Abstracts

Rationale: Depression is the most common, yet under recognized & under treated, psychiatric co-morbidity in patients with epilepsy (PWE). One of the most significant obstacles to treatment for depression may be the failure of adequate screening and diagnosis. This study was designed to: 1)- evaluate the utility of a depression screening tool, the Patient Health Questionnaire (PHQ-9), in identifying comorbid depression in PWE: and 2)- compare the performance of PHQ-9 to the Neurological Disorders Depression Inventory for Epilepsy (NDDI-E), the only depression screening tool that has been specifically validated for PWE. The rationale for this validation is similarly two-fold: First, while NDDI-E is an excellent depression screening tool specific for PWE, its usefulness in a general neurology practice is unclear; and second, while PHQ-9 is conversely a commonly used tool to screen for the presence and severity of depression, it has not been specifically validated in PWE, where multiple somatic, cognitive and/or medication related side effects may confuse the picture. Methods: This is a prospective study that included 235 patients with clinically/EEG diagnosed epilepsy who completed the MINI-International Neuropsychiatric Interview (MINI) to assess diagnostic criteria for major depressive disorder (MDD). A current diagnosis of MDD on MINI defined a depressed patient. Each patient also completed the PHQ-9 & NDDI-E in random order within the following 24 hours. The sensitivity & specificity of the PHQ-9 at cut-points defined in the literature (1-4: minimal, 5-9: mild, 10-14: moderate, 15-19: moderately severe, 20-27: severe depression) was estimated & 95% C.I. constructed. Various PHQ-9 score cut-points for a PHQ-9 positive screen of 10 thru 15 were compared to NDDIE score of >=15 ( NDDIE-positive ). Clinical, demographic, imaging, and EEG patient characteristics were collected.Results: The mean age of our cohort was 42 years (range 19-78), and 69% were women. Most (90%) had focal epilepsy, and overall mean epilepsy duration was 20years (range 0.5-63 yrs). With a score cut-off of 10, the PHQ-9 was 91% sensitive, and 81% specific; whereas a higher cut-off of 15 improved specificity to 95% at the expense of sensitivity (Table 1). Among the 96 patients with data on MINI, PHQ-9, and NDDIE, the estimated ROC area of the PHQ-9 was 0.923 (Standard Error, SE, = 0.036, 95% Wald CI [0.868, 1.0]). The estimated ROC area of the NDDIE scores was 0.957 (SE = 0.023, 95% Wald CI [0.911, 1.0]) (Figure 1). This difference in ROC area is not statistically significant (p = 0.547)Conclusions: PHQ-9 is a reasonable tool to screen for depression in PWE, with excellent sensitivity and acceptable specificity at cut-point >=10. Its ease of administration, and widespread validated use in primary care make PHQ-9 an attractive tool to adequately screen for depression both in specialized epilepsy centers and in general neurology practices.