several factors: under-recognition of NCSE with its spontaneous resolution (thus decreasing the "denominator" of total cases that will have a poor outcome); incorrect diagnosis of NCSE based on misinterpretation of EEG "epileptiform" activity; mis-classification of certain EEG patterns as NCSE (e.g. PLEDs; triphasic waves); and grouping of different populations that have markedly different co-morbidities (ambulatory patients with NCSE together with comatose patients with electrographic seizure activity on EEG). There are almost no prospective studies with premorbid neuropsychometric studies, and retrospective studies typically include isolated cases, or case series that include conditions in which the cause of NCSE itself causes cognitive morbidity.
To summarize available data, absence status (ambulatory generalized non-convulsive status epilepticus) would appear to carry no lasting morbidity. Complex partial status epilepticus in ambulatory patients rarely results in measurable permanent neurologic deficit, although rarely, short or long-standing deficits may clearly occur.
Because intensive treatment with intravenous anticonvulsants (e.g. benzodiazepines or phenytoin) can confer morbidity, the equation has not yet been made as to whether the morbidity of such intensive treatment for all cases of NCSE exceeds the morbidity of the disease itself. Larger, prospective studies will be needed to truly determine the prognosis in the different types of NCSE, stratified according to associated degrees of impairment (minimally impaired, moderately obtunded, comatose).
While convulsive status epilepticus represents at one extreme, a striking example of a prolonged ictal state, there is another group of more subtle types of status epilepticus: namely nonconvulsive status epilepticus (NCSE). A growing body of information and some excellent recent reviews discuss the morbidity of convulsive status epilepticus (CSE); however, determining the morbidity of nonconvulsive states has proven more elusive and contentious [1-5]. Part of the contention resides in the fact that there are several conditions which have collectively been referred to as NCSE . Other problems impairing analysis of outcome arise from an ascertainment bias almost intrinsic to NCSE: that is, that these states (particularly when subtle) evade notice, and hence diagnosis [6, 7]. It is even arguable that numbers of patients in a state of NCSE for hours, following tonic-clonic seizures, may well go unrecognized, with their seizures resolving unnoticed and spontaneously behind the veil of a "post-ictal state". The issue is further complicated by problems in classification, epitomized by what is, or is not NCSE. It might seem simple, perhaps, to formulate a definition such as: a) a behavioral or cognitive change from a patient's baseline state of functioning without convulsive moments, and b) seizure activity on the EEG. But such a broad definition would subsume "PLEDS-plus"; triphasic encephalopathy; arguably some forms of waxing and waning periodic lateralized epileptiform discharges (PLEDS), or even deeply comatose patients with incidental EEG ictal activity. Such a broad grouping would therefore include at one end of the spectrum, ambulatory patients with little cognitive impairment, and at the other, comatose patients with their attendant organ systems failure in an intensive care setting with rudimentary cortical functioning as evaluated by bedside neurologic examination. This grouping would include patients with irreversible major brain damage (e.g. from strokes or trauma), and patients with clearly reversible toxic/metabolic encephalopathies (e.g. uremia, lithium toxicity).
Clearly, when problems in the definition, diagnosis, ascertainment bias, and lack of consensus on "seizure" EEG patterns along with a widely variable patient substrate converge, there will be a significant problem in determining prognosis for this syndrome.
In this review paper, I will address several issues: 1) the evolving definition of NCSE; 2) reasons for misidentification or missed diagnosis of NCSE; 3) problems in determining outcome after NCSE (distinguishing concurrent insult with its own morbid consequences from consequent insult solely attributable to seizure activity); and 4) considerations regarding the morbidity of treatment of NCSE.