Common measure missed many serious infections in febrile infants
by Molly Walker, Staff Writer, MedPage Today September 11, 2017

Action Points
• Complete blood count was a poor predictor of serious bacterial infections, such as bacteremia or bacterial meningitis, in very young infants with a fever.
• Note that a minority of infants with these serious types of infections actually had abnormal white blood cell counts.
Complete blood count was a poor predictor of serious bacterial infections, such as bacteremia or bacterial meningitis, in very young infants with a fever, a secondary analysis of a large cohort study found.
Individual parameters of the complete blood count, such as white blood cell (WBC) count, absolute neutrophil count, and platelet count, could not identify an invasive bacterial infection in infants under 60 days old, reported Andrea T. Cruz, MD, of Baylor College of Medicine in Houston, and colleagues.
Moreover, a minority of infants with these serious types of infections actually had abnormal white blood cell counts, the authors wrote in JAMA Pediatrics.
They noted that the complete blood count (CBC) is the most commonly obtained test for these extremely young infants who present to the emergency department with fevers. But, they added, there is some “diagnostic uncertainty” about detecting the presence of these invasive bacterial infections, including “the lack of reliable physical examination findings and nonspecific symptoms.”
Cruz’s group said that while “several algorithms” can help risk stratify young febrile infants for these serious bacterial infections, the introduction of pneumococcal conjugate vaccines, as well as screening and treating pregnant women for group B streptococcus means that invasive bacterial infections are less common in these “well-appearing” younger febrile infants, which makes “risk stratification even more challenging.”
They examined a cohort of infants from an existing study looking at 26 emergency departments in the Pediatric Emergency Care Applied Research Network from 2008 to 2013. The original study examined 4,795 prospectively enrolled infants, to evaluate the association of host gene expression patterns with bacterial infections. This study included 4,313 of those infants, who were previously healthy, full-term infants 2 months of age (60 days) and younger, presenting with a fever (≥38 ºC) and who had blood cultures obtained.
Overall, 97 infants (2.2%, 95% CI 1.8% to 2.7%) had invasive bacterial infections, 73 had isolated bacteremia (1.7%, 95% CI 1.4% to 2.1%), and 24 infants had meningitis (0.6%, 95% CI 0.4% to 0.8%). Of the infants with meningitis, 11 also had documented bacteremia. A higher portion of infants less than 28 days old had invasive bacterial infections (4.3%, 95% CI 3.2% to 5.3%) compared to infants ages 29 to 60 days (1.4%, 95% CI 1.0% to 1.8%).
However, the authors noted that “no CBC parameters reliably distinguished between infants with or without IBIs,” with low sensitivity and high negative predictive values at commonly used thresholds. They added that children with these infections had higher WBC and absolute neutrophil counts (ANC), and lower platelet counts, “there was not a threshold for these parameters at which IBI could be reliably predicted.”
One-third of infants with invasive bacterial infections would be missed, “even at an optimal ANC threshold” (4,100 cells/μL), while 63% of these infections in infants would be missed using “widely accepted normal ranges” for WBC counts (5,000 to 14,900 cells/μL).
An accompanying editorial by Matthew Mischler, MD, of University of Illinois College of Medicine in Peoria, and colleagues, said that the authors should be “commended for such a large sample size” in what has been a difficult to study population. They added that this research supports an alternative method of risk stratification for febrile infants, such as the “stepwise” method used in one study, replacing the “dogma” of the EBC with C-reactive protein, procalcitonin and urine dipstick.
“Newer markers are available that improve the clinician’s diagnostic acumen and enhance our ability to properly care for this challenging population,” the editorialists wrote. “Hospital administrators and policy makers should focus on making C-reactive protein and procalcitonin increasingly available with rapid turnaround time in any center that cares for infants in this age group.”
Study limitations included the fact that it was a convenience sample of febrile infants, and thus the results may not be generalizable to community emergency departments. Also, critically ill-appearing infants were not included in this study, which might have led to spectrum bias.

This study was supported by the Health Resources and Services Administration, Emergency Services for Children and by the Eunice Kennedy Shriver Institute of Child Health and Human Development, the Health Resources and Services Administration, Maternal and Child Health Bureau, and Emergency Medical Services for Children Network Development Demonstration Program.
Cruz disclosed no relevant relationships with industry. One co-author disclosed support from Abbvie, Janssen, Regeneron, and Pfizer.
Mischler and co-authors disclosed no relevant relationships with industry.
• Reviewed by Robert Jasmer, MD Associate Clinical Professor of Medicine, University of California, San Francisco and Dorothy Caputo, MA, BSN, RN, Nurse Planner
last updated 09.12.2017
• Primary Source
JAMA Pediatrics
Source Reference: Cruz AT, et al “Accuracy of complete blood cell counts to identify febrile infants 60 days or younger with invasive bacterial infections” JAMA Pediatr 2017; DOI:10.1001/jamapediatrics.2017.2927.
• Secondary Source
JAMA Pediatrics
Source Reference: Mischler M, et al “White blood cell count in the evaluation of the febrile infant: Time to revisit the dogma?” JAMA Pediatr 2017; DOI:10.1001/jamapediatrics.2017.2796.