Characteristics of Afebrile Infants ≤60 Days of Age With Invasive Bacterial Infections

We conducted a secondary analysis of a cross-sectional study of infants ≤60 days old with IBI presenting to the emergency departments (EDs) of 11 geographically diverse children’s hospitals from July 1, 2011, to June 30, 2016.¹  We identified positive culture results from blood and/or cerebrospinal fluid (CSF) through microbiology or electronic medical record queries. Infants were included in the parent study if they had growth of an a priori determined pathogenic bacteria not treated as a contaminant, as previously described.¹  For this secondary analysis, we selected afebrile infants, defined by absence of a documented temperature ≥38°C at home, at the referring clinic, or in the ED. This study was approved by the institutional review boards at all participating sites with permission for data sharing.

Site investigators abstracted the following data from the electronic medical record into a standardized Research Electronic Data Capture⁸  tool hosted at Yale University: demographics; past medical history; symptoms before presentation (eg, irritability, lethargy); ED triage vital signs; physical examination findings, including evidence of clinically apparent infection on examination (eg, cellulitis, omphalitis); laboratory test results; and outcomes.

We classified ED triage vital signs as abnormal using the 99th percentile for age, with tachycardia defined as a heart rate ≥181 beats per minute and tachypnea defined as a respiratory rate ≥66 breaths per minute.^9,10  Hypothermia was defined as temperature <36°C.¹¹  Ill appearance was defined as having one of the following terms documented in the ED physical examination: ill appearing, toxic, limp, unresponsive, gray, cyanotic, apnea, weak cry, poorly perfused, grunting, listless, lethargic, or irritable.¹²  We defined an abnormal white blood cell (WBC) count as <5000 or >15 000 cells per μL, an abnormal absolute band count as >1500 cells per μL, and a positive urinalysis as a urinalysis or urine dipstick positive for leukocyte esterase (not including trace), nitrites, or >5 white blood cells per high-power field. These thresholds were chosen because they are used by some clinical decision rules to risk stratify febrile young infants.^13,14 

Bacteremia was defined by growth of pathogenic bacteria from blood culture alone. Bacterial meningitis was defined as either growth of pathogenic bacteria from a CSF culture or bacteremia with CSF pleocytosis (CSF WBC count ≥16 cells per mm³ for age ≤28 days and ≥10 cells per mm³ for age 29–60 days)¹⁵  for infants who received antibiotics before obtaining CSF cultures.¹⁶  Infants who had both bacteremia and meningitis were classified as having meningitis. We examined two adverse outcomes within 30 days of the ED visit: death and neurologic disability (a composite outcome defined as hearing loss, seizures, or any abnormal cranial imaging). These outcomes were identified through medical record review and were chosen because they were evaluated in a previous cohort of febrile infants.¹⁷ 

We described the clinical features, laboratory values, and outcomes of afebrile infants with IBI, with separate analyses of infants with bacteremia without meningitis and infants with bacterial meningitis. Proportions were compared using Fisher’s exact test. Data were analyzed using Stata version 15.0 (Stata Corp, College Station, TX). All tests were 2-tailed, with P < .05 considered statistically significant.

Of 440 infants with IBI, 78 (18%) were afebrile. Among afebrile infants with IBI, 62 (79%) had bacteremia without meningitis and 16 (20%) had bacterial meningitis (10 with concomitant bacteremia).

The most common pathogens were Streptococcus agalactiae (35%), Escherichia coli (16%), and Staphylococcus aureus (16%) (Table 1). Overall, 5 infants (6%) died; all had bacteremia. Eight infants (10%) experienced neurologic disability, 5 with bacteremia and 3 with bacterial meningitis.

Most infants (59 of 78; 76%) had documentation of either poor feeding, irritability, or seizure (Table 2). Nearly half (36 of 78; 46%) of all afebrile infants had at least one vital sign abnormality on ED triage (Table 2). Hypothermia was present in 17 of 77 (22%) infants. Of the 5 infants who died, 4 had at least 1 vital sign abnormality and 1 had missing vital sign data. A total of 52 (67%) infants had at least 1 of the following physical examination findings documented: ill appearance, full or depressed fontanelle, increased work of breathing, or clinically apparent infection. More than three-quarters (60 of 78; 77%) had either an abnormal vital sign or physical examination finding. Forty-three (55%) infants had abnormalities of the WBC count, absolute band count, or urinalysis. Presence of a vital sign, physical examination, or laboratory abnormality occurred in 71 of 78 infants (sensitivity 91% [95% confidence interval 82%–96%]).

Of the 62 infants with bacteremia, 47 (76%) underwent CSF testing. Of the 15 infants without CSF testing, there was no documentation of missed meningitis within 30 days of the index ED visit.

In this multicenter cross-sectional study, we describe the clinical and laboratory characteristics of afebrile young infants with IBI. Only half had abnormalities in the WBC count, absolute band count, or urinalysis. However, the majority of afebrile young infants with an IBI had vital sign, examination, or laboratory abnormalities.

There are few studies in which afebrile infants with IBI are described. Miller et al⁶  found that afebrile infants who undergo evaluation for IBI often present with nonspecific symptoms. In their cohort of 217 afebrile infants, only 2 had IBI (1 with bacteremia and 1 with bacteremic urinary tract infection). Okike et al¹⁸  described 35 infants ≤90 days of age with bacterial meningitis who presented without an abnormal temperature or neurologic features and found that 49% presented with lethargy and 40% presented with irritability.

Our study builds on previous work by describing vital sign and examination abnormalities in afebrile infants with IBI. Hypothermia was found in 22% of our cohort, and in previous studies, authors have reported IBI rates from 3% to 8% in hypothermic infants.¹¹  Tachycardia and tachypnea, present in 27% and 11% of our cohort, respectively, have also been identified as risk factors for serious illness and are included as criteria for systemic inflammatory response syndrome.¹⁹  Ill appearance, present in 47% of our cohort, was the most common examination abnormality noted and more frequent than the 30% reported in febrile infants with ill appearance, suggesting that clinical appearance may be helpful in evaluating afebrile infants.¹⁷  Studies in febrile infants have suggested that clinical appearance is not reliable in identifying IBI,²⁰  but this has not been studied in afebrile infants. Further study should be undertaken to investigate the sensitivity and specificity of these vital sign and examination abnormalities in detection of IBI in afebrile infants.

Once the decision is made to perform an evaluation for IBI, there are no standard criteria guiding interpretation and management for afebrile infants. WBC count, absolute band count, and urinalysis are used for risk stratifying febrile infants,²  even in the office setting,²¹  but abnormalities were present in only 55% of afebrile infants with IBI in our cohort. One-quarter of our cohort had an abnormal urinalysis, which is less than the >50% rate reported in some studies of febrile infants with IBI.^22,23  This suggests that if a laboratory evaluation is pursued in an afebrile infant, a normal complete blood count and urinalysis result should not reassure against the presence of an IBI. However, when these laboratory abnormalities are considered with vital sign and physical examination abnormalities, the overall sensitivity is 91%.

The mortality rate in our cohort (6%) is higher than the 1.5% mortality reported in febrile infants with IBI,¹⁷  suggesting the need for risk stratification criteria to help guide evaluation and management for this population. The present work provides groundwork for future studies of afebrile infants. Newer clinical decision rules for febrile young infants, such as the Step-by-Step approach³  and the Pediatric Emergency Care Applied Research Network clinical prediction rule,⁴  have incorporated biomarkers, such as procalcitonin, which may be helpful to include in future studies evaluating afebrile infants with IBI.

This study has several limitations. First, we did not include a population of afebrile infants without IBI for comparison, and therefore we cannot calculate the prevalence of IBI or the specificity of individual signs and symptoms in this sample. Second, we focused only on IBI, so our results do not apply to afebrile infants with urinary tract infection without IBI or infants with herpes simplex virus infection. Third, data were obtained via chart review, which relies on physician documentation for accuracy. It is possible that symptoms and examination findings may have been present but not documented or that examination findings, such as ill appearance, may have been documented after laboratory test results. We also did not have chief complaint information and cannot conclude what specific factors prompted clinicians to evaluate for IBI. Fourth, absolute band count and urinalysis were not available in a subset of infants. Additionally, few infants had other inflammatory markers available, such as C-reactive protein and procalcitonin (only measured in 24% and 3% of infants, respectively), and thus we cannot evaluate the contribution of those laboratory values to the evaluation of afebrile infants with IBI.

The majority of afebrile young infants with IBI had vital sign, examination, or laboratory abnormalities. Future studies should evaluate the predictive ability of these and other criteria in afebrile infants undergoing evaluation for IBI.

The following collaborators in the Febrile Young Infant Research Collaborative acquired data for this study and/or contributed to the parent study: Elizabeth R. Alpern, MD, MSCE (Ann and Robert H. Lurie Children’s Hospital of Chicago, Chicago, IL); Whitney L. Browning, MD (Vanderbilt University School of Medicine, Nashville, TN); Elana A. Feldman, MD (Lucile Packard Children’s Hospital Stanford, Palo Alto, CA); Catherine E. Lumb, MD (University of Alabama at Birmingham, Birmingham, AL); Russell J. McCulloh, MD (Children’s Mercy Hospital, Kansas City, MO); Christine E. Mitchell, BSN (Children’s Hospital of Philadelphia, Philadelphia, PA); Samir S. Shah, MD, MSCE (University of Cincinnati College of Medicine, Cincinnati, OH); Sarah J. Shin, BSN (Children’s Hospital of Philadelphia, Philadelphia, PA); and Derek J. Williams, MD, MPH (Vanderbilt University School of Medicine, Nashville, TN).