When should induction protocols be used in the diagnostic evaluation of patients with paroxysmal events?

Purpose: To determine the timing of spontaneous psychogenic nonepileptic events (PNEE) during video-EEG telemetry (VEEG), and the need to use induction protocols (IP). Methods: We studied 100 consecutive patients (75 females, 25 males) admitted to our inpatient VEEG unit from July 1994 to June 1996 for differential diagnosis of paroxysmal events. We recorded the time to the first diagnostic spontaneous event, identified by the patient or a family member as typical. Episodes were classified as PNEE, physiologic nonepileptic events (PhysNEE), and epileptic seizures (ES). Results: The mean duration of VEEG was 74 ± SD 54.1 h. In 82 patients, a diagnostic event occurred spontaneously. The first event was an ES in 22 patients, a PNEE in 53, and a PhysNEE in 7. The time to first diagnostic event was significantly shorter for PNEE than for ES [15.0 ± SD 16.3 h (range 5 min to 58 h) vs. 28.6 ± SD 34.0 h (range 1-110 h) F = 15.621, p < 0.0001]. In the first 24 h, 77.4% of the patients with PNEE had an event. By 48 h, all but 2 (96.2%) had had diagnostic events. After the first 58 h of monitoring, all patients with PNEE experienced a spontaneous diagnostic event. Conclusion: Spontaneous events can be expected to occur within 48 h in most patients with PNEE. Therefore, if IP are to be used as a diagnostic tool, we suggest that they be withheld during the initial 48 h of VEEG monitoring.