Abstracts

Rationale: Assess feasibility of computer-assisted health status data acquisition in a high-volume tertiary epilepsy center to determine prevalence and predictors of co-morbid depression. Methods: Subjects included 2,704 patients seen between Nov 2008 and Mar 2010 in a tertiary epilepsy outpatient setting. Patients used touch-screen computer technology to enter responses to standardized surveys of seizure severity and depression: the Liverpool Seizure Severity Scale (LSSS) and the Patient Health Quality-9 (PHQ-9), respectively. Data from one visit per patient were used and included demographics (age, sex, marital status, and race); driving status (driving, not driving, or not driving for reasons other than epilepsy); epilepsy type (focal, generalized, or undefined); seizure control (controlled vs. intractable); and number of antiepileptic drugs (AEDs) used. The PHQ-9, validated in general medical populations, has been shown to have significant correlation with a DSM-IV diagnosis of major depressive disorder when the total score is >9. PHQ-9 total score for the nine items ranges from 0 to 27. Scores of 5, 10, 15, and 20 represent cutpoints for mild, moderate, moderately severe, and severe depression, respectively. Both univariate and multivariate data analyses were used for independent risk factor determination. Results: 2,128 patients (79%) completed the PHQ-9. Mean age was 43.8 (median, 42; range, 19-98; SD, 15.7); 54% were female; 83% were white, 13% black and 4% other; and marital status of most patients (89%) was either single or married with the rest indicating divorce or some other marital status. Mean PHQ-9 for the entire group was 6.4 (range, 0-27, median 4). 561 patients (26.4%) had a PHQ-9 score of >9 with 126 (5.9%) scoring in the severe range. Demographic data were complete for all 2,128; data on epilepsy type and seizure control data were available in only 1,827 (84%), seizure severity in 2,067 (97%), number of AEDs used in 2,024 (95%), and driving status in 2023 (95%). Univariate analysis demonstrated significant associations between depression and all variables listed above. Following multivariate analysis only marital status (p<.0001), seizure severity (p<.0001), and driving status (p<.001) retained independent risk factor status. Seizure control showed a trend in this direction, but did not achieve statistical significance (p=0.0791). All data are summaried in the Table. Conclusions: A nearly 80% completion rate supported the utility of touch-screen computer technology for routine health status assessment in epilepsy outpatients. Roughly one-quarter of all patients endorsed at least a moderate degree of self-rated depression (PHQ-9 >9) with only 6% rating themselves as severely depressed (PHQ-9 >19). Only marital status, driving status and seizure severity proved to be independent predictors of depression whereas number of AEDs prescribed, epilepsy type, seizure control, and race did not. Notably, AED burden was not an independent predictor of depression, a finding that contrasts with the FDA s black-box warning. The large number of study patients adds weight to the significance of these findings.