All in the numbers? Computational prediction of optimal anaesthetic weaning in status epilepticus

Intravenous anaesthetics are the agents of last resort for the treatment of refractory status epilepticus. However, their benefits in terms of achieving seizure suppression must be balanced against the potential harm that could result from inducing an artificial deep coma. Identifying the optimal moment for withdrawal of anaesthetics is therefore key. But whereas the indications for anaesthesia initiation have been clearly outlined, recommendations regarding anaesthesia cessation are lacking. Efforts to establish criteria for weaning anaesthetics are hampered by the diverse aetiologies of patients with refractory status epilepticus, resulting in inhomogeneous cohorts and complicating the definition of precise inclusion criteria. Enrolment of patients is further complicated by ethical issues surrounding a highly vulnerable population unable to provide consent, and funding is unlikely to be obtained from pharmaceutical companies for investigations that could potentially lead to reduced use of their compounds. In this issue of Brain, Rubin and co-workers tackle these issues by showing that the EEG, paired with quantitative measures and machine learning tools, can be used to predict successful weaning from anaesthetics in refractory status epilepticus (Rubin et al., 2020).