Procalcitonin and C-Reactive Protein Are Elevated at the Threshold of Cortical Electrocerebral Silence but Do Not Track Graded EEG Severity in Critically Ill Children: A Single-Centre Retrospective Cohort Study

Background: The relationship between routine inflammatory biomarkers and EEG severity in critically ill children, evaluated using visual (Synek) grading and spectral qEEG, remains unclear. The aim of our study was to examine the associations between procalcitonin (PCT), C-reactive protein (CRP) and EEG severity. Methods: Single-centre retrospective cohort of 97 critically ill children were analysed (PICU, Romania; November 2022–February 2026), each contributing one EEG with an inflammatory biomarker measured within ≤24 h. The primary analysis was the Spearman correlation of PCT and CRP with Synek grade, supported by comparisons of cortical electrocerebral silence with preserved cortical activity (Synek 1–4). The secondary analysis comprised five age-adjusted partial Spearman correlations between biomarkers and spectral qEEG metrics in the spectral cohort (n = 71), BH-corrected within family. Results: PCT and CRP were markedly higher in cortical silence than in preserved activity (Mann–Whitney p-value = 9.6 × 10−6 and 4.1 × 10−6; adjusted OR ≈ 2 per ln-unit). Both markers showed a positive rank association across Synek grades (PCT ρ = +0.49; CRP ρ = +0.41; both BH-adjusted p-value < 10−4), driven by the Synek-5 step: within grades 1–4 neither was graded (PCT ρ = +0.19; CRP ρ = +0.08; both p-value > 0.1). None of the five secondary correlations suggested a moderate-to-strong association (ρ −0.14 to +0.08; all BH-adjusted p-value ≥ 0.668). Conclusions: PCT and CRP were elevated at the threshold of cortical electrocerebral silence but did not provide a graded severity index within Synek grades 1–4, nor did they associate with spectral qEEG, including in the full 10–20 subset. The spectral analysis was underpowered for weak effects and excluded the Synek-5 subset where the primary signal was localised.