Clinical Decision Support in Non-typhoidal Salmonella Bloodstream Infections in Children

In sub-Saharan Africa, non-typhoidal Salmonella (NTS) are a frequent cause of bloodstream infection, display high levels of antibiotic resistance and have a high case fatality rate (15%). In Kisantu hospital in the Democratic Republic of Congo (DR Congo), NTS account for 75% of bloodstream infection in children and many children are co-infected with Plasmodium falciparum (Pf) malaria. NTS bloodstream infection presents as a non-specific severe febrile illness, which challenges early diagnosis and, as a consequence, prompt and appropriate antibiotic treatment.Moreover, at the first level of care, frontline health workers have limited expertise and diagnostic skills and, as a consequence, clinical danger signs that indicate serious bacterial infections are often overlooked.

Basic handheld diagnostic instruments and point-of-care tests can help to reliably detect danger signs and improve triage, referral and the start of antibiotics, but there is need for field implementation and adoption to low-resource settings. Further, it is known that some clinical signs and symptoms are frequent in NTS bloodstream infections. The integration of these clinical signs and symptoms in a clinical decision support model can facilitate the diagnosis of NTS bloodstream infections and target antibiotic treatment.

The investigators aim to develop such a clinical decision support model based on data from children under five years old admitted to Kisantu district referral hospital in the Democratic republic of the Congo. While developing the model, the investigators will focus on the signs and symptoms that can differentiate NTS bloodstream infection from severe Pf malaria and on the clinical danger signs that can be assessed by handheld diagnostic instruments and point-of-care tests. The deliverable will be a clinical decision support model ready to integrate in an electronic decision support system.