Accuracy of Clinical Neuropsychological Versus Statistical Prediction in the Classification of Seizure types

Abstract

Many studies have reported on the pattern of neuropsychological test performance across varied seizure diagnosis populations. Far fewer studies have evaluated the accuracy of the clinical neuropsychologist in formulating an impression of the seizure diagnosis based on results of neuropsychological assessment, or compared the accuracy of clinical neuropsychological judgment to results of statistical prediction. Accuracy of clinical neuropsychological versus statistical prediction was investigated in four seizure classification scenarios. While both methods outperformed chance, accuracy of clinical neuropsychological classification was either equivalent or superior to statistical prediction. Results support the utility and validity of clinical neuropsychological judgment in epilepsy treatment settings.