Postictal language dysfunction in patients with right or bilateral hemispheric language localization

Michael Privitera, Christian Kohler, William T Cahill, Hwa Shain Yeh

Research output: Contribution to journalArticlepeer-review

16 Scopus citations


Purpose: As shown previously, when temporal lobe complex partial seizures (TLCPS) originate from the language dominant hemisphere, patients cannot read a test phrase correctly within 60 s of the end of the total discharge. We wished to assess whether postictal language testing results discordant with this pattern identified patients with non-left (right hemisphere or mixed) language dominance. Methods: Since 1988, all patients undergoing video/EEG monitoring at our institution have been given a test phrase to read aloud as soon as a seizure is detected. Inclusion criteria for this study were (a) postictal language testing within 60 s of seizure end for at least one TLCPS, (b) >90% seizure reduction after anterior temporal lobectomy with 2-year minimum follow-up, and (c) language localization by either intracarotid amobarbital test (IAT) or direct electrical stimulation of left hemisphere cortex. Results: Two hundred twenty-four seizures in 64 patients were analyzed. Discordant postictal language patterns were noted in 10 of 11 patients with IAT documented non-left language dominance and in 15 of 53 with left dominance (p = 0.006; sensitivity 90.9%, specificity 71.7%). Conclusions: Postictal language testing accurately identifies patients with non-left language dominance and may be useful in selecting appropriate patients for IAT.

Original languageEnglish (US)
Pages (from-to)936-941
Number of pages6
Issue number10
StatePublished - 1996
Externally publishedYes


  • Epilepsy
  • Intracarotid amobarbital test
  • Language
  • Seizure
  • Speech

ASJC Scopus subject areas

  • Clinical Neurology
  • Neuroscience(all)


Dive into the research topics of 'Postictal language dysfunction in patients with right or bilateral hemispheric language localization'. Together they form a unique fingerprint.

Cite this