Should we treat patients with impaired consciousness and periodic patterns on EEG?

The significance of periodic EEG patterns in patients with impaired consciousness is controversial. We aimed to determine if treating these patterns influences clinical outcome.

We studied all patients who had periodic discharges on their EEG recordings from January 2007 to December 2009. Patients with clinical seizures within the preceding 24h, or with unequivocal electrographical seizure activity were excluded. Logistic regression was performed to analyze for factors associated with (a) mortality (b) functional status (c) resolution of EEG pattern.

Of the 4246 patients who had EEG, 111 (2.6%) had periodic EEG patterns. 64 met inclusion criteria. In adjusted analysis, higher mortality was associated with acute symptomatic etiology (OR 17.74, 95% CI 1.61-196.07, p=0.019), and presence of clinical seizures (OR 4.73, 95% CI 1.10-20.34, p=0.037). For each unit decrement of GCS, the odds of inpatient mortality and a poorer functional state on discharge increased by 23% (95% CI 7-37%, p=0.009) and 33% (95% CI 9-51%, p=0.011) respectively. Administration of abortive therapy was an independent risk factor for poorer functional status on discharge (adjusted OR 41.39, 95% CI 2.88-594.42, p=0.006), while patients with history of pre-existing cerebral disease appeared more likely to return to baseline functional status on discharge (unadjusted OR 5.00, 95% CI 1.40-17.86, p=0.013).

Treatment of periodic EEG patterns does not independently improve clinical outcome of patients with impaired conscious levels. Occurrence of seizures remote to the time of EEG and lower GCS scores independently predict poor prognoses.