A critical appraisal on the utility of long-term video-EEG monitoring in older adults.

BACKGROUND

World-over, the majority of patients undergoing video-EEG monitoring (VEM) are in their second or third decades. Although elderly represent the fastest growing segment of population with epilepsy, only few of them undergo VEM. We critically evaluated the utility of VEM in the diagnosis and long-term management of older adults with paroxysmal behavioral events.

METHODS

148 consecutive patients aged 45 and above, who underwent long-term (≥8 h) inpatient VEM from 1996 to 2009 formed the study cohort. Utilizing a structured proforma, we gathered their demographic, clinical, electrophysiological and long-term outcome data. VEM was considered as "useful" when it changed the diagnosis and/or management; it was "corroborative" when it helped the treating physician to confirm the diagnosis and "not useful" when it neither helped to improve the diagnosis nor the management.

RESULTS

The mean age was 51.3 (SD 6.4) years; mean duration of VEM was 69.3 h. Out of 117 patients with a diagnosis of epileptic seizures referred for presurgical evaluation or classification, VEM was "useful" or "corroborative" in 111 patients (94.8%; p=0.0001). It was also "useful" or "corroborative" in 29 out of 31 patients (93.5%) referred with a suspicion of associated or pure psychogenic non-epileptic seizures (p=0.0001). None developed any complications during monitoring. At a mean follow-up of 37.7 months there was significant reduction in AED usage in patients with epilepsy (p=0.0001) and epilepsy with associated PNES (p=0.001). At a mean follow-up of 34.2 months, all patients with pure PNES were event-free and medication-free at last follow-up (p=0.002). Twenty-three patients (19.6%) underwent surgery, all except one remaining seizure-free at a mean follow-up of 39.2 months.

CONCLUSIONS

VEM is a safe and cost-effective investigation strategy in older-adults. It aided in improving the diagnosis, offered better treatment including surgery and helped in excluding non-epileptic paroxysmal events in majority.