Read the full EBM Focus and earn CME credit.
Reference - Pediatrics 2016 Aug;138(2):e20154381 (level 1 [likely reliable] evidence)
- Existing criteria have practical limitations in identifying infants ≤ 90 days old presenting with fever without apparent source who are at low risk for invasive bacterial infection (IBI) and as such, these infants are often managed conservatively.
- The Step-By-Step assessment tool sequentially evaluates ill appearance, age, urinalysis, and then blood biomarkers; once risk for IBI is indicated, subsequent factors need not be evaluated.
- In a prospective study of 2,185 infants ≤ 90 days old presenting with fever without apparent source in Europe, the Step-By-Step tool classified 45.4% of infants as having low risk for IBI and had a high negative predictive value (99.3%) in detecting IBI, suggesting that it helps rule out IBI while reducing the need for additional assessments and empiric therapy.
For infants ≤ 90 days old presenting with fever without apparent source, it is important to identify those at low risk of IBI (defined as bacterial pathogen other than contaminants in blood or cerebrospinal fluid culture) so as to limit unnecessary hospitalization and antibiotic therapy. The currently existing criteria require both clinical and laboratory assessments for risk prediction and each have one or more limitations such as reliance on lumbar puncture and/or chest radiography (J Pediatr 1992 Jan;120(1):22, N Engl J Med 1993 Nov 11;329(20):1437), lack of use of the more recently recognized biomarkers, C-reactive protein and procalcitonin (Pediatrics 1994 Sep;94(3):390), or development in a broader age group and therefore may not be applicable to infants ≤ 90 days old (Arch Dis Child 2010 Dec;95(12):968). To identify infants at low risk for IBI, the “Step-By-Step” assessment tool sequentially evaluates factors in descending order of risk for IBI. The first factor evaluated is ill appearance using criteria such as Pediatric Assessment Triangle (Pediatr Emerg Care 2010 Apr;26(4):312), then age (≤ 21 days old), urinalysis (for leukocyturia), and then blood biomarker levels (procalcitonin ≥ 0.5 ng/mL, C-reactive protein > 20 mg/L, absolute neutrophil count > 10,000/mm3). If any criterion is met, the sequential assessment is halted and the infant is classified as at risk for IBI without evaluating subsequent factors; if no criterion is met, the infant is at low risk for IBI. The Step-By-Step tool showed promise in identifying IBI in a retrospective study (Emerg Med J 2014 Oct;31(e1):e19), and the current study prospectively validates its performance in 11 European pediatric emergency departments. In this study, 2,185 infants ≤ 90 days old with fever without apparent source had several required tests—urinalysis, blood tests, and blood culture—as well as additional tests and treatment at the physician’s discretion. All infants were assessed with the Step-By-Step tool and two other tools that do not require lumbar puncture (Rochester criteria and Lab-score), and ultimately diagnosed by blood or cerebrospinal fluid (if collected) culture.
Eighty-seven (3.9%) infants had an IBI by culture. The Step-By-Step tool classified 1,194 (54.6%) as at risk for IBI, and performed well in ruling out IBI, with a negative predictive value (NPV) of 99.3%. This NPV was higher than that for Rochester criteria (98.3%) and Lab-score (98.1%), though pairwise statistical comparisons were not reported. Also, because of its sequential nature, 44.6% of the infants were classified as at risk based on appearance, age, or urinalysis only, without consulting blood biomarkers. Among the seven infants falsely identified to be at low risk for IBI, four were aged from 22 to 28 days.
For assessing the risk of IBI in infants ≤ 90 days old presenting with fever without apparent source, the Step-By-Step tool may be more practical than other tools in that it sequentially evaluates ill appearance, age, urinalysis, and then blood biomarkers. It classified 45.4% of infants as having low risk for IBI and had a high negative predictive value, making it useful for ruling out IBI while decreasing the need for additional assessments and empiric antibiotic therapy. However, for detection of non-invasive but potentially serious infections, only limited analyses were reported. Also, the authors point out that individualized approaches, rather than standardized assessment tools, are often used. Thus, a comparison of the Step-By-Step tool to “usual care” may be more useful than a comparison to other tools that may or may not be used. Finally, the fact that four of the seven infants who were falsely identified to be at low risk for IBI in the Step-by-Step approach were aged 22 to 28 days suggests that the Step-by-Step approach should be used with caution, if at all, in infants < 29 days old. Keeping these caveats in mind, this study advances our understanding on how to assess infants with fever without apparent source.
For more information see the Fever without apparent source in infants less than 3 months old topic in DynaMed Plus. DynaMed users click here.