Rationale:Our goal was to determine the frequency of repeated intracarotid amobarbital test (IAT) at our center and to estimate the retest reliability of the IAT for both language and memory lateralization.

Methods: A total of 1,249 consecutive IATs on 1,190 patients were retrospectively reviewed for repeat tests.

Results: In 4% of patients the IAT was repeated in order to deliver satisfactory information on either language or memory lateralization. Reasons for repetition included obtundation and inability to test for memory lateralization, inability to test for language lateralization, no hemiparesis during first test, no aphasia during first test, atypical vessel filling, and bleeding complications from the catheter insertion site. Language lateralization was reproduced in all but one patient. Repeated memory test results were less consistent across tests, and memory lateralization was unreliable in 63% of the patients.

Discussion: In spite of test limitations by a varying dose of amobarbital, crossover of amobarbital from one side to the other, testing of both hemispheres on the same day, practice effects, unblinded observers, fluctuating cooperation of the patients, and a biased sample of patients language lateralization was reproduced in all but one patient. In contrast, repeated memory test results were frequently contradictory. Memory results on IAT therefore seem much less robust than the results of language testing. Gain of reliable information versus the risks of complications and failed tests has to be considered when a patient is subjected to an IAT.