Neonates and young infants are at increased risk of invasive bacterial infection (IBI), defined as bloodstream infection (BSI) and/or bacterial meningitis (BM). Fever (rectal body temperature ≥38°C) may be the only presenting clinical symptom fever without source (FWS), and many of those infants are clinically well-appearing.1 Despite substantial variation in clinical management of infants below 2 months of age with FWS—resulting in excessive invasive diagnostics, hospital admission, and empiric antibiotic therapy (ABT)—the past decade has seen the advent of more risk-adapted approaches. A major milestone was the publication of the American Academy of Pediatrics (AAP) Clinical Practice Guideline in 2021, providing structured algorithms for the evaluation and management of well-appearing febrile infants >7 and ≤60 days old. Since then, several pediatric emergency care registers, including networks from Spain, France and the United Kingdom,2–4 have provided additional evidence and raised important questions regarding the generalizability of this guidance across different healthcare systems. We here aim to summarize key developments since 2021, critically appraise new evidence, and discuss the applicability of contemporary risk-stratification strategies to European clinical settings, with a focus on pragmatic, evidence-informed clinical decision-making. BACKGROUND Over recent decades, numerous clinical algorithms have been proposed to identify neonates and young infants presenting with FWS who are at low risk of IBI, integrating age, clinical assessment, and laboratory parameters.5 However, subtle and often nonspecific clinical presentations of IBI in early life, limited multicenter validation of earlier tools, and poor negative predictive values (NPV) of biomarkers in young infants led to a prevailing conservative “one-size-fits-all” approach.6 This strategy favors comprehensive sepsis evaluation, hospital admission, and ABT for all febrile infants ≤60, or even ≤90, days of age. More recently, large multicenter research consortia in North America and Europe have generated prospective registry data on young infants with FWS that challenge this paradigm.7 The 2021 AAP guideline represents a synthesis of these data, introducing age- and risk-adapted diagnostic and therapeutic algorithms for well-appearing neonates and young infants >7 and ≤60 days old with FWS.8 Nevertheless, differences in access to healthcare, organizational structures, admission policies, diagnostic pathways and microbiological epidemiology may substantially influence both baseline IBI risk and performance of risk-stratification tools.9 Several international groups therefore proposed alternative or modified approaches3,7,10 or evaluated existing algorithms11,12 in diverse populations. Understanding convergent and divergent findings across these studies is essential for contextualizing contemporary management strategies.12,13 Against this background, we synthesize the evidence from 109 PubMed-indexed articles published since 2021 (available on request to the authors), compare key elements of current risk-stratification strategies, and comment on their applicability to European clinical practice. By highlighting areas of agreement, uncertainty, and ongoing controversy, we seek to inform pragmatic, evidence-based decision-making in the management of well-appearing febrile neonates and young infants. Epidemiology and Risk of IBI Among infants ≤60 days old presenting with FWS, approximately 1%–4% have IBI,7,14,15 including BSI in 1%–2% and BM in 0.3%–0.9%, both of which are associated with mortality and long-term neurodevelopmental morbidity. Urinary tract infections (UTIs), particularly in the second month of life, are substantially more common (10%–25%) and therefore necessitate appropriate urine sampling to reliably rule out or confirm infection in all infants. The most frequent causative pathogens of IBI are Escherichia coli (~40%–55%) and group B streptococci (GBS, ~20%–25%), followed by S. aureus (~10%) and Enterococcus species (~10%).16,17 As the epidemiology of IBIs shifts with age, the relative contribution of Streptococcus pneumoniae increases beyond the neonatal period, whereas infections due to Listeria monocytogenes have become exceedingly rare and are largely confined to the neonatal period.18 Isolated BSI accounts for approximately 80% of IBI cases, isolated BM for 6%, and concomitant BSI and BM for about 15%.17 However, pathogen distribution varies across regions and over time; therefore, local epidemiology and national surveillance data—if available—should always be considered when making management decisions. Defining the Well-appearing Infant Accurate classification of infants as “well-appearing” is a prerequisite for any risk-adapted management strategy.7,8 It requires exclusion of established high-risk features such as preterm birth, immunodeficiency or suppression, hospital-acquired fever, congenital malformations, focal infection, previous ABT and underlying chronic illness.8 Clinically, the assessment relies on the Pediatric Assessment Triangle, encompassing general appearance, work of breathing, and skin color/perfusion.19 Infants are considered well-appearing if no abnormalities are present in any of these domains; however, this classification requires substantial clinical experience, particularly in neonates and young infants. Ill-appearing febrile infants have an IBI incidence of 10%–20%1,3,10 and require immediate comprehensive sepsis evaluation and empiric ABT covering the relevant pathogens. Notably, well-appearing infants who are afebrile at presentation but had a documented fever earlier on the day of evaluation carry a comparable risk profile and should therefore be managed equivalently.20,21 Importantly, recently vaccinated infants represent a distinct subgroup with a different risk profile, in whom bacterial infections are largely confined to UTIs.22 Age-related Differences: Neonates Versus Infants >28 Days of Age Age remains one of the strongest predictors of IBI risk.3,7,23 Neonates (≤28 days) have a substantially higher incidence of both BSI and BM compared with infants 29–60 days of age (Fig. 1). This particularly vulnerable group of patients should always be admitted to a pediatric hospital and observed as inpatients for at least 24 hours. This allows for repeated careful clinical examination and, if necessary, repeated measurement of biomarkers over time. BSI occurs in approximately 3%–6% of neonates,24 while BM remains present in up to 2.3% of febrile infants younger than 29 days in the Spanish registry,4,25 compared with about 0.2% in those 29–60 days of age.14,23 Although the risk of BM declines markedly (by 90%) after the neonatal period, the reduction in BSI risk remains less pronounced at around 50%.14 However, well-appearing infants >28 and ≤60 days of age may be monitored in the pediatric emergency department or day unit until sufficient diagnostic information and assessment of the clinical course over time (4–6 hours) are available to justify a well-informed decision regarding inpatient admission and ABT.26 A substantial proportion of these infants will require neither lumbar puncture (LP) nor ABT.26 Several studies have highlighted the challenges of defining an earlier age cutoff within the neonatal period.8,10 Spanish registry data demonstrate that, despite the improved performance of risk-stratification algorithms incorporating C-reactive protein (CRP) and procalcitonin (PCT), a small but clinically relevant number of infants 22–28 days old with IBI remain misclassified as low risk.10 This limitation reflects the inherent latency of inflammatory biomarker responses between fever onset and clinical evaluation. As mentioned above, reported IBI incidences vary considerably between countries,27 underscoring the influence of population characteristics and healthcare structures.9 These observations support—in the absence of robust, locally derived epidemiological data—treating the entire neonatal period as a high-risk phase when interpreting risk-stratification strategies.FIGURE 1.: Practical recommendations for the management of well-appearing neonates and infants with fever without source > 7 and ≤ 60 days of age. ABT, antibiotic therapy; ANC, absolute neutrophil count; BC, blood culture; BSI, bloodstream infection; CBC, complete blood count; CRP, C-reactive protein; CSF, cerebrospinal fluid; HSV, herpes simplex virus; IV, intravenous; L, leukocytes; LP, lumbar puncture; PCR, polymerase chain reaction; PCT, procalcitonin; SDM, shared decision-making; UTI, urinary tract infection. aOne aerobic pediatric blood culture bottle (pediatric blood culture vials: liquid media for children 15/µL in neonates and >9/µL in infants) should be treated for suspected meningitis until proven otherwise (CSF culture + PCR panel). For neonatal HSV meningoencephalitis (and the empirical use of acyclovir), refer to current literature and guidelines. If LP is unsuccessful (eg, bloody tap), empirical ABT may be started and subsequently adjusted according to clinical course, biomarker trends and BC results. Persistently normal inflammatory markers over 36 hours make IBI highly unlikely. hInfants > 28 days of age who are not admitted immediately should be observed, monitored and clinically reevaluated over at least 4–6 hours (eg, short-stay/observation unit) to reliably assess the clinical course. Selection of infants for an outpatient management approach should consider patient- and parent-specific considerations: (I) parental ability to understand and agree to the management plan, (II) likelihood of good compliance, (III) feasibility of attending outpatient follow-up, (IV) availability of family support, and (V) parental ability to recognize potential clinical deterioration and to return promptly for medical reassessment if needed.Urine Analysis and Blood Culture in All Infants ≤60 Days UTI remains the most frequent bacterial infection in young infants and must always be excluded.14,28 Although young infants with UTI have a higher risk of BSI than those without UTI, concomitant meningitis is rare beyond the neonatal period.27,29 Urine collection should ideally be performed using clean-catch techniques, although bag urine samples may be used for initial screening. For microbiological confirmation, specimens should be obtained via clean-catch, transurethral catheterization, or suprapubic aspiration. Blood cultures are recommended in all febrile infants ≤60 days of age and are an essential prerequisite to the initiation of antimicrobial therapy. This is based on the recognition that BSI rates remain clinically relevant beyond the neonatal period, with reported incidences of up to 3.5% extending into the third month of life.27 Notably, the association between UTI and concomitant BSI is not confined to infants beyond the neonatal period.14,27 Evidence Supporting Invasive Evaluation in All Febrile Neonates Despite a well-appearing clinical presentation, neonates with FWS carry a sufficiently high risk of BM (>2%) to generally justify LP and ABT, even in the absence of focal or systemic signs of severe infection.25 Since the presence of UTI does not reduce the risk of BM in neonates,14,25,27 LP should always be considered in neonates presenting with FWS, even if they are well-appearing, irrespective of urine results or inflammatory biomarkers. However, recent evidence from pooled data across 4 large international research networks raises optimism that more targeted risk-stratification strategies may be developed—potentially omitting routine LP—even in neonates.30 These findings underscore the critical importance of analyzing large, population-based datasets and of prospectively evaluating any consented management algorithm as a continuously moving target in different contexts and healthcare systems. Stepwise Approaches in Infants 29–60 Days Old In contrast, the markedly lower risk of BM in well-appearing infants 29–60 days old allows for a more differentiated and stepwise approach (Fig. 1). Among infants with negative urinalysis, the risk of BSI is approximately 0.1%, supporting observation without empiric ABT.28 In this group, normal inflammatory markers reduce the estimated risk of BM to 2500, supporting observation without routine LP. Elevated inflammation markers in the absence of UTI should prompt consideration of an LP, as BM represents a plausible explanation for raised biomarkers. Conversely, infants with normal biomarkers may be observed without LP or ABT.3,10 In cases initially managed without an LP, where blood cultures eventually yield a plausible bacterial pathogen, the estimated risk of concurrent BM is approximately 5.4% in those >28 days old and 19% in neonates. Thus, given the immediate implications for choice and treatment duration of antibiotics, a low LP threshold is recommended in these cases. While a single aerobic pediatric blood culture bottle is sufficient, the recommended blood volume for young infants is 1–3 mL of whole blood. Strategies to Enhance Safety of Management The safety of risk-adapted strategies should be enhanced by specialist-level clinical assessment and structured observation. Hospital admission or prolonged observation for at least 6 hours allows detection of evolving clinical signs and integration of biomarker dynamics into the decision-making process. While no biomarker alone can reliably exclude IBI at presentation, established cutoffs for CRP, PCT, and absolute neutrophil count provide valuable adjunctive information. Persistently normal CRP and PCT values after 24–36 hours substantially reduce the likelihood of IBI, even in neonates. Multiple studies consistently demonstrated superior predictive performance of PCT compared with CRP for IBI risk stratification in young infants.31,32 PCT provides higher positive and NPVs, greater sensitivity (78% vs. 65%) and specificity (85% vs. 80%), improved overall diagnostic accuracy, lower interstudy heterogeneity, and the advantage of a validated international cutoff (0.5 ng/mL vs. variable CRP thresholds around 13–20 mg/L). Viral Testing Polymerase chain reaction (PCR) testing for respiratory viruses from throat or nasopharyngeal aspirates can provide useful additional information, particularly during seasonal outbreaks. However, viral PCRs are positive in up to 50% of febrile neonates and infants ≤60 days of age presenting with FWS, may remain positive for weeks after infection, and should therefore not be used as sole parameter to guide management decisions. Bacterial and viral co-infections may coexist, limiting the NPV of multiplex PCR panels for excluding bacterial disease.33 Although the number needed to test to detect IBI is higher in virus-RNA/DNA-positive infants, this does not justify the omission of a full septic work-up in neonates. In infants >28 days old, viral detection contributes more meaningfully to overall risk assessment but should never delay or replace appropriate bacteriological investigations. Shared Decision-making and Antimicrobial Stewardship Structured and transparent shared decision-making, including parents/legal caregivers, can improve acceptance of invasive procedures and reduce parental stress in the management of febrile young infants.34 Despite the justified goal of limiting invasive diagnostics and ABT in young infants to what is necessary, the elevated IBI risk in this vulnerable population must always be considered.35 Given increasing evidence on the long-term consequences of early-life exposure to broad-spectrum antibiotics, initiating empiric antimicrobial treatment in low-risk infants only as a “safety precaution” should be avoided. Instead, admission for observation or structured outpatient monitoring following a minimum observation period of 4–6 hours under specialist supervision represents a reasonable alternative. In many cases, however, hospital admission and longer observation periods remain necessary to ensure safety.26
Fortmann et al. (Fri,) studied this question.