Bronchopneumonia

Lower respiratory tract infections (LRTIs) remain a leading cause of hospitalisation and death among young children in Africa. Although viral infections account for a substantial proportion of LRTIs, antibiotics are frequently prescribed because distinguishing between viral and bacterial infections can be challenging. This contributes to unnecessary antibiotic exposure and the rising global public health challenge of antimicrobial resistance. This pragmatic, unblinded randomised controlled study will evaluate whether the combination of point-of-care (POC) C-reactive protein (CRP) testing and the PREPARE clinical severity score can safely reduce antibiotic use among children aged 2 to 24 months hospitalised with LRTI at Chris Hani Baragwanath Academic Hospital (CHBAH), South Africa. Participants will be randomly assigned in a 1:1 ratio to either the intervention group, where antibiotic decisions are guided by serial POC CRP measurements and repeated PREPARE severity assessments, or a control group receiving routine clinical care. The primary objective of the study is to compare antibiotic use during hospitalisation between the intervention and control groups. Secondary objectives are to evaluate the safety of this approach by comparing mortality, hospital readmission within 28 days, adverse events, need for invasive ventilation, duration of hospitalisation, and other clinical outcomes between study groups. In addition, the study will compare healthcare-related costs, including antibiotic expenditure and costs associated with hospitalisation. A parallel mixed-methods component will explore clinician attitudes, beliefs, and determinants of antibiotic prescribing for children hospitalised with LRTI using surveys, focus groups discussions, and individual interviews informed by the Theoretical Domains Framework (TDF). Findings will help identify behavioural and system-level factors that influence antibiotic prescribing and inform future antimicrobial stewardship interventions. This study aims to generate evidence on whether biomarker-guided and risk-stratified care can safely optimise antibiotic use among young African children hospitalised with LRTI while improving antimicrobial stewardship and reducing healthcare costs.

The purpose of this research is to develop a body voice artificial intelligence (AI) recognition device, also referred to as an AI-assisted body sound identification device, by utilizing a deep learning-based novel AI algorithm in conjunction with a big body voice model. It could identify normal and abnormal heart, breath, and bowel sounds, and to provide early screening and auxiliary diagnosis of congenital heart disease (CHD), respiratory infections, diarrhea and other common multi-occurring diseases.