Added value of somato-sensory evoked potentials amplitude for prognostication after cardiac arrest

Abstract

AimsBilateral absence of cortical somato-sensory evoked potentials (SSEPs) robustly predicts poor outcome after cardiac arrest (CA), but it is uncertain if SSEP amplitudes provide additional information. Here, we examined the prognostic value of cortical SSEP amplitude in comparison with other known outcome predictors. MethodsWe retrospectively determined SSEP amplitudes in a prospective CA registry, identified an amplitude cut-off for worst Cerebral Performance Category (CPC) within three months, and examined correlations of SSEP amplitude with pupillary light reflex (PLR), myoclonus, peak serum neuron specific enolase (NSE), and 24–36 h and 36–72 h EEG (reactivity, epileptiform features). ResultsAmong 158 patients, 54% awoke. Amplitudes correlated with EEG findings, present PLR, myoclonus, NSE. A cut-off for cortical SSEP ≤ 0.41 μV was 100% specific for poor outcome (95% CI: 96–100%); sensitivity increased marginally vs. SSEPs absence [47% (35–59%) vs 46% (34–58%)] for CPC 4–5. Adding SSEPs ≤0.41 μV to a multimodal prognostic model including EEG, clinical features, and NSE improved prediction for mortality, but not for CPC 3–5 at three months. No statistical correlation between amplitudes and good outcome was observed. SSEP amplitudes correlated inversely with CPC at three months in the overall cohort (r = −0.332; p < 0.0001) but not in the subgroup with present SSEPs (r = −0.102; p = 0.256). ConclusionDecreased SSEPs amplitudes are associated with poor outcome after cardiac arrest; however, adding this to a multimodal prognostic approach including EEG, clinical and blood biomarkers, improves slightly prediction of mortality, but not of poor or good outcome.