New onset epilepsy among patients with periodic discharges on continuous electroencephalographic monitoring.

OBJECTIVE

To evaluate the incidence of new onset epilepsy and associated risk factors in patients with periodic patterns on continuous electroencephalography (cEEG) during critical illness.

METHODS

The local cEEG database and then medical records were reviewed from January 1, 2013 to June 30, 2013 to find adult patients with no history of epilepsy who had periodic discharges-either lateralized (LPDs) or generalized (GPDs)-or nonperiodic/nonepileptogenic (NP/NE) findings on cEEG and ≥3 months of clinical follow-up. Clinical seizure after discharge was the primary outcome. Chi-square test, Kruskal-Wallis test, and Cox proportional hazards models were used for statistical analysis.

RESULTS

A total of 195 patients (median age = 67.8 years) were included. There were 53 (27%), 73 (37%), and 69 (35%) patients with LPDs, GPDs, and NP/NE findings on cEEG, respectively. These three groups did not differ by demographic or clinical variables. A total of 29 (15%) patients (LPDs = 20 [38%], GPDs = 4 [6%], and NP/NE = 5 [7%]) developed epilepsy during a median follow-up of 32.1 (95% confidence interval [CI] = 13.2-42.8) months. The hazard ratio for epilepsy development among LPD patients was 7.7 (95% CI = 2.9-20.7) times compared to the NP/NE group, and the risk further increased to 11.4 (95% CI = 4-31.4) times if they also had electrographic seizures. This association remained significant despite adjusting for each covariate at a time.

SIGNIFICANCE

Patients with LPDs on cEEG during critical illness are at least seven times more likely to develop epilepsy compared to patients with NP/NE findings. This risk is further increased if patients with LPDs have electrographic seizures. In comparison, the presence of GPDs does not seem to impact the risk for developing epilepsy. cEEG findings at the time of acute insult have potential to serve as prognostic biomarkers for epilepsy development.