O2051: Effect of Co-morbid Viral Illness on the Utility of Procalcitonin as a Predictor of Invasive Bacterial Infection in Infants 8-60 Days of Age

Background: Febrile infants under 60 days-old undergo an extensive evaluation for infection. Recent AAP guidelines recommend using procalcitonin (PCT) in febrile patients 8-60 days-old to aid in predicting the presence of an invasive bacterial infection (IBI). Viruses are theorized to induce cytokines that inhibit PCT production, potentially altering PCT levels in the setting of an IBI. Few studies focus on PCT levels in this population, and none consider the effect of a co-morbid confirmed viral infection (CVI). The objectives are: 1. To determine if the presence of co-morbid CVI alters the mean PCT level of infants 8-60 days-old with IBI. 2. To compare the sensitivity and specificity of PCT in detection of an IBI in infants with and without a co-morbid CVI.

Methods: A retrospective cohort study examined infants 8-60 days-old presenting to a tertiary children’s hospital with fever between 2018-2020. Patients were identified via ICD-10 codes for fever or IBI (bacteremia, meningitis, urinary tract infection). Exclusion criteria included prematurity ( < 37 weeks), previous admission for infection or to intensive care, current or prior antibiotic use, or complex care needs. Data collected included PCT level, bacterial cultures, and viral PCR test results. Patients with CVI alone were excluded from further analysis. The Wilcoxon rank sum test compared PCT values between CVI+IBI, IBI only, and neither CVI nor IBI groups. Area under the curve (AUC) analysis was performed to calculate the sensitivity and specificity of PCT in IBI detection for infants with and without co-morbid CVI.

Results: 506 patients met inclusion criteria. CVI+IBI presented in 47 (9%) patients, IBI only in 157 (31%), and neither CVI nor IBI in 302 (60%). Significant differences in mean PCT levels were detected between infants with IBI only (3.5ng/mL) vs. those with IBI+CVI (1.8ng/mL [p=0.01]), and between patients with IBI+CVI (1.8ng/mL) vs. those with neither IBI nor CVI (0.5ng/mL [p < 0.01]) (Table 1). AUC analysis using current AAP guidelines with PCT cut-off of 0.5ng/mL demonstrated decreased sensitivity of PCT in IBI detection (from 0.61 to 0.43) when a co-morbid CVI was present, while specificity remained unchanged at 0.94 (Table 2).

Conclusion: In febrile infants aged 8-60 days, the presence of a co-morbid CVI significantly decreased PCT levels when compared to infants with IBI alone, but still resulted in a significantly higher PCT level when compared to infants with neither IBI nor CVI. The mean PCT was >0.5 in those with IBI+CVI, indicating that infants would likely not be missed under current AAP guidelines, however the sensitivity of PCT as a marker for IBI in this group was decreased. Practitioners should consider the impact of co-morbid CVI when using PCT levels to aid in the prediction of IBI in febrile neonates.