Do changes in SSEP amplitude over time predict the outcome of comatose survivors of cardiac arrest?

Aim: To assess if the amplitude of the N20 wave (N20Amp) of somatosensory evoked potentials (SSEPs) changes between 12–24 h and 72 h from the return of spontaneous circulation (ROSC) after cardiac arrest and if an N20Amp decrease predicts poor neurological outcome (CPC 3–5) at six months. Setting: Retrospective analysis of the ProNeCA multicentre prognostication study dataset. (NCT03849911). Methods: In adult comatose cardiac arrest survivors whose SSEPs were recorded at both 12–24 h and 72 h after ROSC, we measured the median N20Amp at each timepoint and the individual change in N20Amp across the two timepoints. We identified their cutoffs for predicting poor outcome with 0% false positive rate (FPR) and compared their sensitivities. Results: We included 236 patients. The median [IQR] N20Amp increased from 1.90 [0.78–4.22] µV to 2.86 [1.52–5.10] µV between 12–24 h and 72 h (p = 0.0019). The N20Amp cutoff for 0% FPR increased from 0.6 µV at 12–24 h to 1.23 µV at 72 h, and its sensitivity increased from 56[48–64]% to 71[63–77]%. Between 12–24 h and 72 h, an N20Amp decrease > 53% predicted poor outcome with 0[0–5]% FPR and 26[19–35]% sensitivity. Its combination with an N20Amp < 1.23 µV at 72 h increased sensitivity by 1% to 72[64–79]%. Conclusion: In comatose cardiac arrest survivors, the median N20Amp and its cutoff for predicting poor neurological outcome increase between 12–24 and 72 h after ROSC. An N20Amp decrease greater than 53% between these two timepoints predicts poor outcome with 0% FPR, confirming the unfavourable prognostic signal of a low N20Amp at 72 h.