PhD defence W. (Wendy) van Herk

Below is a brief summary of the dissertation:

Neonatal early-onset sepsis (EOS) is one of the leading causes of neonatal mortality and morbidity worldwide and the start of early antibiotic treatment is crucial. Uncertainty about the presence of infection in newborns at risk for EOS can lead to unnecessary and prolonged antibiotic treatment. Antibiotics can be life-saving but also harmful due to, for example, increasing resistance with overuse of antibiotics. There is a need for better antibiotic management in newborns suspected of EOS. This thesis provides an overview of the current (clinical) challenges. A possible solution could be a combined stratified risk approach based on neonatal characteristics, but this is challenging because of the dynamic clinical presentation of EOS. Non-specific clinical signs and conventional laboratory parameters complicate the diagnosis of EOS. A potentially promising diagnostic tool for reducing antibiotic treatment in neonates with suspected EOS is the biomarker procalcitonin. Our study shows that procalcitonin-driven decision-making may be a valuable tool for reducing the duration of antibiotic treatment in neonates with suspected EOS without changing clinical outcomes or healthcare costs. We also found that biomarkers are superior to clinical symptoms and risk factors for the prediction of blood culture-positive EOS. To optimize clinical decision-making for treatment of neonates suspected of EOS, we propose that biomarkers should be considered in algorithms; there is a role for combining biomarkers (PCT-guided decision-making) and already existing alternative strategies such as machine learning (neonatal sepsis calculator).