Lipocalin-2 is a Sensitive and Specific Marker of Bacterial Infection in Children

Jethro Herberg; Honglei Huang; Marie L. Thezenas; Victoria Janes; Michael Carter; Stuart Gormley; Melisa S. Hamilton; Benedikt Kessler; Michael Levin; Climent Casals-Pascual

Introduction. Bacterial infection remains the major infectious cause of death in children (1) . Identifying the small number of children with serious bacterial infection (SBI -microbiologically confirmed bacterial infection with systemic symptoms) amongst the majority with viral infections remains difficult, as clinical features are unreliable. Laboratory markers including white cell count, neutrophil proportion and C-reactive protein (CRP) have the strongest predictive value in children with clinically obvious SBI, rather than in the more numerous group with clinically uncertain infection status (2). Blood cultures, considered the gold standard for diagnosing SBI, are unhelpful for early antibiotic decisions as results are not immediately available, and are often negative if pre-admission antibiotics have been administered. In the UK there has been a 30% increase in hospital admission for children with probable infection (3), and the majority of children admitted with fever are discharged without a pathogen diagnosis. Many children with self-limiting viral illnesses are admitted to hospital and receive unnecessary antibiotics whilst awaiting the results of cultures, contributing Table 1 . Demographic and clinical data of recruited subjects. Table footnote: NS -not significant (P>0.05); IQR -interquartile range; N/A -not applicable; PICUpaediatric intensive care unit; WBC -white cell count; a Not all patients had samples collected for virological investigation; b Some patients had both respiratory and non-respiratory viruses detected. The identified bacterium is listed for patients with Definite Bacterial infection. Two age comparisons were significant: the Probable Bacterial median age was higher than the Definite Viral group (p<0.01), and the Definite Viral group were younger than the Controls (p<0.05). The proportion of children with male sex or with Caucasian ethnicity (mode ethnicity) was not significantly different between groups. Severity of illness, as indicated by the number of children requiring PICU admission, inotrope use and ventilation, was highest in the Definite Bacterial group and lowest in the Definite Viral group. Deaths in each group were not significantly different.