Evaluating a Higher Temperature Threshold for the Emergency Department Workup of Young Febrile Infants

Background: Young infants who develop fever are at an increased risk for serious infection. Determining which infants are at highest risk for serious infection, and in need of hospitalization for intravenous antibiotics, can be particularly challenging. Since young infants can be difficult to examine and often look well despite having serious infection, most physicians will perform laboratory testing on these well-appearing infants once a certain temperature threshold is met. It is unclear, however, what temperature cut-off would be optimal to begin evaluating these infants, as some criteria use different thresholds.

Objective: We sought to determine the percentage of infants presenting to the Emergency Department (ED) with a temperature <38.2℃ who develop serious infection compared to those with higher temperatures.

Methods: We used a publicly-available dataset from the Pediatric Emergency Care Applied Research Network (PECARN). Patients were included if they were ≤60 days old, had a documented rectal temperature ≥38.0℃ in the ED or had a history of fever within 24 hours prior to presentation to the ED, and were being evaluated for serious infection. We used the same exclusion criteria as the original PECARN study but further excluded those who were ill-appearing (Yale Observation Score >10). Serious infections included any of the following: urinary tract infection (UTI), bacteremia, bacterial or herpes meningitis, bacterial pneumonia, or bacterial enteritis. Data were described using frequencies (percentages), and compared between groups using a chi-square test.

Results: Of the 4,619 eligible infants, 1,311 (28.4%) had a temperature <38.2℃, and 4,554 (98.6%) had at least one test for serious infection. While the majority of infants had urine, blood, and cerebrospinal fluid (CSF) cultures obtained, a minority had CSF sent for herpes testing, or a chest radiograph or stool culture obtained. Infants with temperatures <38.2℃ were significantly less likely to have a serious infection compared to those with higher temperatures (97 [7.5%] vs 365 [11.2%], p<0.001). Of the infants with temperature <38.2℃ who were tested, 67 (5.8%) had a urinary tract infection, 10 (0.8%) had bacteremia, 3 (0.4%) had bacterial meningitis, 3 (2.9%) had herpes meningitis, 17 (4.5%) had bacterial pneumonia, and 2 (4.8%) had bacterial enteritis.

Conclusion: In this study, we analyzed data from a previous large prospective multicenter dataset and found that infants with temperatures <38.2℃ were significantly less likely to have a serious infection than those with higher temperatures. Using an evaluation cutoff of 38.2℃, however, would likely miss a clinically important number of well-appearing infants with serious infections.