Human Rhinovirus Detection by PCR in Febrile Infants and Risk of Concomitant Bacterial Infection

BACKGROUND: Studies have shown that well-appearing febrile infants (FI) with viral respiratory infections have a reduced risk of bacterial infections (BI; urinary tract infection, bloodstream infection, meningitis). Respiratory testing by PCR allows detection of human rhinovirus (HRV), but few data exist on the risk of concomitant BI in HRV-positive FI. METHODS: We identified well-appearing FI 1–90 days old within Intermountain Healthcare evaluated in the ED or inpatient setting (IP) with viral respiratory testing by PCR (RVPCR) from August 2007 to August 2016. Respiratory viruses detected by RVPCR included: adenovirus, coronavirus, human metapneumovirus, influenza A/B, parainfluenza 1–4, RSV and HRV. We used relative risk (RR) to compare the risk of BI for infants with HRV vs. non-HRV viruses detected. Similarly, we used RR to compare risk of UTI and invasive bacterial infection (IBI; bacteremia and meningitis) for infants with HRV detected compared with those who were virus negative. RESULTS: 10,964 FI were evaluated in the ED/IP during the study period. 4037 (37%) had RVPCR and were included. 2212 (55%) FI were positive for a respiratory virus and 73% were 29–90 days old. HRV was detected alone in 1392 (34%) and non-HRV viruses were detected in 820 (20%). The overall frequency of BI in the cohort was 9.5%. FI with HRV were more likely to have BI when compared with those with non-HRV viruses [7.8% vs 3.7% P < 0.0001; RR 2.12 (95% CI; 1.43–3.15)]. When compared with virus-negative infants, HRV detection in infants 1–28 days did not decrease the risk for UTI [RR 0.87 (95% CI 0.58–1.29)]; risk of IBI was statistically decreased [RR 0.41 (95% CI 0.19–0.88)] but with wide CI approaching 1 suggesting that this may not be clinically meaningful. Similarly, UTI risk in infants 29–90 days was statistically lower with HRV detection [RR 0.78 (95% CI 0.65–0.95)], but unlikely to be clinically important. For infants 29–90 days with HRV, risk of IBI was statistically decreased [RR 0.52 (95% CI 0.34–0.80)] with possible clinical relevance. CONCLUSION: HRV detection was common in young febrile infants. Infants with HRV were at higher risk of BI than infants with non-HRV infection. Detection of HRV did not meaningfully change risk for UTI at any age or meaningfully impact risk of IBI in infants 1–28 days. HRV detection may be associated with a decreased risk for IBI in infants 29–90 days. DISCLOSURES: A. J. Blaschke, BioFire Diagnostics LLC: Collaborator, Have intellectual property in BioFire Diagnostics through the University of Utah and Investigator, Licensing agreement or royalty and Research support; J. Daly, Biofire: Grant Investigator, Grant recipient; C. L. Byington, BioFire: Collaborator and Grant Investigator, Licensing agreement or royalty and Research grant