Abstracts

Rationale: Stimulus-induced rhythmic, periodic, or ictal discharges (SIRPIDs) can present a diagnostic challenge to neurophysiologists and treating physicians, as the discharges may be interpreted as seizures and/or epileptogenic phenomena; if the electrographic pattern is unclear, clinicians face a dilemma when determining whether to and how aggressively to treat the patient with anti-seizure medications. Methods: We observed through clinical work that in several patients, a cyclical or “sputtering” termination was seen following SIRPIDs that may lend itself to detection on processed EEG (see Figure 1). To determine whether this pattern actually is a characteristic signature of SIRPIDs, we screened the Johns Hopkins Hospital continuous video-EEG (vEEG) database for patients with SIRPIDs. In these patients, the initial determination of SIRPIDs was made via serial correlations of EEG findings with video observations showing stimulation, a time-intensive process in the clinical setting. We identified consecutive patients from the continuous vEEG database with seizures for comparison.  Twenty-five patients with SIRPIDs (all of whom were adults) and 25 adult patients with seizures were included. We processed the files using Persyst 12 software (Prescott, Arizona) and identified the termination on fast fourier tranform (FFT) panels.  Two neurophysiologists blinded to the diagnosis graded the termination pattern as abrupt or cyclical/“sputtering;” when there was a disagreement, a third neurophysiologist also graded the pattern.  The grading was then compared to the clinical/neurophysiologic diagnosis. Patient characteristics (age, days of hospitalization) were compared using a Wilcoxon rank-sum test.  Etiologies, hospital locations, and outcomes were compared using a chi-squared test.  To account for the likelihood that patients with seizures had been triaged to a Neurology or Neurocritical care unit, the location distribution of patients with SIRPIDs was compared to the locations (i.e., neurology floor, medical intensive care unit) of all continuous vEEG monitoring done over 1 year. Results: The inter-rater agreement (kappa) for termination pattern was 0.64 (substantial agreement).  A cyclical or “sputtering” termination pattern on FFT had 88% sensitivity and 87% specificity for SIRPIDs, while an abrupt termination pattern on FFT had 84% sensitivity and 88% specificity for seizures.  Median patient ages were 65 (SIRPIDs) and 59 (seizures), which was not a significant difference.  The hospital day on which SIRPIDs were identified was significantly higher than the hospital day on which seizures were recorded (mean hospital days 5 and 12, p=0.03), which may reflect a delay in obtaining EEG in the absence of motor signs.  SIRPID patients were more likely to have a systemic infection or toxic/metabolic diagnosis than were seizure patients (p=0.049).  SIRPIDs were significantly more likely to be identified on non-neurological units such as the medical intensive care unit compared to the locations of all patients (p < 0.001).  Outcomes were similar in both groups. Conclusions: The termination pattern helped distinguish SIRPIDs from seizures with 88% sensitivity and 87% specificity.  Quantitative EEG provides a rapid way of visualizing and comparing the termination pattern in this challenging scenario, which may influence clinical decision-making.  The distinct terminations support a difference in the underlying mechanism of SIRPIDs and seizures; SIRPIDs may represent an abnormality of the arousal network rather than an ictal pattern. Funding: Not applicable