Combined short- and mid-latency SSEP amplitudes for early prediction of favorable neurological recovery after cardiac arrest.

Abstract

Early identification of patients with preserved neurological recovery after cardiac arrest remains challenging. We evaluated whether mid-latency somatosensory evoked potentials (SSEPs) provide additional prognostic information during the acute post-cardiac arrest period. In this retrospective cohort study, we included comatose adult cardiac arrest patients treated with targeted temperature management between 2009 and 2023 who underwent SSEP recording within 24 h after return of spontaneous circulation (ROSC). Short- and mid-latency SSEP amplitudes (N20, N20-P25, P25, P25-N30, and N30-P45) and latencies were quantified. Favorable neurological outcome was defined as a 6-month Cerebral Performance Category (CPC) score of 1-2. Discriminative performance was compared with other early prognostic variables obtained within the same time window, including neurological examination, brain computed tomography, electroencephalography, and neuron-specific enolase. Among 73 patients, 29 (40%) achieved favorable neurological outcome. Cortical SSEP amplitudes were consistently higher in patients with favorable outcomes. P25-N30 amplitude demonstrated strong discrimination (AUC 0.943, 95% CI 0.862-0.984), achieving 100% sensitivity (95% CI 88.1-100%) and 84.1% specificity (95% CI 69.9-93.4%). Its performance was comparable to that of the conventional N20-P25 amplitude (AUC 0.922, 95% CI 0.836-0.972; p = 0.417). The addition of non-SSEP predictors did not significantly improve overall discrimination. Quantitative assessment of the P25-N30 amplitude within 24 h after cardiac arrest was associated with favorable neurological outcome and may contribute to early multimodal prognostication.