Failure to record typical events during long-term video/EEG monitoring (LTM) leaves the diagnosis uncertain. The purpose of this study was to analyze predictors of an initial nondiagnostic LTM study and to evaluate the yield of a repeat study.

We reviewed all adult LTM studies performed at Vanderbilt University from January 2004 to June 2008. We identified 150 patients with no typical events on LTM and matched them with consecutive controls with typical events. We compared patient demographics, epilepsy risk factors, history of antiepileptic use/failures, and prior EEG and MRI results.

Of 2397 LTM studies, 380 (15.8%) failed to record typical events. Absence of epilepsy risk factors and normal outpatient EEG predicted this outcome. A repeat LTM study was successful in 18 of 45 patients (42%).

The prediction of a nondiagnostic LTM can help to reduce the cost of investigating spells of unknown nature.