Micro-managing infection: discovery of microRNA signatures with diagnostic potential for paediatric infections

Validation was performed by RT-qPCR, normalised to UniSp6. Fold-change and machine learning analyses identified an optimal biomarker panel for ruling out SBI, with performance compared to CRP and PCT.<br><br><b>Results: </b>Small RNA sequencing identified 21 candidate miRNA biomarkers of potential diagnostic relevance, and 17 miRNAs were validated by RT-qPCR. Optimal diagnostic performance was achieved using a combined biomarker panel of miRNAs, CRP, and PCT (AUC = 0.83, sensitivity = 90%, specificity = 59%), outperforming CRP alone (AUC = 0.70, sensitivity = 75%, specificity = 54%).

The novel panel is predicted to prevent 55 unnecessary antibiotic prescriptions per 1,000 children, representing a possible 11% reduction in inappropriate antibiotic use.<br><br><b>Conclusion:</b> This study identifies and validates miRNA biomarkers that enhance diagnostic accuracy for paediatric infection when combined with existing tests. While further validation in larger cohorts is required, these findings support the potential of miRNA-based diagnostic panels to improve early clinical decision-making and reduce inappropriate antibiotic use.<br><br><i>Thesis is embargoed until 31 July 2031.</i><br>