Clinical prediction rule to identify febrile infants 60 days and younger at low risk for serious bacterial infections

Abstract

Kupperman et al. recently published results of a prospective multi-centre observational study that created and validated a clinical prediction rule identifying febrile infants ≤60 days old at low risk of serious bacterial infection (SBI), defined as cultureconfirmed bacteraemia, urinary tract infection or meningitis. A sample of 908 infants ≤60 days old presenting with fever >38 C across 22 emergency departments in the USA were enrolled to derive the rule, and it was validated on a further 913 infants. Infants were excluded if they appeared critically unwell, had received antibiotics in the preceding 48 h, had pre-existing medical conditions or indwelling devices, a history of prematurity <36 weeks or soft tissue infections. The combination of absolute neutrophil count <4 × 10/L, procalcitonin <1.7 ng/mL and negative urinalysis was shown to most accurately determine low risk of SBI. In the group identified as low risk by the rule, 99.8% did not have an SBI. No cases of meningitis were missed. This study highlights the potential utility of procalcitonin, and has validated a purely laboratory-based scoring system to determine which febrile infants are at low risk of an SBI. This has potential to reduce the number of infants who receive unnecessary lumbar punctures and antibiotics.