Abstracts

Rationale: The risk of generalized tonic-clonic seizures (GTCs)is relevant to the safety of patients during epilepsy monitoring, during which drugs are often reduced. High risk may extend into the early post-discharge period, before full medical control is re-established. We searched for factors predicting GTC occurrence during these times. Methods: We conducted a chart analysis and phone interview of 100 patients from two cohorts of 50 consecutive patients admitted to the University of Alabama Hospital seizure monitoring unit who met criteria: 1) history of GTCs before admission or observed GTCs during monitoring 2)able to provide a history of events during the 72 hours after discharge and 3) no psychogenic seizures during monitoring. Data on suspected preadmission seizure types, inhospital seizures, medication reductions, time from last inhospital GTC to discharge, time from medication restart to discharge, and use or nonuse of short-term lorazepam (LZP) were collected. Patients were surveyed by telephone to determine seizure occurrences, emergency visits, and rehospitalizations within the 72 hours after discharge. Results: 442 patients were admitted during the period of observation. Of the 98 patients who had a history of GTCs before admission, 55 (56%)had at least one GTC inhospital. In comparison, only 2 of 344 patients admitted during this time without a history of GTC had a documented GTC inhospital (p <0.001). 107 of these 344 had observed psychogenic events. Two of the 100 patients in the study had an early posthospital GTC- within 3 days after discharge, though 11 others had complex partial and 2 others simple partial seizures during these 72 hours. 7 of 46 receiving LZP (usually 1 mg TID for 3 days) had early posthospital seizures vs. 8 of 54 who did not (NS). The mean length of stay for the 100 patients in the study, none of whom had psychogenic seizures, was 3.3 days. No evidence of premature discharge was found among those with early posthospital seizures. Conclusions: Patients who have a history of GTCs are at very high risk for GTCs during seizure monitoring, and should be watched carefully. Those without such a history are unlikely to have a GTC precipitated. However, we recognize that some patients with a history of GTCs may be found to have GTC-like psychogenic events. Ideally, patients with any GTC history should be under direct observation by staff or family during monitoring. A brief course of LZP made no difference in the rate of early posthospital seizures, though doses used were low.