Bloodstream infections are frequent in children admitted to the hospital for severe febrile illness in sub-Saharan Africa.Ongoing blood culture surveillance at Kisantu Hospital showed non-typhoidal Salmonella (NTS) as the first cause of bloodstream infections in children. Bloodstream infections have a high case fatality (15 - 20%). Outcome of bloodstream infections is dependent on timely diagnosis and treatment. However, observations at Kisantu Hospital showed that many children arrive late and die early after admission.

By interviewing caregivers of severely ill children admitted to Kisantu Hospital, the investigators aim to study their health itinerary, i.e. the sequence of all actions of health care seeking and care provision between the onset of febrile illness and the admission at the hospital. The investigators aim to assess the health itinerary according to the "three delays" model. The three delays model studies delays and practices at the level of health care seeking, of transport and of start of antibiotic treatment.10 Visits to referring health centers will provide complementary information about diagnosis, treatment and referral practices. In hospital follow-up will allow to assess the outcome according to the duration of health itinerary. The results of routine laboratory tests upon hospital admission will allow to stratify the health itinerary according to fever etiology.

The results of this study will allow to understand the duration of the health itinerary, its possible association with case-fatality, and factors explaining for delays at every level. This information is expected to orient local health policy makers towards interventions shortening the duration of the health itinerary and in that case improve and monitor the referral system. In addition, the study results are expected to orient towards further research to understand health seeking behavior (i.e. focus-group discussions and community-based studies).

The study is designed as a hospital-based cohort study at the pediatric ward of Kisantu general referral hospital. The investigators foresee a study inclusion period of 5 months. The study period is currently determined as such that it includes approximately 2 months of the dry season and 3 months of the rainy season. Seasonality is taken into account, because the rain affects the condition of the road and local disease epidemiology and can thus influence the study results. Most data will be collected retrospectively, i.e. all data on the health itinerary collected from the questionnaire and referral letter and the data from the health records in the referring health centers. Only the data from the in-hospital laboratory tests and clinical outcome will be collected prospectively.

The data from the health records at referring health centers will be consulted after termination of the five months study inclusion period to avoid any influence from the study in the patient management at health centers during the study.

Inclusion Criteria:

• Be a child between 28 days and 5 years old
• Be admitted to Kisantu Hospital
• Have a suspected bloodstream infection, which is defined as the presence of objective fever, hypothermia or a history of fever during the past 48 hours and at least one of the following criteria: Hypotension, confusion or increased respiratory rate/Suspicion of severe localized infection: pneumonia, meningitis, osteomyelitis, complicated urinary tract infection, abscess, skin/soft tissue infection or abdominal infection/ Suspicion of typhoid fever/ Suspicion of severe malaria
• Having a caregiver willing and able to provide written informed consent

Exclusion Criteria:

-

The study setting is the Kisantu health zone, which has a population of 202 451 people, of which 48 043 are under 5 years old. The health zone is subdivided in 17 health areas that are each served by at least one health center, by smaller health posts and by community health workers.

The formal health care system is composed of 51 health care facilities. Next to the formal health care system of community health workers, health posts and health centers, many informal health care providers exist (Nurses and doctors provide care, private pharmacies or drug vendors exist and traditional healers).

Local disease prevalence:

• High P. falciparum malaria endemicity: 24% of children between 6-59 months had a positive blood microscopy test in the last national demographic and health survey.
• High prevalence of malnutrition: 46% and 11% of children under five were respectively chronically and acutely malnourished in the last national demographic and health survey.