A practical risk score for EEG seizures in hospitalized patients

The American Academy of Neurology (AAN) 70th Annual MeetingA Struck, B Ustun, AR Ruiz, J Lee, S LaRoche, L Hirsch, E Gilmore, J Vlachy, H Haider, C Runi, M Westover | April 23, 2018

Advertisement

Summary: Researchers sought to create a simple scoring system with which to predict the probabiity of electrographic seizures using seizure risk factors--including epileptiform and periodic discharges--from continuous EEG monitoring (cEEG) in patients with acute illness. They concluded that the 2HELPS2B score makes possible an accurate prediction of seizures by adding points for six readily available variables from patient history and initial EEG.

Methods:

  • Researchers used a prospective database to derive a dataset containing 24 clinical and electrographic variables for 5427 > 24hr cEEG sessions, with which they created a scoring system model to estimate seizure risk in patients undergoing cEEG.
  • The model was built using a new machine learning method (RiskSLIM) designed to produce accurate, risk-calibrated, scoring systems with a limited number of variables and small integer weights.
  • Researchers validated the predictive accuracy and risk-calibration of this model with cross-validation, and compared its performance to models built with state-of-the-art logistic regression methods.

Results:

  • Researchers found that their final model (2HELPS2B) had an AUC of 0.821 and average calibration error of 2.7%.
  • The model includes six variables with the following point assignments:
    • (i) brief potentially ictal rhythmic discharges (B(I)RDs) (2 points);
    • (ii) presence of LPD, LRDA, or BIPDs (1 point);
    • (iii) prior seizure (1 point);
    • (iv) sporadic epileptiform discharges (1 point);
    • (v) frequency > 2.0Hz for any periodic or rhythmic pattern (1 point);
    • (vi) presence of “plus” features (superimposed, rhythmic, sharp, or fast activity) (1 point).
  • Corresponding predicted seizure risk for each score was: 0: 5%, 1: 12%, 2: 27%, 3: 50%, 4: 73%, 5: 88%, 6-7: > 95%.


This study was supported by a Research Infrastructure award from the American Epilepsy Society and the Epilepsy Foundation.

Advertisement