PHAs following a single FBTCS/GTCS in the EMU are significantly associated with the duration of ictal/postictal hypoxemia. It is possible that FBTCS/GTCS-associated hypoxemia may trigger fatal cardiac arrhythmias in a subset of susceptible patients dying of sudden unexpected death in epilepsy.

Summary Objective No biomarkers reliably predict risk for sudden unexpected death in epilepsy (SUDEP). Postictal generalized electroencephalography (EEG) suppression (PGES) is a possible biomarker for SUDEP risk. However, its utility in predicting SUDEP remains uncertain. We had observed that postictal tonic electromyography (PTEMG) activity follows some generalized convulsive seizures (GCS). PTEMG activity and PGES may have a common pathophysiologic basis. PGES is associated with periictal respiratory distress. There is evidence that tonic EMG occurs with brain hypoxia. Thus PTEMG activity may be related to seizure‐associated hypoxemia. Pronounced variation occurs among expert clinicians in identifying PGES, thereby limiting its utility as a biomarker. Characteristics of PTEMG activity and its relationship to preceding GCS have not been explored. We studied PTEMG activity characteristics, its relationship to the preceding seizure and associated respiratory dysfunction. Methods We reviewed 145 GCS in 66 patients undergoing video‐EEG telemetry (VET). The presence of PTEMG activity was defined when tonic EMG occurred for at least 3 seconds following seizure termination and was identified with filter settings at 5‐200 Hz. Duration of PTEMG activity, the seizure, PGES, seizure‐associated peripheral capillary oxygen saturation (SpO2) change, and end‐tidal CO2 were analyzed. We compared data from GCS with and without PTEMG activity. Results Ninety of 145 seizures with GCS had PTEMG activity. The remainder had postictal slowing without PTEMG activity, and cessation of activity was followed by EEG slowing. Duration of the initial PTEMG discharge was 39.1 (mean) ± (standard deviation) 17.9 seconds. SpO2 nadir was lower (P = 0.005) in seizures with PTEMG activity than in those without (72% vs 77%). End‐tidal CO2 was higher (P = 0.05) in seizures with PTEMG activity than in those without (63  vs 56 mm Hg). PGES duration was 35.6 ± 22.2 seconds and associated with duration of PTEMG activity (P < 0.001). Significance The novel finding is that PTEMG activity occurs following 62% of GCS and that seizures with PTEMG activity have greater severity of respiratory dysfunction than seizures without. PTEMG activity is readily discerned by visual analysis of VET at appropriate filter settings and has the potential of being a complementary or surrogate biomarker of PGES for assessing SUDEP risk.