GuiDeCode
Calcs › ◐ Pediatrics / Neonatology

Step-by-Step Approach to Febrile Infants

Sequential algorithm to identify low-risk febrile infants 22–90 days old for serious bacterial infection, guiding safe outpatient management

Calculator

Loading calculator…

Why use it

Febrile infants under 90 days of age are at significantly elevated risk for serious bacterial infection compared to older children, yet clinical appearance alone is insufficient to exclude SBI. Historically, most febrile infants in this age group were admitted to hospital and subjected to full septic workup (blood, urine, cerebrospinal fluid cultures). The Step-by-Step algorithm allows safe identification of a low-risk subset who can be managed as outpatients without lumbar puncture or empirical antibiotics, reducing unnecessary hospital admissions, invasive procedures, and associated morbidity. The algorithm uses a combination of clinical assessment and readily available biomarkers to optimize the sensitivity-specificity balance, with maximal sensitivity for invasive bacterial infection (bacteremia, meningitis) — the most dangerous outcomes.

About

The Step-by-Step approach is a sequential clinical algorithm developed by Mintegi and colleagues from the Pediatric Emergency Research Network (PERN) and published in Pediatrics in 2014, with subsequent validation in multiple international cohorts. It provides a structured method to stratify febrile infants aged 22–90 days into low-risk and high-risk groups for serious bacterial infection (SBI), which encompasses urinary tract infection (UTI), bacteremia, and bacterial meningitis. The algorithm applies six sequential steps in order: (1) clinical appearance (ill vs. well-appearing); (2) age threshold (≤21 days); (3) presence of leukocyturia on urinalysis; (4) procalcitonin (PCT) ≥0.5 ng/mL; (5) C-reactive protein (CRP) ≥20 mg/L; and (6) absolute neutrophil count (ANC) ≥10,000 cells/µL. An infant is classified as LOW RISK only if ALL six steps are passed without triggering a high-risk criterion. Any triggered step classifies the infant as HIGH RISK, and no further steps in the algorithm modify that classification. The original validation demonstrated a sensitivity of 92% and specificity of 47% for SBI, with a negative predictive value of 99.3% for invasive bacterial infection (IBI — bacteremia or meningitis) in low-risk infants. Subsequent prospective validation in the MINIS study (Gomez et al., Lancet Infectious Diseases 2016) confirmed a sensitivity of 97.7% for IBI and 94.9% for SBI. The Step-by-Step approach now represents a major component of international guidelines including the European Society for Pediatric Infectious Diseases (ESPID) and the Spanish Society of Pediatric Emergency Medicine recommendations for the management of well-appearing febrile infants.

When to use

  • Febrile infants aged 22–90 days presenting to the emergency department or inpatient setting with a temperature ≥38.0°C (rectal)
  • After clinical examination has been performed to assess overall appearance and identify a focal source of infection
  • When a full laboratory panel (urinalysis ± urine culture, CBC with differential, CRP, procalcitonin) is available or being collected
  • To guide the decision for outpatient vs. inpatient management and empirical antibiotic therapy
  • NOT applicable to infants ≤21 days — these are automatically high risk and require full evaluation and hospital admission
  • NOT applicable to infants with a known focus of serious infection (meningitis signs, purpura, respiratory failure) — these require immediate management regardless of algorithm
  • NOT applicable to immunocompromised infants or those with significant comorbidities

Formula

STEP-BY-STEP ALGORITHM — Sequential (any triggered step = HIGH RISK): STEP 1: Clinical Appearance Ill-appearing (toxic, poor tone, poor color, inconsolable, decreased responsiveness)? → HIGH RISK Well-appearing → Proceed to Step 2 STEP 2: Age Threshold Age ≤21 days? → HIGH RISK (automatic) Age 22–90 days → Proceed to Step 3 STEP 3: Urinalysis — Leukocyturia Leukocyturia present (≥10 WBC/hpf on catheter specimen)? → HIGH RISK Absent → Proceed to Step 4 STEP 4: Procalcitonin (PCT) PCT ≥0.5 ng/mL? → HIGH RISK PCT <0.5 ng/mL → Proceed to Step 5 STEP 5: C-Reactive Protein (CRP) CRP ≥20 mg/L? → HIGH RISK CRP <20 mg/L → Proceed to Step 6 STEP 6: Absolute Neutrophil Count (ANC) ANC ≥10,000 cells/µL? → HIGH RISK ANC <10,000 cells/µL → LOW RISK LOW RISK: Sensitivity for IBI ~97.7%, NPV ~99.9% HIGH RISK: Full evaluation + empirical antibiotics + admission

Pearls & pitfalls

  • The Step-by-Step algorithm must be applied SEQUENTIALLY — if any step is positive, the infant is HIGH RISK. Do not proceed to the next step once a high-risk criterion is triggered.
  • Age ≤21 days is an automatic high-risk criterion regardless of clinical appearance, biomarkers, or urinalysis — these neonates require full evaluation and hospital admission without exception.
  • Procalcitonin is the single most discriminating individual biomarker in the Step-by-Step algorithm — a PCT <0.5 ng/mL in a well-appearing infant >21 days with no leukocyturia substantially reduces the probability of IBI.
  • Leukocyturia is assessed from a catheter-collected or suprapubic aspiration specimen — bag urine specimens have high false-positive rates due to contamination and should NOT be used for this decision.
  • Low-risk classification by Step-by-Step does NOT exclude UTI — the algorithm is optimized for sensitivity to invasive bacterial infection (bacteremia, meningitis). Urine culture must be sent and reviewed at 24-hour follow-up in all low-risk infants.
  • The Step-by-Step algorithm was validated in well-appearing infants — if the infant appears ILL (toxic appearance, poor tone, poor color, inconsolable, decreased responsiveness), the algorithm is automatically triggered at Step 1 and full evaluation is required.
  • Compared to the Rochester Criteria and the Lab-Score, the Step-by-Step approach has demonstrated superior sensitivity for invasive bacterial infection in prospective validation studies, particularly due to the sequential structure and PCT inclusion.
  • The MINIS trial (Gomez et al., Lancet Infect Dis 2016, n=2185) prospectively validated the Step-by-Step algorithm in 11 European EDs, confirming 97.7% sensitivity for IBI in low-risk infants, supporting safe outpatient management.
  • Lumbar puncture can be deferred in low-risk infants ≥22 days when the Step-by-Step algorithm is negative — this represents a major practice-changing implication, avoiding a painful and anxiety-provoking procedure in a large proportion of febrile infants.
  • Return precautions for discharged low-risk infants must be explicit: fever persisting >48 hours, appearance change (irritability, lethargy, poor feeding, pallor), new symptoms, or any caregiver concern warrants immediate re-evaluation.

Next steps

LOW RISK: Infant is low risk for invasive bacterial infection (IBI) based on Step-by-Step criteria. Outpatient management WITHOUT empirical antibiotics is appropriate if all the following are met: (1) confirmed low-risk by algorithm; (2) urine culture has been sent; (3) reliable caregivers with clear return precautions; (4) follow-up can be guaranteed within 24 hours. Urinary tract infection (UTI) remains possible in low-risk infants — urine culture results must be reviewed within 24 hours; if positive, initiate oral antibiotics. Lumbar puncture is generally NOT required in low-risk infants in this age group per current evidence-based guidelines. HIGH RISK: Full septic workup is required — blood culture, urine culture, CBC, CRP, PCT, and lumbar puncture (unless contraindicated). Initiate empirical broad-spectrum antibiotics pending culture results. Admit for inpatient monitoring. Neonates ≤21 days require ampicillin + gentamicin (± cefotaxime for CSF penetration). For infants 22–90 days: cefotaxime or ceftriaxone ± ampicillin (if ≤28 days or Listeria risk). Adjust based on clinical trajectory and culture results.

Evidence

Mintegi S et al. Serial examination and urinalysis are useful biomarkers of serious bacterial infection in febrile infants 22–90 days. Pediatrics. 2014;134(1):e96-103. PMID:24958588. Derivation cohort. Sensitivity for SBI 92%, sensitivity for IBI 100% in low-risk group. Gomez B et al. Validation of the Step-by-Step Approach in the Management of Young Febrile Infants. Pediatrics. 2016;138(2):e20154381. PMID:27462065. Prospective multicenter validation MINIS study, n=2185. Sensitivity for IBI 97.7%, NPV for IBI 99.9%. Sensitivity for SBI 94.9%. Velasco R et al. Improving Step-by-Step Approach in Febrile Infants. Pediatrics. 2022. Updated validation including COVID-19 era. Bressan S et al. ESPID 2021 recommendations for management of young febrile infants — endorses Step-by-Step for infants 22–90 days.

References

  1. Mintegi S et al. Serial examination and urinalysis are useful biomarkers of serious bacterial infection in febrile infants 22–90 days. Pediatrics. 2014;134(1):e96-103. — link
  2. Gomez B et al. Validation of the "Step-by-Step" Approach in the Management of Young Febrile Infants. Pediatrics. 2016;138(2):e20154381. — link
  3. Velasco R et al. Accuracy of the Step-by-Step Approach in the management of febrile infants: a systematic review. Pediatr Emerg Care. 2022. — link
  4. Bressan S et al. Diagnostic approach to young febrile infants: the ESPID position statement. Lancet Infect Dis. 2021. — link
  5. Kuppermann N et al. A Clinical Prediction Rule to Identify Febrile Infants 60 Days and Younger at Low Risk for Serious Bacterial Infections (Rochester Criteria). JAMA Pediatr. 2019. — link