Abstracts

The Epilepsy Foundation of America Epilepsy Concerns Index (Gilliam et al, 1999) is a 20-question scale in which patients rate epilepsy-specific concerns on a 5-point scale. Although this scale is increasingly used to characterize both disease burden and treatment outcomes, the scores from the different questions are either studied individually or summed into a single measure to reflect overall epilepsy concerns. We conducted Principal Components Analysis (PCA) to better understand the dimensions being assessed by this scale. We performed PCA of the EFA Concerns Index scores of 189 patients undergoing evaluation for epilepsy surgery that were retrospectively identified. The patients averaged 34 years of age (SD=11) with an average age of habitual seizure onset of 17 years (SD=13). Factor scores were then compared to independent measures including the QOLIE-89, Personality Assessment Inventory (PAI), and neuropsychological measures of IQ and memory using multiple linear regression. Based upon the scree plot, the best model consisted of a 5 factor solution in which there were no varimax-rotated factors represented by fewer than 2 variables. The first factor appears to reflect affective impact on enjoyment of life and is identified by questions 16, 18, 19, 20, and to a smaller degree, questions 14 and 17. The second factor is defined by questions 1, 4, 5, 6, and 12, and appears to reflect concerns regarding general autonomy. Factor 3 is defined by 2, 7, 14, and 15 (questions 14 and 15 also load on factor 1) and reflects a fear of seizure occurrence. Factor 4 is defined by questions 9, 10, and 11, and reflects concern of being a burden to ones family. Factor 5 is defined by questions 8 and 13, and appears to reflect a perceived lack of understanding by others. Only a single question (3, having to take medication) failed to load on any of the factors. Multiple regression showed that the EFA factor scores were predicted by seven scales from the PAI, seven scales from the QOLIE-89, Performance IQ, and measures of verbal and visual memory. By contrast, only two PAI scores and four QOLIE-89 scores were associated with a single, summary EFA score. This analysis confirms that the EFA Concerns Index is multidimensional, and that using a global score based upon all items may mask important specific concerns that may be relevant when applied to individual patients. Affective concerns is the strongest underlying factor, followed by autonomy, fear of seizures, family burden, and lack of understanding. These data support the validity and clinical utility of scoring the EFA Index by the 5 factors we have identified rather than relying on a single summary score.
(Gilliam F, et al. Patient-oriented outcome assessment after temporal lobectomy for refractory epilepsy. Neurology. 1999;53:687-694).