Literature Monitor

Can You Predict How Often Patients Will Have Seizures?

Ann Clin Transl Neurol; 2018 Jan 9; Goldenholz et al.


A mathematical formula may help determine the variability in the frequency of patients’ seizures, according to researchers from the National Institutes of Health, Beth Israel Deaconess Medical Center, and other institutions.

  • Because clinicians and researchers have yet to develop a reliable way to predict the range of each patient’s seizure counts, investigators analyzed 3 independent seizure diary databases to look for patterns.
  • The databases included 3106 entries in Seizure Tracker, 93 from the Human Epilepsy Project, and 15 from NeuroVista.
  • The analysis looked at the relationship between mean seizure frequency and the standard deviation of seizure frequency.
  • The analysis revealed that the logarithm of the mean seizure count had a linear relationship with the log of the standard deviation (R 2 >0.83).
  • Using this mathematical relationship, researchers were able to predict variability in seizure frequency with a 94% accuracy rate, compared to only 77% using traditional prediction methods.

Goldenholz DM, Goldenholz SR, Moss R, et al. Is seizure frequency variance a predictable quantity? Ann Clin Transl Neurol. 2018;5(2):201-207.

Next Article: