campbell, M Weber, K Smith, and T Duke. The prevalence of hypoxaemia among ill children in developing countries: a systematic review. Lancet Infect Dis. 2009;9:219-27. DOI: 10.1016/S1473-3099(09)70071-4. [PMID:19324294]

[40] T Duke, S Graham, M Cherian, A Ginsburg, M English, and S Howie. Oxygen is an essential medicine: a call for international action. Int J Tuberc Lung Dis. 2010;14:1362-8. [PMID:20937173]

Correspondence to:
Hamish R Graham
University of Melbourne Department of Paediatrics
The Royal Children’s Hospital Melbourne
50 Flemington Road
Parkville 3052
Victoria
Australia
[email protected]