Occult Serious Bacterial Infection in Infants Younger Than 60 to 90 Days With Bronchiolitis: A Systematic Review
Ralston S, Hill V, Waters A
Arch Pediatr Adolesc Med. 2011;165:951-956
Study Summary
Improved vaccines and aggressive prenatal screening of mothers have reduced the incidence of serious bacterial infection in young children. Evidence is also building that infants and young children with viral respiratory infections have an even lower incidence of serious bacterial infection than infants without viral respiratory infections. This study attempted to assess the risk of serious bacterial infection in children with bronchiolitis by combining data from other studies. For this study, serious bacterial infection was defined as bacterial meningitis, bacteremia, or urinary tract infection.
Study children were all younger than 90 days of age. By conducting database searches for key terms, the investigators initially identified 14 studies for inclusion. Review of those studies identified an additional 4 studies, for a total of 18 studies that met study criteria for final consideration. Eleven of the studies allowed the investigators to calculate age-specific frequencies of serious bacterial infection. Six of those 11 studies were retrospective, and 6 studies used a clinical definition of bronchiolitis rather than a specific, positive RSV test. The studies were published between 1997 and 2008. When looking at frequency of urinary tract infection, the combined estimate of the prevalence was 3.3% with a 95% confidence interval of 1.9%-5.7%. The investigators did not conduct a formal analysis of combined data for the rates of bacteremia, but 5 children (of a total of 1749 children) had bacteremia, corresponding to a unadjusted frequency of 0.3% (or roughly 3 infants with bacteremia per 1000 studied). Not a single case of meningitis was found in the studies included in the analyses, making it impossible to estimate frequencies of meningitis in children with bronchiolitis.
Viewpoint
It has been fascinating to follow the evolving data about serious bacterial infections and febrile infants with viral illnesses. I previously reviewed some of the individual studies used in this meta-analysis for Medscape. Although the data were drawn from studies that varied in many aspects, the overall approach was the same for the most part and allows for rough estimations of urinary tract infection frequency. The consensus of the data suggests that if one wanted to limit evaluation for serious bacterial infection to urine cultures among children with clinically diagnosed bronchiolitis, the data would support that approach. The investigators are cautious to point out that the same assumptions should not be extended to infants less than 30 days of age because data were insufficient for analyses for these very young infants.
Abstract