The use of microbiological cultures for diagnosing bacterial infections in young febrile infants have substantial limitations, including false positive and false negative cultures, and non-ideal turn-around times. Analysis of host genomic expression patterns (“RNA biosignatures”) in response to the presence of specific pathogens, however, may provide an alternate and potentially improved diagnostic approach. This study was designed to define bacterial and non-bacterial RNA biosignatures to distinguish these infections in young febrile infants.

A total of 279 febrile infants and 19 healthy afebrile control infants aged 0-6 months (for a total of 298 samples) for microarray analysis. For analytic purposes, we classified patients into two groups, those with bacterial infections (n=89) and those with non-bacterial infections (n=190).144 of the samples were run on Illumina HT12 V4 R1 chips. Of these, there were 34 bacterial infections, 105 non-bacterial infections, and 5 healthy afebrile controls.154 of the samples were run on Illumina HT12 V4 R2 chips. Of these, there were 55 bacterial infections, 85 non-bacterial infections, and 14 healthy afebrile controls.

Bacterial infection (2021 ICD-10-CM code* = A49.9)

The data were normalized using average normalization in GenomeStudio’s gene expression module.

https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE64456