[Skip to Content]
[Skip to Content Landing]
Views 366
Citations 0
Comment & Response
August 19, 2019

Efficacy of a Clinical Prediction Rule to Identify Febrile Young Infants at Low Risk for Serious Bacterial Infections—Reply

Author Affiliations
  • 1School of Medicine, Departments of Emergency Medicine and Pediatrics, University of California, Davis, Sacramento
  • 2Department of Emergency Medicine, University of Michigan School of Medicine, Ann Arbor
  • 3Center for Vaccines and Immunity, Division of Pediatric Infectious Diseases, Nationwide Children's Hospital, The Ohio State University School of Medicine, Columbus
JAMA Pediatr. 2019;173(10):998-999. doi:10.1001/jamapediatrics.2019.2656

In Reply We thank Burstein and Papenburg, Verd and Moll, and Bonadio for the letters pertaining to our study1 in which we derived and validated a prediction rule for identifying febrile infants 60 days and younger with serious bacterial infections (SBI). Burstein and Papenburg question the use of the less widely available procalcitonin (PCT) rather than the C-reactive protein. They also suggest considering results of viral multiplex testing in the prediction rule. We chose PCT because of its superiority as a screening biomarker compared with C-reactive protein for invasive bacterial infections (bacteremia and/or bacterial meningitis) in young febrile infants.2,3 It is widely available in Europe and its use is appropriately increasing in the United States. We did not assess C-reactive protein because there is insufficient biological plausibility for added benefit vs PCT as a biomarker for invasive bacterial infections, and the urinalysis is the best screen for urinary tract infections. The volume of blood obtained was also a limiting factor after drawing separate samples for routine laboratory tests/cultures, RNA expression profiles,4 and PCT. Although we (and others) have documented the decreased risk of SBI in viral-positive febrile infants,5 at the time of our study, the cost of viral multiplex testing was prohibitive and turnaround times insufficiently rapid to affect emergency department decision-making. We agree that further study of the marginal value of adding viral infection data to febrile infant prediction rules would be useful. Fortunately, we have ongoing studies in which we are prospectively enrolling more febrile infants and are collecting specimens for PCT and multiplex viral testing in addition to genomic samples.

×