Abstracts

Traditional neuropsychological testing may fail to detect abnormalities in declarative memory tests in epileptic patients. One possible explanation is that such tests do not adequately measure memory in real life activities. Prospective memory (PM), i.e. cognitive abilities related to remembering a planned intention in the correct setting, may be a more ecological measurement of memory function in daily activities. We studied the performance of patients with left and right mesial temporal sclerosis and normal controls in a prospective memory battery and compared results with performance in the Rey Auditory Verbal Learning Test (RAVLT) and with a self assessment questionaire for memory impairment (QMI). The battery consisted of six time and event related tasks to be recalled in the adequate situation, such as remembering to tell the examiner that they needed a new prescrition after fifteen minutes and remembering to ask for an personal object taken by the examiner at the end of a 105 minute session.Patients were submitted a neuropsychological test battery during the session, including the RAVLT and were asked to fill the QMI (24 items). Spontaneous and cued recall of the 15 RAVLT words was tested after 7 days. Performance in the PM tests was compared to RAVLT scores and to QMI score. A patient was considered impaired in either MP or RAVLT if performance fell below 2SD of controls. Significant memory compaints were considered if QMI score was 1SD above controls. Sensitivities and specificities for impairment measured by PM and RAVLT were calculated in relation to QMI assessed impairment (gold standard). We studied 28 patients with mesial temporal sclerosis (MTS) (17 left) and 18 normal controls, matched by age (38+/-10 vs 40+/-12 yrs) and education (11+/-2 vs 11+/-2)(age 40 +/-2). Left MTS patients performed significantly worse than controls in PM and spontaneous recall and in QMI score (p[lt]0.05). Both left and right MTS performed significantly worse than controls in 7-day cued recall (p[lt]0.05). Sensitivities in relation to QMI were 56.2% for PM, 50% for RAVLT, 37.5% RAVLT and PM and 68,8% for RAVLT or PM. Specificities were 100% for PM, 83.3% for RAVLT, 100% for RAVLT and PM and 91.6% for PM or RAVLT measured impairment. This prospective memory test was able to differentiate memory performance between left MTS patients and controls. It is has a better specificity than RAVLT in relation to a self assessment memory questionaire. Its sensitivity is increased when used in combination with RAVLT. Further studies should better delineate its use in combination with other tests in the assessment of memory dysfunction in epileptic patients and in patients with memory impairment of other etiologies.