The Rochester criteria help physicians triage febrile children into high- and low-risk categories, depending on how likely they are to have a serious bacterial infection. The extent to which children with viral illness have a concomitant serious bacterial infection is not known, nor have any studies shown whether identifying a viral illness in febrile children reduces their risk of having a serious bacterial infection. Byington and colleagues examined whether children with viral illnesses are less likely to develop serious bacterial infections than children without viral illness.

This prospective study enrolled infants one to 90 days of age with temperatures of at least 38°C (100.4°F), testing them for respiratory viruses, including respiratory syncytial virus (RSV). All infants underwent a complete septic work-up. Infants were assigned Rochester classifications (high-risk or low-risk) and were subsequently classified as having serious bacterial infection based on bacterial culture results.

Viral testing was performed in1,385 infants, 491 (35 percent) of whom had one or more viral infections. Of the 1,385 participants, 922 (67 percent) were classified as high-risk, and 456 (33 percent) were classified as low-risk; a determination could not be made in the remaining seven infants. Of the 1,385 infants, 131 (9.5 percent) had a serious bacterial infection. Of the 491 virally infected children, 21 (4.2 percent) had a serious bacterial infection compared with 110 (12.3 percent) of the 894 children who were not virally infected.

Bacterial meningitis occurred in six children without viruses and in none of the children with viral infections. When these findings were combined with the Rochester classifications, high-risk infants with viral infections were significantly less likely to have serious bacterial infections than high-risk infants without viral infections. The occurrence of serious bacterial infections was not significantly different in low-risk, virus-positive infants versus low-risk, virus-negative infants. High-risk virus-negative infants had a 13.67 greater occurrence of serious bacterial infections than low-risk, virus-positive infants.

RSV and enteroviruses were the most commonly identified viruses. The authors point out that technology to identify these viruses is readily available. Thus, viral testing could be used as an additional tool to triage febrile infants. Even high-risk infants who are virally infected are less likely to have bacteremia, urinary tract infection, or soft-tissue infection, whereas high-risk children without viral infection are 3.5 times more likely to have a serious bacterial infection.