The diagnosis of late onset sepsis in neonates is complex and therefore usually late, resulting in increased risks. The Digi-NewB project proposes a novel solution to this problem, by designing a non-invasive Decision Support System (DSS) which will use vital signs, images and sounds to measure the risk of sepsis in the preterm infant and therefore support clinicians in diagnosis. The introduction of any new system to a neonatal intensive care unit (NICU) environment presents a challenge for designers who must account for a technology laden environment and a demanding work-load for clinicians. To define the user needs and therefore build the first use cases for such a system, a multi-method approach was adopted and is described in this paper. This approach consisted of a period of ethnography, eleven semi-structured interviews and the application of a prototyping exercise based on the principles of participatory design.