Prevalence of serious bacterial infections and management of febrile infants ≤60 days in Swedish Pediatric Emergency Departments

Abstract: BACKGROUND. The reported prevalence of serious bacterial infections (SBI: urinary tract infection (UTI), bacteremia, and meningitis) among febrile infants ≤60 days of age varies from 8% to 23%. There are indications that the prevalence of infections differs between the sexes. Infants with fever at home but afebrile at the pediatric emergency department (PED) might have a lower risk of SBIs. Furthermore, the management of febrile infants and adherence to guidelines varies among PEDs. There is a paucity of knowledge regarding how physicians decide whether to follow management recommendations. AIM. To describe the age- and sex-specific prevalence of SBIs in infants aged ≤60 days with fever without source (FWS) at four PEDs in Sweden. To evaluate whether there is a difference in the prevalence between infants with reported fever at home who are afebrile at the PED and those who are still febrile. To describe the clinical management and outcomes of febrile infants. To investigate physicians’ decision-making process when managing febrile infants aged ≤60 days and to describe the factors that influenced this decision.METHOD. This thesis is comprised of 2 separate projects. The first is a retrospective cross-sectional study in previously healthy, full-term febrile infants ≤60 days with FWS who presented at 4 PEDs in 2014-2020. The second is a qualitative study with a phenomenographic approach based on focus group discussions with physicians active in 2 of the PEDs. RESULTS. There were included 2237 febrile infants aged ≤60 days with FWS. The prevalence of SBIs was 12.6% (95% CI, 11.0-14.3), of UTI 11.0% (95% CI, 9.5-12.6), of bacteremia 1.5% (95% CI, 1.0-2.2), and of meningitis 0.5% (95% CI, 0.2–0.9). In infants aged ≤28 days, the prevalence of meningitis did not differ (p=1.000) between girls 0.8% (95% CI, 0.1-2.9) and boys 0.9% (95% CI, 0.2-0.2.7). Similarly, there was no difference in the risk of meningitis between infants aged ≤28 days with reported fever at home who were febrile at the PED and those still febrile, with a risk ratio (RR) of 1.05 (95% CI, 0.18–6.23). In infants aged ≤28 days, lumbar puncture (LP) was performed in 13% (95% CI, 11–16), blood culture in 40% (95% CI, 36–40), broad-spectrum antibiotics were administered in 30% (95% CI, 26–34), and 67% (95% CI,63–71) were hospitalized. Of the infants who did not receive antibiotics at the initial approach, 0.3% (95% CI, 0.1–0.8) were diagnosed with meningitis or bacteremia. Three main factors influenced the decision-making process on whether to perform an LP:1) a possible focus of infection that could explain the origin of the fever; 2) questioning whether the temperature at home reported by the parents was a fever, especially if it was ≤38.2°C; and 3) the infant’s general condition and questioning the need for LP in case of well-appearing infants.CONCLUSIONS. The prevalence of meningitis and bacteremia was low in infants aged ≤60 days with FWS. A different meningitis risk estimation is not justified for infants aged ≤28 days with reported fever at home afebrile at the PED. The management of febrile infants aged ≤28 days did not constitute of routine LP, blood culture, antibiotic treatment, and hospitalization but was not associated with increased adverse outcomes. Such management should be investigated further since it could reduce unnecessary investigations, antibiotic treatments, and hospitalizations. The primary factors that influenced physicians to omit performing LP were general appearance, presence of fever, and possible focus of infection.