536 - Diagnostic Performance of Biomarker Rate of Increase in Febrile Young Infants With Fever of Very Recent Onset

Publication Number: 2521.536

Laura de Tovar, Universitat Autònoma de Barcelona, Sabadell, Catalonia, Spain; Rosario López, Hospital la Paz Madrid, Madrid, Madrid, Spain; Arístides Rivas, Gregorio Marañón University Hospital, Alcobendas, Madrid, Spain; Elizabeth Domingo, Hospital Santa Creu i Sant Pau, Barcelona, Catalonia, Spain; Inés Perdomo Delgado, Sociedad Española de Urgencias Pediátricas, Las Palmas de Gran Canaria, Canarias, Spain; Julio Baena, Consorci Sanitari del Maresme, Sant Antoni de Vilamajor, Catalonia, Spain; Manuel Gijón, Hospital Universitario 12 de Octubre, MADRID, Madrid, Spain; Mariano Plana, Hospital de Barbastro, Binéfar, Aragon, Spain; M. Natali Campo Fernandez, Rio Hortega University Hospital, Valladolid, Castilla y Leon, Spain; Raquel Porto, Hospital Puerta de Hierro, Madrid, Madrid, Spain; Susanna Hernández-Bou, Hospital Sant Joan de Déu Barcelona, Barcelona, Catalonia, Spain; Cristina Ibarzo Port, Hospital Parc Taulí, Sabadell, Catalonia, Spain; María Suárez-Bustamante, Hospital Infantil Universitario Niño Jesús, Madrid, Madrid, Spain; Borja Gómez, Hospital Universitario Cruces, Barakaldo, Pais Vasco, Spain; Jose Antonio Alonso Cadenas, Hospital Infantil Universitario Niño jesús, Madrid, Madrid, Spain; Roberto Velasco, Hospital Universitari Parc Taulí, SABADELL, Catalonia, Spain

Background: In young infants, biomarkers and clinical decision rules have limited sensitivity (Sn) in cases of fever of very recent onset, so repeated testing is often required in the first hours (h). Applying the same decision rule cutoffs to repeated biomarkers in patients with only a few hours of fever may increase false positives. In adults, the diagnostic performance of the rate of rise of C-reactive protein (CRP) has been investigated, but data are lacking for pediatric patients.

Objective: To evaluate the diagnostic accuracy of the rate of rise in procalcitonin (PCT), CRP, leukocyte count, and absolute neutrophil count (ANC) for detecting invasive bacterial infection (IBI) in well-appearing infants ≤90 days old with fever without a source (FWS) lasting < 6 h, and more specifically in patients with < 2 h of fever.

Design/Methods: Study conducted in 7 pediatric emergency departments. Infants ≤90 days with FWS < 6 h and at least 2 measurements of any studied biomarker within 24 h were included. Data were collected retrospectively (2011–2024). A subanalysis was performed in infants with < 2 h of fever, as the biomarkers’ performance in this group has been proved as poorer.
IBI was defined as isolation of a pathogenic bacterium from blood or cerebrospinal fluid. Rate of increase was calculated as [(Biomarker₂ − Biomarker₁)/hours between samples].
Patient inclusion continues; results are provisional.

Results: We included 515 infants, with a median fever duration of 2 h (IQR 1–3). Of these, 21 (4.1%; CI95% 2.7-6.2) were diagnosed with an IBI. The median interval between biomarker determinations was 12 hours. Figure 1 shows the rate of increase in each biomarker by the presence or absence of IBI. Table 1 presents the area under the curve for each biomarker’s rate of rise in all patients and among infants with fever < 2 h.
Among the 231 (34.7%) infants with < 2 h of fever, 11 (4.8%) were diagnosed with an IBI. Seven of them (63.6%) showed normal initial PCT, CRP, and ANC values, including 5 of the 6 infections by S. agalactiae. In patients with < 2 h of fever, applying Step-by-Step cutoffs to the second measurement yielded a Sn of 81.8% (CI95% 52.3-94.9) and a specificity (Sp) of 74.1% (CI95% 67.9-79.4) for IBI diagnosis. Meanwhile, a PCT increase rate ≥0.0325 ng/mL/h achieved 80% (CI95% 49.0-94.3) Sn and 91.2% (CI95% 86.5-94.3) Sp, while CRP increase rate >0.454 mg/L/h reached 90.0% (CI95% 59.6-98.2) Sn and 76.2% (CI95% 70.0-81.4) Sp.

Conclusion(s): In febrile infants ≤90 days old, especially when fever duration is very short ( < 2 h), the rate of increase of PCT and CRP might be an useful biomarker to diagnose IBI.

Figure 1. Values of the rate of increase for each biomarker based on the presence or absence of an invasive bacterial infection (IBI).
Figure 1 6 hours.pdf(ANC: absolute neutrophil count; CRP: C-reactive protein; PCT: procalcitonin; WBC: white blood cell count)

Table 1. Area under the curve for each biomarker’s rate of rise (∆) for the diagnosis of invasive bacterial infection (IBI) based on the duration of fever.
Table 1.pdf(ANC: absolute neutrophil count; CRP: C reactive protein; PCT: procalcitonin; WBC: white blood cell count)