Optimising Antibiotic Use for Paediatric Lower Respiratory Tract Infections in Africa

Antimicrobial resistance (AMR) has emerged as one of the most pressing global health threats and is driven largely by the inappropriate use of antibiotics. Strengthening antibiotic stewardship is therefore a critical public health priority to slow the emergence of resistant pathogens and preserve the effectiveness of existing therapies. Stewardship in inherently complex, particularly in the management of lower respiratory tract infections (LRTIs) in children. Optimising appropriate antibiotic use requires good clinical acumen, diagnostic accuracy, and sound clinical decision-making. This decision-making is influenced by multiple environmental, social, behavioural and psychological factors.

LRTIs remain a leading cause of mortality and morbidity in children under five years of age, particularly in low- and middle-income countries (LMICs). Despite progress in pneumococcal conjugate vaccine (PCV) and Haemophilus influenzae type b (Hib) vaccine coverage, LRTIs continue to account for a significant proportion of paediatric hospital admissions and deaths. There has been a marked shift in the aetiology of LRTIs over the last two decades, with a now greater proportion of LRTIs attributable to respiratory viruses and a reduced proportion caused by bacterial infections. Data from The Pneumonia Etiology Research for Child Health (PERCH) study demonstrated that viruses accounted for approximately two-thirds of childhood pneumonia cases, while bacteria accounted for approximately one-quarter.

The World Health Organisation (WHO) pneumonia case definition is intentionally highly sensitive in order to reduce mortality, particularly in resource-limited settings. Importantly, the WHO definition does not differentiate between viral, bacterial or mixed aetiology LRTI. Consequently, many children fulfilling the criteria for severe pneumonia receive antibiotic treatment despite the predominance of viral causes. This challenge is particularly relevant in infants and young children where bronchiolitis and bronchopneumonia frequently present with overlapping clinical features and are often difficult to distinguish clinically.

Although bronchiolitis is generally managed with supportive care alone and bronchopneumonia is treated with antibiotics, clinicians frequently struggle to differentiate between these two conditions, resulting in over-prescribing of antibiotics. Diagnostic uncertainty, concerns regarding clinical deterioration, fear of missing serious bacterial infection, and limitations in available diagnostic tools all contribute to antibiotic prescribing decisions. While this approach may reduce the risk of undertreatment, it also contributes to unnecessary antibiotic exposure, increased healthcare costs, adverse drug effects, disruptions of the microbiome, and the development of antimicrobial resistance.

Clinical severity assessment tools may help identify children at higher risk of adverse outcomes. The Pneumonia Research Partnership to Assess WHO Recommendations (PREPARE) clinical severity score was derived from data obtained from more than 20 countries and has demonstrated good external validity in predicting pneumonia-related mortality among children aged 2-59 months hospitalised with pneumonia. The PREPARE score incorporates patient age, sex, weight-for-age z-score, body temperature, respiratory rate, level of consciousness, cyanosis, and oxygen saturation to produce a risk score that can assist with clinical risk stratification.

While risk assessment tools may identify children at risk of severe disease and adverse outcomes, they do not determine disease aetiology. Point-of-care (POC) biomarkers may assist clinicians in differentiating between viral and bacterial infections. C-reactive protein (CRP) is an inflammatory biomarker that has been evaluated as a tool to guide antibiotic prescribing practices. Studies conducted in Uganda and Tanzania have demonstrated that CRP-guided management strategies can safely reduce antibiotic prescribing among children with respiratory illnesses in outpatient settings. However, evidence regarding CRP-informed risk-differentiated care among African children hospitalised with LRTI is lacking.

The implementation of a clinical severity scoring system alongside serial POC CRP testing may provide a structured strategy to support safe antibiotic prescribing while reducing unnecessary antibiotic use. To our knowledge, no studies have assessed the safety of biomarker-guided and severity-score-guided withholding or early termination of antibiotics among young African children hospitalised with LRTI.

The primary component of this study is a 12-month pragmatic, unblinded randomised controlled trial conducted at Chris Hani Baragwanath Academic Hospital (CHBAH), a large tertiary referral hospital in Soweto, South Africa. Children aged 2 to 24 months admitted with a clinician diagnoses of LRTI will be eligible for enrolment following written informed consent from a parent or legal guardian.

Participants will be randomised win a 1:1 ratio to either an intervention arm or routine care arm. Children allocated to the intervention arm will undergo a baseline assessment using a decision-making algorithm incorporating both the PREPARE clinical severity score and a POC CRP test. A conservative CRP threshold of less than 10mg/L together with a PREPARE score of 4 or less will be used to identify children who are at low risk of adverse outcomes and more likely to have viral disease. Repeat PREPARE assessments and POC CRP measurements will be performed on days 1, 2 and 3 following enrolment.

The intervention is intended to support clinical decision-making regarding antibiotic initiation, continuation and early termination. Treating clinicians will retain the authority to override the algorithm at any time based on the patient's clinical condition to ensure patient safety.

Children allocated to the control arm will receive routine clinical care according to existing hospital and unit-specific practices. A baseline PREPARE score will be recorded for research purposes, but clinical management decisions will not be guided by the study algorithm. Participants will be followed throughout hospitalisation and until 28 days following randomisation. Data collected will include antibiotic use, duration of antibiotic exposure, route of administration, clinical outcomes, need for respiratory support, adverse events, readmissions and mortality,

The primary outcome is antibiotic use during hospitalisation, including whether antibiotics were initiated or withheld, route of administration, and duration of therapy. Secondary outcomes include death within 28 days of randomisation, hospital readmission within 28 days, serious adverse events, escalation of care to invasive ventilation, hospitalisation exceeding seven days, duration of oxygen supplementation, duration of antibiotic exposure, duration of hospital stay, escalation of antibiotic therapy, post-discharge antibiotic use, and clinical complications including sepsis, lung abscess, and empyema.

In addition to the randomised trial, a mixed-methods study will explore clinician attitudes, beliefs and behaviours regarding antibiotic prescribing for children hospitalised with LRTI. Despite awareness of antimicrobial stewardship principles, clinicians continue to over-prescribe antibiotics because of diagnostic uncertainty, fear of adverse outcomes, time constraints, parental expectations, influence of colleagues and senior clinicians, and limitations in diagnostic resources.

The qualitative component will be guided by the Theoretical Domains Framework (TDF) and will involve surveys, focus group discussions and semi-structured interviews among registrars and paediatricians working at CHBAH. The study aims to identify behavioural, social and environmental determinants of prescribing practices and to better understand barriers and facilitators to appropriate antibiotic use.

A health economic analysis will also be undertaken to compare healthcare-related costs between the intervention and routine care groups, including antibiotic expenditure, duration of hospitalisation, level of care received, and costs associated with respiratory support. The study will determine whether the intervention leads to cost savings through reductions in antibiotic use and shorter hospital stays.

This study aims to determine the safety and effectiveness of combining point-of-care CRP testing with clinical severity assessment to support antibiotic prescribing decisions among children aged 2-24 months hospitalised with LRTI. In addition, it seeks to improve understanding of clinician prescribing behaviour and the contextual factors that influence antibiotic use. The findings will contribute to the evidence base for paediatric antimicrobial stewardship and may inform future implementation of biomarker-guided and severity-guided antibiotic prescribing strategies in resource-limited settings.