When would you start antiepileptic drugs in a critically ill patient who develops myoclonic jerks but has not yet had an EEG?

Treatment depends on the setting in which these myoclonic jerks are seen.

In a non-cardiac arrest patient, such myoclonic jerks are often due to medications, organ dysfunction (e.g., uremia, etc), electrolyte imbalance, or non-convulsive seizures, etc, and workup for this is recommended with labs, head imaging, and EEG (can use rapid EEG systems that can give a quick answer). I would not recommend treating empirically unless there is EEG evidence of seizures.

In a post-cardiac arrest patient, gu...

I generally err on the side of caution by starting ASM when unclear about the etiology for myoclonic jerks, especially in the setting of cardiac arrest/concern for anoxic brain injury, especially if there are delays in obtaining an EEG. If the EEG monitoring does not reveal seizures or epileptiform discharges and a low probability for seizures, the ASM can be discontinued or weaned off.