Cocito et al have raised an important consideration in the differential diagnosis of patients with altered mental status. Recent data demonstrate that NCSE is more common than previously appreciated and may not be accompanied by any physical manifestations of continuous seizure activity. Our patients were not routinely studied with EEGs, so we cannot comment on the incidence of NCSE in our population. No large-scale prospective study of confused or delirious patients has been conducted using EEG to identify NCSE. Cocito et al have observed NCSE in 6% (5/84) of their cancer patients with coma or delirium. Other reports reveal that comatose patients have an 8% incidence of unrecognized NCSE1; we had only 3 (2%) comatose patients in our series of 140 patients. Likewise, only 9% of our patients (13/140) had seizures as a manifestation of their encephalopathy, and no patient had convulsive status epilepticus, with which the incidence of subsequent nonconvulsive status can be as high as 14%.2 These are important risk factors for NCSE, and they were present in a small proportion of our patients or not at all. Furthermore, it is notable that 67% of our patients recovered in the absence of any specific treatment focused on controlling prolonged seizure activity. These data do not eliminate NCSE as a factor in our patients, but they reduce the probability. Nevertheless, NCSE remains an important consideration in any patient with prolonged unexplained impairment of mental status, and EEG can be diagnostic.
In reply. Arch Neurol. 2001;58(8):1310. doi: